InfoVis:Wiki - User contributions [en]

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T16:57:43Z

UE-InfoVis0506 9335259: /* The Basic Algorithm */

== Information Murals ==

=== Basics ===

{{Quotation|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazine, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available screen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of a vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appending elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

The initial point for the basic algorithm is an input-image (of M x N elements) which is scaled into a mural of I x J pixels. On the one hand the algorithm needs the data structures which store the information, on the other hand it requires an I x J array of floats.

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a colourscale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Information as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in an elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural', s.a.: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazine, 1993] BYTE-Magazine, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T16:57:30Z

UE-InfoVis0506 9335259: /* The Basic Algorithm */

== Information Murals ==

=== Basics ===

{{Quotation|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazine, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available screen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of a vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appending elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

The initial point for the basic algorithm is an input-image (of M x N elements) which is scaled into a mural of I x J pixels. On the one hand the algorithm needs the data structures which store the information, on the other hand it requires an I x J array of

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a colourscale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Information as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in an elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural', s.a.: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazine, 1993] BYTE-Magazine, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T16:56:21Z

UE-InfoVis0506 9335259: /* The Basic Algorithm */

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T16:55:45Z

UE-InfoVis0506 9335259: /* The Basic Algorithm */

== Information Murals ==

=== Basics ===

{{Quotation|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazine, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available screen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of a vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appending elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

The initial point for the basic algorithm is an input-image (of M x N elements) which is scaled into a mural of I x J pixels. On the one hand the algorithm needs the data structures which store the information, on other hand it requires an I x J array of floats.

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a colourscale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Information as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in an elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural', s.a.: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazine, 1993] BYTE-Magazine, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T13:37:44Z

UE-InfoVis0506 9335259: /* Earthlight-Information as 'Realmural' */

== Information Murals ==

=== Basics ===

{{Quotation|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazine, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available screen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of a vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appending elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a colourscale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Information as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in an elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural', s.a.: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazine, 1993] BYTE-Magazine, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T13:01:43Z

UE-InfoVis0506 9335259: /* Basics */

== Information Murals ==

=== Basics ===

{{Quotation|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazine, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available screen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of a vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appending elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a colourscale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Information as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural', s.a.: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazine, 1993] BYTE-Magazine, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4

2006-02-01T12:59:53Z

UE-InfoVis0506 9335259:

== '''Gruppe G4''' ==

Die Gruppe G4 besteht bis jetzt aus den folgenden Mitgliedern: 
The following people are part of the G4 group by now:

* [[User:UE-InfoVis0506_0026030|Muster, Anna]]
* [[User:UE-InfoVis0506_9335259|Puchta, Jürgen]]
* [[User:UE-InfoVis0506_0026754|Rainer, Christian]]
* [[User:UE-InfoVis0506_0026629|Sölder, Christoph]]

---- 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 0|Aufgabe 0]] 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 1|Aufgabe 1]] 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 2|Aufgabe 2]] 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3|Aufgabe 3]] 

[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe|Zusatzaufgabe - 'Information Murals']]

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4

2006-02-01T12:59:35Z

UE-InfoVis0506 9335259:

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4

2006-02-01T12:59:22Z

UE-InfoVis0506 9335259:

== '''Gruppe G4''' ==

Die Gruppe G4 besteht bis jetzt aus den folgenden Mitgliedern: 
The following people are part of the G4 group by now:

* [[User:UE-InfoVis0506_0026030|Muster, Anna]]
* [[User:UE-InfoVis0506_9335259|Puchta, Jürgen]]
* [[User:UE-InfoVis0506_0026754|Rainer, Christian]]
* [[User:UE-InfoVis0506_0026629|Sölder, Christoph]]

---- 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 0|Aufgabe 0]] 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 1|Aufgabe 1]] 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 2|Aufgabe 2]] 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3|Aufgabe 3]] 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe|Zusatzaufgabe - Information Murals]]

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:58:42Z

UE-InfoVis0506 9335259:

== Information Murals ==

=== Basics ===

{{Quotation|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazine, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available screen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appending elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a colourscale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Information as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural', s.a.: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazine, 1993] BYTE-Magazine, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:56:37Z

UE-InfoVis0506 9335259: /* Earthlight-Informations as 'Realmural' */

== Information Murals ==

=== Basics ===

{{Quotation|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazine, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available screen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appending elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a colourscale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Information as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural', s.a.: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazine, 1993] BYTE-Magazine, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:55:50Z

UE-InfoVis0506 9335259:

== Information Murals ==

=== Basics ===

{{Quotation|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazine, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available screen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appending elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a colourscale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural', s.a.: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazine, 1993] BYTE-Magazine, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:54:26Z

UE-InfoVis0506 9335259: /* Earthlight-Informations as 'Realmural' */

== Information Murals ==

=== Basics ===

{{Quotation|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazine, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available screen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appending elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a colourscale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural', s.a.: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazine, 1993] BYTE-Magazine, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:54:16Z

UE-InfoVis0506 9335259: /* Earthlight-Informations as 'Realmural' */

== Information Murals ==

=== Basics ===

{{Quotation|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazine, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available screen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appending elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a colourscale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural' s.a.: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazine, 1993] BYTE-Magazine, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

File:Earthlights dmsp.jpg

2006-02-01T12:54:03Z

UE-InfoVis0506 9335259: /* Beschreibung */

== Beschreibung ==
Figure 3: 'Realmural', s.a.: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

== Copyright-Status ==

== Quelle ==
[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

File:Earthlights dmsp.jpg

2006-02-01T12:53:40Z

UE-InfoVis0506 9335259: /* Beschreibung */

== Beschreibung ==
Figure 3: 'Realmural', see: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

== Copyright-Status ==

== Quelle ==
[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

File:Earthlights dmsp.jpg

2006-02-01T12:53:31Z

UE-InfoVis0506 9335259: /* Beschreibung */

== Beschreibung ==
Figure 3: 'Realmural' see: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

== Copyright-Status ==

== Quelle ==
[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:53:18Z

UE-InfoVis0506 9335259:

== Information Murals ==

=== Basics ===

{{Quotation|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazine, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available screen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appending elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a colourscale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural' see: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazine, 1993] BYTE-Magazine, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:42:35Z

UE-InfoVis0506 9335259: /* Basics */

== Information Murals ==

=== Basics ===

{{Quotation|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazin, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available sreen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appendant elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a color scale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural' nach: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazin, 1993] BYTE-Magazin, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:42:17Z

UE-InfoVis0506 9335259: /* Basics */

== Information Murals ==

=== Basics ===

{{FloatingQuote|... Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form. ...| [BYTE-Magazin, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available sreen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appendant elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a color scale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural' nach: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazin, 1993] BYTE-Magazin, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:40:38Z

UE-InfoVis0506 9335259: /* Basics */

== Information Murals ==

=== Basics ===

{{Quotation|Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form.| [BYTE-Magazin, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available sreen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appendant elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a color scale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural' nach: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazin, 1993] BYTE-Magazin, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:40:23Z

UE-InfoVis0506 9335259: /* Basics */

== Information Murals ==

=== Basics ===

{{Quotation|Data visualization is about comprehension, not graphics. Think of it as a range of techniques that enable you to display abstract numerical data and statistics in graphical form.| BYTE-Magazin, 1993}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available sreen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appendant elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a color scale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural' nach: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazin, 1993] BYTE-Magazin, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:38:48Z

UE-InfoVis0506 9335259: /* Basics */

== Information Murals ==

=== Basics ===

{{Quotation|Data visualization is about comprehension, not graphics.
Think of it as a range of techniques that enable you to display
abstract numerical data and statistics in graphical form.| [BYTE-Magazin, 1993]}}

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available sreen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appendant elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a color scale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural' nach: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazin, 1993] BYTE-Magazin, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:36:37Z

UE-InfoVis0506 9335259:

== Information Murals ==

=== Basics ===

'Data visualization is about comprehension, not graphics.
Think of it as a range of techniques that enable you to display
abstract numerical data and statistics in graphical form.' [BYTE-Magazin, 1993]

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available sreen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appendant elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a color scale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Earthlights dmsp.jpg|Earthlights taken under perfect atmospheric conditions]]

Figure 3: 'Realmural' nach: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazin, 1993] BYTE-Magazin, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:34:36Z

UE-InfoVis0506 9335259: /* Earthlight-Informations as 'Realmural' */

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:33:20Z

UE-InfoVis0506 9335259: /* Sun Spots */

== Information Murals ==

=== Basics ===

'Data visualization is about comprehension, not graphics.
Think of it as a range of techniques that enable you to display
abstract numerical data and statistics in graphical form.' [BYTE-Magazin, 1993]

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available sreen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appendant elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|Sunspots recorded per month]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a color scale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Sunspot.gif|Sunspots recorded per day]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Occam.jpg|William of Ockham]]

Figure 3: 'Realmural' nach: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazin, 1993] BYTE-Magazin, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:32:31Z

UE-InfoVis0506 9335259: /* Sun Spots */

== Information Murals ==

=== Basics ===

'Data visualization is about comprehension, not graphics.
Think of it as a range of techniques that enable you to display
abstract numerical data and statistics in graphical form.' [BYTE-Magazin, 1993]

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available sreen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appendant elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Sunspot_mo.gif|William of Ockham]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a color scale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Occam.jpg|William of Ockham]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Occam.jpg|William of Ockham]]

Figure 3: 'Realmural' nach: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazin, 1993] BYTE-Magazin, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

File:Earthlights dmsp.jpg

2006-02-01T12:31:14Z

UE-InfoVis0506 9335259: Figure 3: 'Realmural' nach: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

== Beschreibung ==
Figure 3: 'Realmural' nach: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]
== Copyright-Status ==

== Quelle ==
[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

File:Sunspot mo.gif

2006-02-01T12:29:47Z

UE-InfoVis0506 9335259: Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

== Beschreibung ==
Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]
== Copyright-Status ==

== Quelle ==
[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

File:Sunspot.gif

2006-02-01T12:28:29Z

UE-InfoVis0506 9335259: /* Quelle */

== Beschreibung ==
Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

== Copyright-Status ==

== Quelle ==
[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

File:Sunspot.gif

2006-02-01T12:27:50Z

UE-InfoVis0506 9335259: Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

== Beschreibung ==
Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

== Copyright-Status ==

== Quelle ==
http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe

2006-02-01T12:05:59Z

UE-InfoVis0506 9335259:

== Information Murals ==

=== Basics ===

'Data visualization is about comprehension, not graphics.
Think of it as a range of techniques that enable you to display
abstract numerical data and statistics in graphical form.' [BYTE-Magazin, 1993]

In case of displaying very large information spaces traditional information visualizations reach their limits rather fast, often right from the design or planning stage. Without specific techniques and tools respectively the considered information is often useless. The primary goal is to provide effective display and access mechanisms. According to the vast number of data in large information spaces it is a real challenge to create effective presentations using only the limited number of pixels on a screen.

The Information Mural technique is a possibility for 2D visual representation even if the number of data-elements greatly rises above the available sreen-pixels. This technique was invented by John Stasko and Dean Jerding at the 'Georgia Institute of Technology'. Other current procedures for depicting large information spaces normally apply abstraction, sampling or overplotting to produce the aspired view of the vast space. Some kind of loss of information that might be useful to the observer is accepted. Unlike to these used techniques Information Murals follow different approaches.

Goals of the Information Mural technique:

*Create a representation of an entire (large) information space that fits completely within a display window or screen.

*Mimic what the original visual representation of the information would look like if it could be viewed in its entirety, ie. containing the same visual patterns.

*Minimize the loss of information in the compressed view. [Jerding et al., 1996]

The generated visualization consists of contextual information and supports analytical as well as navigational tasks for a better user-interaction with the informational display. Jerding and Stasko give several different types of information spaces which could be represented using their new technique:

*A text file or document usually does not fit entirely on the screen, because its vertical dimension far exceeds its horizontal dimension. Typically, a text editor displays only a portion of the file being edited.

*Graphs of data often require some compression technique to fit on the screen. Scaling and rounding of data values is often necessary to draw the entire graph. Other alternatives are to display an average of the data values, or only a subset of the data.

*Program visualizations often span many computer screens if laid out completely. This is especially true for those views where one dimension corresponds to time.

*Images might be represented using Information Murals. Although an image usually fits on a screen, it is often desirable to change the size of the image. As an image is shrunk, information in the image is inevitably lost. [Jerding et al., 1996]

=== Used Technique ===

The Information Mural tries to fit a large information space in some area of i x j pixels by scaling the position of each information element to fit into the available space. For each element at a certain pixel different amounts of 'ink' are saved according to the correspondent element-value . While 'drawing' each element, the amount of ink will build up at certain pixel-positions, depending on the grade of overlap of the appendant elements.

A final Information Mural is generated by mapping the amount of ink at each pixel (the information density) to some visual attribute.

Visual attributes might be:

*Grayscale Mural: the shade of each pixel corresponds proportionally to the amount of ink saved for each element.

*Colourscale Mural: uses an equalized intensity variation over the entire color scale.

*'Raindrop' Mural: the amount of ink at each position makes a 'puddle' centered around that pixel. In that case pixels with 'more' ink will appear larger. In addition to that color can be added to the mural to highlight other attributes of the informational elements, while still preserving the density mapping.

=== The Basic Algorithm ===

1) for each i,j set mural_array[i][j] to zero

2) for each element m,n of information
a) compute x = m / M * I, y = n / N * J
b) determine the proportion of this point that lies in each of
the four surrounding mural_array entries (totals to 1.0):
mural_array[floor(x)][floor(y)]
mural_array[floor(x)][ceil(y)]
mural_array[ceil(x)][floor(y)]
mural_array[ceil(x)][ceil(y)]
c) add each of the proportions determined in the previous step to the
existing values of each corresponding mural_array entry
i) update max_mural_array_value to keep track of the
maximum mural_array[][] value

3) for each i,j in the mural_array
a) map the value mural_array[i][j] / max_mural_array_value
to a grayscale or color intensity varying scale, or to pixel size,
depending on the type of mural being created
b) color and draw the pixel at i,j of the mural based on mapping
computed in the previous step

[Jerding et al., 1996]

The rather simple basic algorithm for creating Information Murals listed above will not be explained further because it would blast the scope of this summary.

=== Examples ===

==== Sun Spots ====

The number of sun spots have been recorded since the 1700s by Astronomers. Because of the vast dataset, it is typically plotted by showing the monthly averages. The figure below shows a plot of the average number of sun spots per month recorded from 1850-1993 at the Carnegie Mellon University.

[[image:Occam.jpg|William of Ockham]]

Figure 1: Plot of average number of sun spots recorded per month, 1850-1993. [Jerding et al., 1996]

There is no concern about the size of the dataset using the Information Mural technique. The figure below shows an anti-aliased mural of the number of sun spots recorded daily from 1850-1993, over 52,000 readings. Instead of using grayscale to depict density, a color scale is used to highlight outliers. The scale goes from dark blue (lowest data density) to bright white (highest data density).

[[image:Occam.jpg|William of Ockham]]

Figure 2: Mural of the number of sun spots recorded daily, 1850-1993. [Jerding et al., 1996]

==== Earthlight-Informations as 'Realmural'====

Following image is a panoramic view of the world from a space station. This picture was taken on a perfect night with no obscuring atmospheric conditions. It is a night photo with the lights clearly indicating the populated areas. Light just adds up perfectly. This picture 'highlights' the principles of th Information Mural Technique in a elemental way.

[[image:Occam.jpg|William of Ockham]]

Figure 3: 'Realmural' nach: http://antwrp.gsfc.nasa.gov/apod/ap001127.html [NASA, 2000]

=== References ===

[BYTE-Magazin, 1993] BYTE-Magazin, State of the Art, April 1993, page 120-147

[Jerding et al., 1996] Dean F. Jerding, John T. Stasko, The Information Mural - Overview Article, created at: March 1996, retrieved at: 22.01.2006, http://www.cc.gatech.edu/gvu/softviz/infoviz/information_mural.html

[NASA, 2000] NASA, APOD-Astronomy Picture of the Day, C. Mayhew, R. Simmon, created at: 27.11.2000, retrieved at: 22.01.2006, http://antwrp.gsfc.nasa.gov/apod/ap001127.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4

2006-02-01T11:50:55Z

UE-InfoVis0506 9335259: /* '''Gruppe G4''' */

== '''Gruppe G4''' ==

Die Gruppe G4 besteht bis jetzt aus den folgenden Mitgliedern: 
The following people are part of the G4 group by now:

* [[User:UE-InfoVis0506_0026030|Muster, Anna]]
* [[User:UE-InfoVis0506_9335259|Puchta, Jürgen]]
* [[User:UE-InfoVis0506_0026754|Rainer, Christian]]
* [[User:UE-InfoVis0506_0026629|Sölder, Christoph]]

---- 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 0|Aufgabe 0]] 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 1|Aufgabe 1]] 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 2|Aufgabe 2]] 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3|Aufgabe 3]] 
[[Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Zusatzaufgabe|Zusatzaufgabe]]

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-12T03:01:40Z

UE-InfoVis0506 9335259: /* Specification of used Techniques / applied Principles */

== Topic ==
'''MP3 Archive Visualization - "Interpret Analyser"'''

== Specification of the Application Area and the given Dataset ==
=== Application area Analysis ===

Since we have chosen the MP3 Archive Visualization, our job will be the creation of a clearly arranged visualization for a big music archive consisting of several thousand files.

This can be achieved by using the already existing values of the container format ID3, additional attributes of the iTunes library, attributes of the music files themselves, as well as some system values.

Due the fact that these given sources already provide loads of different Information, we will try to create and present additional Information by combining some prior chosen values in a reasonable way.

Thus we have to keep in mind that ID3 for example theoretically indeed supports a huge amount of input values but in most cases only the most common values like Album, Interpret or Year are specified correctly.

Therefore we will only use some of these entries in our prototype.

=== Dataset Analysis ===

The values we will use in our project consist of nominal, discrete and ordinal data types and are for themselves all one-dimensional.

The table below shows a complete listing:

[[Image:tabelle_datentyp_g4.png]]

The complete data set is multi-dimensional and consists of all attributes listed above.

== Analysis of the Target User Group ==

=== Who should use this kind of visualization technique? ===

This visualization technique is mainly meant for the 'end-users', that is someone who collects lots of MP3s. With 'lots of MP3s' we mean quite a few GBs, just more than 30 GBs. Our visualization should help the user to get an overview of his collection and his listening-habits.
This visualization technique could also be interesting for the band and the music industry, if they want to produce a new album. for example: a band (like "Radiohead") who changed their music-style over the years wants to know which style is preferred more. But therefore they have to compare these datas from many users.

=== What are the characteristics of the target group? ===

People of this group are music enthusiasts. They have thousands of mp3s on their hard-disk and love it to collect them. Most of them have lost track of their collection, on the strength of the abundance of their collection. These people mainly receive their mp3s from the internet instead of buying CDs, because they like to see their whole music-collection at a glance.

=== Are there any known / often used Methods / Visualisation Techniques? ===

No, we don't know any similar visualization technique. ITunes only shows textbased info about how often a song was heard.

== Intended Purpose of our Visualization ==

=== What should be achieved with this visualization? ===
A better information representation of the MP3s should be achieved. Our visualization should help the user to get an overview of his collection and his listening-habits. He will get information about the tracks, in reference to a special artist, which he often listens to and to those which he has never heard before. The representation of the data should be expressive, precise and self-explanatory.

=== Which tasks should be solved? ===
By using this visualization technique, the user will get information about a chosen artist and his discography. For example: in his database the user has got the band "Radiohead", who produced albums over 15 years and in this period they changed their music-style from alternative rock to experimental electronic. The visualization will show him from which producing period he has got more MP3s and which period he likes more, by counting the number of listenings of each song. The result could be that he has got more MP3s from their early years, but likes the experimental electronic tracks more.

=== Questions that should be solved with this visualization technique ===

== Proposal of Design ==

=== Kind of Visualization / Visualization Details ===

When the user opens the "Interpret-Analyser" he will be prompted via a text-message in the main-window to click on an artist/band in the right upper window. The artists/bands are sorted alphabetically and the subject of interest can be found by scrolling the window vertically. If the user chose an artist/band in the upper right window, the main window will visualize him following details on the demanded item:

*The x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data
*The y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library

The respective maximum on the y-axis will give a first overview on how many songs the specific library contains per artist/band per year. Though the users' library might not be complete the visualization allows drawing conclusions according to the artist/band-activities over the last years. In any cases we assume that the user applies the "Interpret-Analyser" to artists/bands whereof he collected the whole discography and not only one song.

For each song (= one data point) one horizontal bar is drawn along the y-axis. That means for example if the library contains 34 songs by the band "Queen" with publication-year "1985", 34 bars are drawn at the x-axis value "1985" along the vertical y-axis.

In addition to that each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently).

Further on the user can interact and influence the characteristic of the visualization by using a slider positioned in the lower right in the graphic below. Via the slider a more objective image can be drawn according to the actual point of interest. This slider with a value-range from "0" to "10" represents the counts how often a song was played. It allows setting a threshold. The default value is "3" and means that songs that were played less than 3 times do not appear coloured, but as grey bars vertically above the coloured ones along the y-axis. If for example someone drags the slider to the position with value "10" and only 1 song out of 27 with a special publishing year was played more than 9 times the "Interpret-Analyser" shows 1 coloured and 26 grey bars at the according year. This could for instance help if someone is on the way to filter out his absolute favourites of an artist/band.

As it is mentioned above the "Interpret-Analyser" represents highly interesting visualizations for End-Users but it might also prevent outstanding features for the Music-Industry respectively bands, who work on a Come-Back. This however would assume to arising the data of a rather big audience, what could for example be achieved via a contest.

=== Visual Mapping ===

'''2D Diagram''':

*'''"X-AXIS":''' the x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data.

*'''"Y-AXIS":''' the y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library. The height of the vertical bar-column represents the "Song-Occurrence" per year.

*'''"Colour":''' each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently). Grey bars represent songs that did not pass the adjusted threshold.

=== Specification of used Techniques / applied Principles ===

*'''Bar X Plot:''' In this plot, one vertical bar is drawn for each data point [StatSoft, 2003]

*'''Histograms, 2D:''' 2D histograms present a graphical representation of the frequency distribution of the selected variable(s) in which the columns are drawn over the class intervals and the heights of the columns are proportional to the class frequencies. [StatSoft, 2003]

*'''Colour-Range, Linking & Brushing:''' A colour-Range representing the levels between "not played for a long time" and "recently played". (s. slide 100 of Info_Vis0.pdf handed out in the course 188.305 VO InfoVis)

*'''Scatterplot, 2D:''' The scatterplot visualizes a relation (correlation) between two variables X and Y (e.g., weight and height). Individual data points are represented in two-dimensional space (see below), where axes represent the variables (X on the horizontal axis and Y on the vertical axis). The two coordinates (X and Y) that determine the location of each point correspond to its specific values on the two variables. [StatSoft, 2003]

*'''Dynamic Queries:''' Adjusting the slider generates dynamic queries. [StatSoft, 2003]

=== Possibilities of User-Interaction ===

*Select item of interest (artist/band)
**Get artist/band-details

*Adjust the slider to influence the threshold
**Get individual Visualizations according to the users' point of interest

=== Mockup / Fake Screenshot ===

[[Image:Interpret-Analyser.png]]

== References ==

[Wikipedia, 2005a] ID3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[Wikipedia, 2005b] MP3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://en.wikipedia.org/wiki/Mp3

[Id3.org, 2004] ID3v2 frames, Id3.org, Last updated: 28. February, 2004, Retrieved at: November 22, 2005, http://www.id3.org/frames.html

[Montclaire, 2000] Data Types, Department of Science and Mathematics at Montclair State University, Last updated: 3. August, 2000, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[StatSoft, 2003] Graphical Analytic Techniques, Last updated: 2003, Retrieved at: November 24, 2005, http://www.statsoft.com/textbook/stgraph.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-12T03:01:24Z

UE-InfoVis0506 9335259: /* Specification of used Techniques / applied Principles */

== Topic ==
'''MP3 Archive Visualization - "Interpret Analyser"'''

== Specification of the Application Area and the given Dataset ==
=== Application area Analysis ===

Since we have chosen the MP3 Archive Visualization, our job will be the creation of a clearly arranged visualization for a big music archive consisting of several thousand files.

This can be achieved by using the already existing values of the container format ID3, additional attributes of the iTunes library, attributes of the music files themselves, as well as some system values.

Due the fact that these given sources already provide loads of different Information, we will try to create and present additional Information by combining some prior chosen values in a reasonable way.

Thus we have to keep in mind that ID3 for example theoretically indeed supports a huge amount of input values but in most cases only the most common values like Album, Interpret or Year are specified correctly.

Therefore we will only use some of these entries in our prototype.

=== Dataset Analysis ===

The values we will use in our project consist of nominal, discrete and ordinal data types and are for themselves all one-dimensional.

The table below shows a complete listing:

[[Image:tabelle_datentyp_g4.png]]

The complete data set is multi-dimensional and consists of all attributes listed above.

== Analysis of the Target User Group ==

=== Who should use this kind of visualization technique? ===

This visualization technique is mainly meant for the 'end-users', that is someone who collects lots of MP3s. With 'lots of MP3s' we mean quite a few GBs, just more than 30 GBs. Our visualization should help the user to get an overview of his collection and his listening-habits.
This visualization technique could also be interesting for the band and the music industry, if they want to produce a new album. for example: a band (like "Radiohead") who changed their music-style over the years wants to know which style is preferred more. But therefore they have to compare these datas from many users.

=== What are the characteristics of the target group? ===

People of this group are music enthusiasts. They have thousands of mp3s on their hard-disk and love it to collect them. Most of them have lost track of their collection, on the strength of the abundance of their collection. These people mainly receive their mp3s from the internet instead of buying CDs, because they like to see their whole music-collection at a glance.

=== Are there any known / often used Methods / Visualisation Techniques? ===

No, we don't know any similar visualization technique. ITunes only shows textbased info about how often a song was heard.

== Intended Purpose of our Visualization ==

=== What should be achieved with this visualization? ===
A better information representation of the MP3s should be achieved. Our visualization should help the user to get an overview of his collection and his listening-habits. He will get information about the tracks, in reference to a special artist, which he often listens to and to those which he has never heard before. The representation of the data should be expressive, precise and self-explanatory.

=== Which tasks should be solved? ===
By using this visualization technique, the user will get information about a chosen artist and his discography. For example: in his database the user has got the band "Radiohead", who produced albums over 15 years and in this period they changed their music-style from alternative rock to experimental electronic. The visualization will show him from which producing period he has got more MP3s and which period he likes more, by counting the number of listenings of each song. The result could be that he has got more MP3s from their early years, but likes the experimental electronic tracks more.

=== Questions that should be solved with this visualization technique ===

== Proposal of Design ==

=== Kind of Visualization / Visualization Details ===

When the user opens the "Interpret-Analyser" he will be prompted via a text-message in the main-window to click on an artist/band in the right upper window. The artists/bands are sorted alphabetically and the subject of interest can be found by scrolling the window vertically. If the user chose an artist/band in the upper right window, the main window will visualize him following details on the demanded item:

*The x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data
*The y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library

The respective maximum on the y-axis will give a first overview on how many songs the specific library contains per artist/band per year. Though the users' library might not be complete the visualization allows drawing conclusions according to the artist/band-activities over the last years. In any cases we assume that the user applies the "Interpret-Analyser" to artists/bands whereof he collected the whole discography and not only one song.

For each song (= one data point) one horizontal bar is drawn along the y-axis. That means for example if the library contains 34 songs by the band "Queen" with publication-year "1985", 34 bars are drawn at the x-axis value "1985" along the vertical y-axis.

In addition to that each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently).

Further on the user can interact and influence the characteristic of the visualization by using a slider positioned in the lower right in the graphic below. Via the slider a more objective image can be drawn according to the actual point of interest. This slider with a value-range from "0" to "10" represents the counts how often a song was played. It allows setting a threshold. The default value is "3" and means that songs that were played less than 3 times do not appear coloured, but as grey bars vertically above the coloured ones along the y-axis. If for example someone drags the slider to the position with value "10" and only 1 song out of 27 with a special publishing year was played more than 9 times the "Interpret-Analyser" shows 1 coloured and 26 grey bars at the according year. This could for instance help if someone is on the way to filter out his absolute favourites of an artist/band.

As it is mentioned above the "Interpret-Analyser" represents highly interesting visualizations for End-Users but it might also prevent outstanding features for the Music-Industry respectively bands, who work on a Come-Back. This however would assume to arising the data of a rather big audience, what could for example be achieved via a contest.

=== Visual Mapping ===

'''2D Diagram''':

*'''"X-AXIS":''' the x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data.

*'''"Y-AXIS":''' the y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library. The height of the vertical bar-column represents the "Song-Occurrence" per year.

*'''"Colour":''' each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently). Grey bars represent songs that did not pass the adjusted threshold.

=== Specification of used Techniques / applied Principles ===

*'''Bar X Plot:''' In this plot, one vertical bar is drawn for each data point [StatSoft, 2003]

*'''Histograms, 2D:''' 2D histograms present a graphical representation of the frequency distribution of the selected variable(s) in which the columns are drawn over the class intervals and the heights of the columns are proportional to the class frequencies. [StatSoft, 2003]

*'''Colour-Range, Linking & Brushing:''' A colour-Range representing the levels between "not played for a long time" and "recently played". (s. slide 100 of Info_Vis0.pdf handed out in the course 188.305 VO InfoVis)

*'''Scatterplot, 2D:''' The scatterplot visualizes a relation (correlation) between two variables X and Y (e.g., weight and height). Individual data points are represented in two-dimensional space (see below), where axes represent the variables (X on the horizontal axis and Y on the vertical axis). The two coordinates (X and Y) that determine the location of each point correspond to its specific values on the two variables. [StatSoft, 2003]

*'''Dynamic Queries:''' Adjusting the slider generates dynamic queries [StatSoft, 2003]

=== Possibilities of User-Interaction ===

*Select item of interest (artist/band)
**Get artist/band-details

*Adjust the slider to influence the threshold
**Get individual Visualizations according to the users' point of interest

=== Mockup / Fake Screenshot ===

[[Image:Interpret-Analyser.png]]

== References ==

[Wikipedia, 2005a] ID3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[Wikipedia, 2005b] MP3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://en.wikipedia.org/wiki/Mp3

[Id3.org, 2004] ID3v2 frames, Id3.org, Last updated: 28. February, 2004, Retrieved at: November 22, 2005, http://www.id3.org/frames.html

[Montclaire, 2000] Data Types, Department of Science and Mathematics at Montclair State University, Last updated: 3. August, 2000, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[StatSoft, 2003] Graphical Analytic Techniques, Last updated: 2003, Retrieved at: November 24, 2005, http://www.statsoft.com/textbook/stgraph.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-12T03:01:08Z

UE-InfoVis0506 9335259: /* Visual Mapping */

== Topic ==
'''MP3 Archive Visualization - "Interpret Analyser"'''

== Specification of the Application Area and the given Dataset ==
=== Application area Analysis ===

Since we have chosen the MP3 Archive Visualization, our job will be the creation of a clearly arranged visualization for a big music archive consisting of several thousand files.

This can be achieved by using the already existing values of the container format ID3, additional attributes of the iTunes library, attributes of the music files themselves, as well as some system values.

Due the fact that these given sources already provide loads of different Information, we will try to create and present additional Information by combining some prior chosen values in a reasonable way.

Thus we have to keep in mind that ID3 for example theoretically indeed supports a huge amount of input values but in most cases only the most common values like Album, Interpret or Year are specified correctly.

Therefore we will only use some of these entries in our prototype.

=== Dataset Analysis ===

The values we will use in our project consist of nominal, discrete and ordinal data types and are for themselves all one-dimensional.

The table below shows a complete listing:

[[Image:tabelle_datentyp_g4.png]]

The complete data set is multi-dimensional and consists of all attributes listed above.

== Analysis of the Target User Group ==

=== Who should use this kind of visualization technique? ===

This visualization technique is mainly meant for the 'end-users', that is someone who collects lots of MP3s. With 'lots of MP3s' we mean quite a few GBs, just more than 30 GBs. Our visualization should help the user to get an overview of his collection and his listening-habits.
This visualization technique could also be interesting for the band and the music industry, if they want to produce a new album. for example: a band (like "Radiohead") who changed their music-style over the years wants to know which style is preferred more. But therefore they have to compare these datas from many users.

=== What are the characteristics of the target group? ===

People of this group are music enthusiasts. They have thousands of mp3s on their hard-disk and love it to collect them. Most of them have lost track of their collection, on the strength of the abundance of their collection. These people mainly receive their mp3s from the internet instead of buying CDs, because they like to see their whole music-collection at a glance.

=== Are there any known / often used Methods / Visualisation Techniques? ===

No, we don't know any similar visualization technique. ITunes only shows textbased info about how often a song was heard.

== Intended Purpose of our Visualization ==

=== What should be achieved with this visualization? ===
A better information representation of the MP3s should be achieved. Our visualization should help the user to get an overview of his collection and his listening-habits. He will get information about the tracks, in reference to a special artist, which he often listens to and to those which he has never heard before. The representation of the data should be expressive, precise and self-explanatory.

=== Which tasks should be solved? ===
By using this visualization technique, the user will get information about a chosen artist and his discography. For example: in his database the user has got the band "Radiohead", who produced albums over 15 years and in this period they changed their music-style from alternative rock to experimental electronic. The visualization will show him from which producing period he has got more MP3s and which period he likes more, by counting the number of listenings of each song. The result could be that he has got more MP3s from their early years, but likes the experimental electronic tracks more.

=== Questions that should be solved with this visualization technique ===

== Proposal of Design ==

=== Kind of Visualization / Visualization Details ===

When the user opens the "Interpret-Analyser" he will be prompted via a text-message in the main-window to click on an artist/band in the right upper window. The artists/bands are sorted alphabetically and the subject of interest can be found by scrolling the window vertically. If the user chose an artist/band in the upper right window, the main window will visualize him following details on the demanded item:

*The x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data
*The y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library

The respective maximum on the y-axis will give a first overview on how many songs the specific library contains per artist/band per year. Though the users' library might not be complete the visualization allows drawing conclusions according to the artist/band-activities over the last years. In any cases we assume that the user applies the "Interpret-Analyser" to artists/bands whereof he collected the whole discography and not only one song.

For each song (= one data point) one horizontal bar is drawn along the y-axis. That means for example if the library contains 34 songs by the band "Queen" with publication-year "1985", 34 bars are drawn at the x-axis value "1985" along the vertical y-axis.

In addition to that each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently).

Further on the user can interact and influence the characteristic of the visualization by using a slider positioned in the lower right in the graphic below. Via the slider a more objective image can be drawn according to the actual point of interest. This slider with a value-range from "0" to "10" represents the counts how often a song was played. It allows setting a threshold. The default value is "3" and means that songs that were played less than 3 times do not appear coloured, but as grey bars vertically above the coloured ones along the y-axis. If for example someone drags the slider to the position with value "10" and only 1 song out of 27 with a special publishing year was played more than 9 times the "Interpret-Analyser" shows 1 coloured and 26 grey bars at the according year. This could for instance help if someone is on the way to filter out his absolute favourites of an artist/band.

As it is mentioned above the "Interpret-Analyser" represents highly interesting visualizations for End-Users but it might also prevent outstanding features for the Music-Industry respectively bands, who work on a Come-Back. This however would assume to arising the data of a rather big audience, what could for example be achieved via a contest.

=== Visual Mapping ===

'''2D Diagram''':

*'''"X-AXIS":''' the x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data.

*'''"Y-AXIS":''' the y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library. The height of the vertical bar-column represents the "Song-Occurrence" per year.

*'''"Colour":''' each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently). Grey bars represent songs that did not pass the adjusted threshold.

=== Specification of used Techniques / applied Principles ===

*'''Bar X Plot:''' In this plot, one vertical bar is drawn for each data point [StatSoft, 2003]

*'''Histograms, 2D:''' 2D histograms present a graphical representation of the frequency distribution of the selected variable(s) in which the columns are drawn over the class intervals and the heights of the columns are proportional to the class frequencies. [StatSoft, 2003]

*'''Colour-Range, Linking & Brushing:''' A colour-Range representing the levels between "not played for a long time" and "recently played" (s. slide 100 of Info_Vis0.pdf handed out in the course 188.305 VO InfoVis)

*'''Scatterplot, 2D:''' The scatterplot visualizes a relation (correlation) between two variables X and Y (e.g., weight and height). Individual data points are represented in two-dimensional space (see below), where axes represent the variables (X on the horizontal axis and Y on the vertical axis). The two coordinates (X and Y) that determine the location of each point correspond to its specific values on the two variables. [StatSoft, 2003]

*'''Dynamic Queries:''' Adjusting the slider generates dynamic queries [StatSoft, 2003]

=== Possibilities of User-Interaction ===

*Select item of interest (artist/band)
**Get artist/band-details

*Adjust the slider to influence the threshold
**Get individual Visualizations according to the users' point of interest

=== Mockup / Fake Screenshot ===

[[Image:Interpret-Analyser.png]]

== References ==

[Wikipedia, 2005a] ID3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[Wikipedia, 2005b] MP3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://en.wikipedia.org/wiki/Mp3

[Id3.org, 2004] ID3v2 frames, Id3.org, Last updated: 28. February, 2004, Retrieved at: November 22, 2005, http://www.id3.org/frames.html

[Montclaire, 2000] Data Types, Department of Science and Mathematics at Montclair State University, Last updated: 3. August, 2000, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[StatSoft, 2003] Graphical Analytic Techniques, Last updated: 2003, Retrieved at: November 24, 2005, http://www.statsoft.com/textbook/stgraph.html

Teaching talk:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-12T02:59:11Z

UE-InfoVis0506 9335259: /* überarbeitetes/neues Konzept */

'''Kommentar zum Konzept:'''

*Die Datentypen für die Attribute sind nicht korrekt - z.B. Title, Interpret, Album sind nominale Datentypen, usw.
*Verwendet bitte bei den Datentypen ein durchgängiges Schema:
** nominal, ordinal, discrete, continuous, binary
** (quantitative kann nämlich discrete oder continuous bedeuten)
*Also entweder haben die Gruppen 9 und 10 von euch diese Tabelle abgekupfert oder umgekehrt? :-/
*Der Datensatz ist als solcher ''multidimensional''
*Subgrouping ist soweit ich weiss bei ID3 tags nicht möglich (ein Genre als Unterpunkt eines anderen) - Woher kommt die Zusatzinformation, die so nicht in den Daten vorkommt?
*IMHO ist die Anzahl verschiedener Genres in einer MP3 Sammlung nicht wirklich "enorm" (ich habe z.B. über 30GB MP3s, aber nur knapp über 100 verschiedene Genres...)
*Für die relativ wenig Information, die in eurer graphischen Repräsentation visualisiert wird, wird sehr viel Platz verwendet. (Ein normaler TreeView mit dazugehängten Balken würde dieselbe Information transportieren, allerdings wesentlich weniger Platz benötigen und damit eine bessere Übersicht ermöglichen.)
*Ihr visualisiert eigentlich nur 2 verschiedene Datenattribute: Genre und Anzahl von Songs innerhalb des Genres. Dafür braucht man eigentlich nicht wirklich besondere, interaktive InfoVis Methoden (die sollten dazu dienen, möglichst viele Parameter übersichtlich und sinnvoll miteinenader zu kombinieren um neues Wissen zu generieren)
*Ihr geht auf die selben Variablen ein wie Gruppe 4 -- der Kopierverdacht erhärtet sich...
*Kurz und gut: überdenkt euer Konzept noch einmal und versucht mehr Information zu integrieren bzw. andere Darstellungsvarianten dafür...

-- [[User:Iwolf|Wolfgang Aigner]] 14:31, 25 November 2005 (CET)

== Kopierverdacht + Korrektur ==

*Bezüglich Ihres Verdachts, dass die Datensatzanalyse abgeschrieben sein könnte, kann ich nur sagen, dass ich diesen Teil selbst erarbeitet habe.
:'''''--> also das halte ich für eine sehr gewagte Theorie und es fällt mir ziemlich schwer das zu glauben... ihr habt die fragliche Tabelle wesentlich später online gestellt und haargenau dieselben Fehler wie die Gruppen 9 und 10 (sogar die Rechtschreibfehler!) - also Zufall hin oder her - 3x die (fast) gleiche Tabelle und identische inhaltliche sowie Rechtschreibfehler...'''''

** --> Rechtschreibfehler konnte ich bei meiner Tabelle keine finden. Zudem haben wir in unserer Angabe als einzige den Datentyp "Quantitative" angegeben. Bei einem Vergleich mit den anderen Gruppen konnte ich ebenfalls Ähnlichkeiten feststellen, die aber wohl eher damit zu begründen sind, dass wir alle das Thema MP3 gewählt haben.

::'''''--> Stimmt, da habt ihr recht mit den Rechtschreibfehlern und Datentypen - sorry. Nichts desto trotz sehen sich die Datensatzanalysen sehr ähnlich. Aber gut, das soll jetzt nicht das große Problem sein. Viel wichtiger ist die inhaltliche Verbesserung eures Konzeptes - da solltet ihr noch einige Überlegungen reinstecken!''''' -- [[User:Iwolf|Wolfgang Aigner]] 08:56, 28 November 2005 (CET)

*Der Gründe für die Fehler in meiner Ausarbeitung liegen in folgenden Punkten:
**Leider hatte ich Probleme im Internet (für mich) schlüssige Unterlagen zu diesem Thema finden.
** Zudem waren die Definitionen, die ich zu den diversen Datentypen überhaupt finden konnte oft widersprüchlich.
** So war es für mich schwierig die Daten eindeutig zuzuordnen.
:'''''--> Datentypen sollten eigentlich Grundwissen z.B. aus Statistik sein. Hier nochmal eine kompakte Zusammenfassung: [[Data Type]].'''''

*Bei den Referenzen unserer Ausarbeitung können sie die Internetseite finden, die ich schlussendlich hauptsächlich als Ausgangsbasis für meine Ausarbeitung verwendet habe. Diese Seite stellt eine der detailliertesten Angaben dar, die ich finden konnte.

*Das sich die Fehler in der Analyse zum Teil überdecken, kann ich mir persönlich nur damit erklären, dass die anderen Gruppen mit ähnlichen Schwierigkeiten zu kämpfen hatten.

*Natürlich werden wir unsere Abgabe nochmals überarbeiten bzw. überdenken. Daher möchten wir gerne wissen, bis wann Sie uns hierfür Zeit geben.
:'''''--> Ich erwarte mir eine komplette Überarbeitung eures Konzepts (inkl. Datensatzanalyse) und vor allem einen verbesserten Designentwurf. Eine Deadline für die Überarbeitung gibt es nicht - es ist in eurem Interesse, das so bald wie möglich zu tun, schliesslich ist die Abgabe des Prototypen in weniger als 4 Wochen.'''''

-- [[User:Iwolf|Wolfgang Aigner]] 12:52, 26 November 2005 (CET)

== überarbeitetes/neues Konzept ==

*ein überarbeitetes bzw. generell neu ausgelegtes Konzept ist jetzt online.

*Wir glauben, mit unserem neuen Konzept die Aufgabenstellung nun wesentlich besser "getroffen" zu haben und bitten Sie wieder um ein kurzes Feedback, damit wir mit unserer Implementierung beginnen können.

*Vielen Dank, G4

Teaching talk:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-12T02:58:25Z

UE-InfoVis0506 9335259: /* überarbeitetes/neues Konzept */

'''Kommentar zum Konzept:'''

*Die Datentypen für die Attribute sind nicht korrekt - z.B. Title, Interpret, Album sind nominale Datentypen, usw.
*Verwendet bitte bei den Datentypen ein durchgängiges Schema:
** nominal, ordinal, discrete, continuous, binary
** (quantitative kann nämlich discrete oder continuous bedeuten)
*Also entweder haben die Gruppen 9 und 10 von euch diese Tabelle abgekupfert oder umgekehrt? :-/
*Der Datensatz ist als solcher ''multidimensional''
*Subgrouping ist soweit ich weiss bei ID3 tags nicht möglich (ein Genre als Unterpunkt eines anderen) - Woher kommt die Zusatzinformation, die so nicht in den Daten vorkommt?
*IMHO ist die Anzahl verschiedener Genres in einer MP3 Sammlung nicht wirklich "enorm" (ich habe z.B. über 30GB MP3s, aber nur knapp über 100 verschiedene Genres...)
*Für die relativ wenig Information, die in eurer graphischen Repräsentation visualisiert wird, wird sehr viel Platz verwendet. (Ein normaler TreeView mit dazugehängten Balken würde dieselbe Information transportieren, allerdings wesentlich weniger Platz benötigen und damit eine bessere Übersicht ermöglichen.)
*Ihr visualisiert eigentlich nur 2 verschiedene Datenattribute: Genre und Anzahl von Songs innerhalb des Genres. Dafür braucht man eigentlich nicht wirklich besondere, interaktive InfoVis Methoden (die sollten dazu dienen, möglichst viele Parameter übersichtlich und sinnvoll miteinenader zu kombinieren um neues Wissen zu generieren)
*Ihr geht auf die selben Variablen ein wie Gruppe 4 -- der Kopierverdacht erhärtet sich...
*Kurz und gut: überdenkt euer Konzept noch einmal und versucht mehr Information zu integrieren bzw. andere Darstellungsvarianten dafür...

-- [[User:Iwolf|Wolfgang Aigner]] 14:31, 25 November 2005 (CET)

== Kopierverdacht + Korrektur ==

*Bezüglich Ihres Verdachts, dass die Datensatzanalyse abgeschrieben sein könnte, kann ich nur sagen, dass ich diesen Teil selbst erarbeitet habe.
:'''''--> also das halte ich für eine sehr gewagte Theorie und es fällt mir ziemlich schwer das zu glauben... ihr habt die fragliche Tabelle wesentlich später online gestellt und haargenau dieselben Fehler wie die Gruppen 9 und 10 (sogar die Rechtschreibfehler!) - also Zufall hin oder her - 3x die (fast) gleiche Tabelle und identische inhaltliche sowie Rechtschreibfehler...'''''

** --> Rechtschreibfehler konnte ich bei meiner Tabelle keine finden. Zudem haben wir in unserer Angabe als einzige den Datentyp "Quantitative" angegeben. Bei einem Vergleich mit den anderen Gruppen konnte ich ebenfalls Ähnlichkeiten feststellen, die aber wohl eher damit zu begründen sind, dass wir alle das Thema MP3 gewählt haben.

::'''''--> Stimmt, da habt ihr recht mit den Rechtschreibfehlern und Datentypen - sorry. Nichts desto trotz sehen sich die Datensatzanalysen sehr ähnlich. Aber gut, das soll jetzt nicht das große Problem sein. Viel wichtiger ist die inhaltliche Verbesserung eures Konzeptes - da solltet ihr noch einige Überlegungen reinstecken!''''' -- [[User:Iwolf|Wolfgang Aigner]] 08:56, 28 November 2005 (CET)

*Der Gründe für die Fehler in meiner Ausarbeitung liegen in folgenden Punkten:
**Leider hatte ich Probleme im Internet (für mich) schlüssige Unterlagen zu diesem Thema finden.
** Zudem waren die Definitionen, die ich zu den diversen Datentypen überhaupt finden konnte oft widersprüchlich.
** So war es für mich schwierig die Daten eindeutig zuzuordnen.
:'''''--> Datentypen sollten eigentlich Grundwissen z.B. aus Statistik sein. Hier nochmal eine kompakte Zusammenfassung: [[Data Type]].'''''

*Bei den Referenzen unserer Ausarbeitung können sie die Internetseite finden, die ich schlussendlich hauptsächlich als Ausgangsbasis für meine Ausarbeitung verwendet habe. Diese Seite stellt eine der detailliertesten Angaben dar, die ich finden konnte.

*Das sich die Fehler in der Analyse zum Teil überdecken, kann ich mir persönlich nur damit erklären, dass die anderen Gruppen mit ähnlichen Schwierigkeiten zu kämpfen hatten.

*Natürlich werden wir unsere Abgabe nochmals überarbeiten bzw. überdenken. Daher möchten wir gerne wissen, bis wann Sie uns hierfür Zeit geben.
:'''''--> Ich erwarte mir eine komplette Überarbeitung eures Konzepts (inkl. Datensatzanalyse) und vor allem einen verbesserten Designentwurf. Eine Deadline für die Überarbeitung gibt es nicht - es ist in eurem Interesse, das so bald wie möglich zu tun, schliesslich ist die Abgabe des Prototypen in weniger als 4 Wochen.'''''

-- [[User:Iwolf|Wolfgang Aigner]] 12:52, 26 November 2005 (CET)

== überarbeitetes/neues Konzept ==

*ein überarbeitetes bzw. generell neu ausgelegtes Konzept ist jetzt online.

*Wir glauben, mit unserem neuen Konzept die Aufgabenstellung nun wesentlich besser "getroffen" zu haben und bitten sie wieder um ein kurzes Feedback, damit wir mit unserer Implementierung beginnen können.

*Vielen Dank, G4

Teaching talk:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-12T02:58:12Z

UE-InfoVis0506 9335259: /* überarbeitetes/neues Konzept */

'''Kommentar zum Konzept:'''

*Die Datentypen für die Attribute sind nicht korrekt - z.B. Title, Interpret, Album sind nominale Datentypen, usw.
*Verwendet bitte bei den Datentypen ein durchgängiges Schema:
** nominal, ordinal, discrete, continuous, binary
** (quantitative kann nämlich discrete oder continuous bedeuten)
*Also entweder haben die Gruppen 9 und 10 von euch diese Tabelle abgekupfert oder umgekehrt? :-/
*Der Datensatz ist als solcher ''multidimensional''
*Subgrouping ist soweit ich weiss bei ID3 tags nicht möglich (ein Genre als Unterpunkt eines anderen) - Woher kommt die Zusatzinformation, die so nicht in den Daten vorkommt?
*IMHO ist die Anzahl verschiedener Genres in einer MP3 Sammlung nicht wirklich "enorm" (ich habe z.B. über 30GB MP3s, aber nur knapp über 100 verschiedene Genres...)
*Für die relativ wenig Information, die in eurer graphischen Repräsentation visualisiert wird, wird sehr viel Platz verwendet. (Ein normaler TreeView mit dazugehängten Balken würde dieselbe Information transportieren, allerdings wesentlich weniger Platz benötigen und damit eine bessere Übersicht ermöglichen.)
*Ihr visualisiert eigentlich nur 2 verschiedene Datenattribute: Genre und Anzahl von Songs innerhalb des Genres. Dafür braucht man eigentlich nicht wirklich besondere, interaktive InfoVis Methoden (die sollten dazu dienen, möglichst viele Parameter übersichtlich und sinnvoll miteinenader zu kombinieren um neues Wissen zu generieren)
*Ihr geht auf die selben Variablen ein wie Gruppe 4 -- der Kopierverdacht erhärtet sich...
*Kurz und gut: überdenkt euer Konzept noch einmal und versucht mehr Information zu integrieren bzw. andere Darstellungsvarianten dafür...

-- [[User:Iwolf|Wolfgang Aigner]] 14:31, 25 November 2005 (CET)

== Kopierverdacht + Korrektur ==

*Bezüglich Ihres Verdachts, dass die Datensatzanalyse abgeschrieben sein könnte, kann ich nur sagen, dass ich diesen Teil selbst erarbeitet habe.
:'''''--> also das halte ich für eine sehr gewagte Theorie und es fällt mir ziemlich schwer das zu glauben... ihr habt die fragliche Tabelle wesentlich später online gestellt und haargenau dieselben Fehler wie die Gruppen 9 und 10 (sogar die Rechtschreibfehler!) - also Zufall hin oder her - 3x die (fast) gleiche Tabelle und identische inhaltliche sowie Rechtschreibfehler...'''''

** --> Rechtschreibfehler konnte ich bei meiner Tabelle keine finden. Zudem haben wir in unserer Angabe als einzige den Datentyp "Quantitative" angegeben. Bei einem Vergleich mit den anderen Gruppen konnte ich ebenfalls Ähnlichkeiten feststellen, die aber wohl eher damit zu begründen sind, dass wir alle das Thema MP3 gewählt haben.

::'''''--> Stimmt, da habt ihr recht mit den Rechtschreibfehlern und Datentypen - sorry. Nichts desto trotz sehen sich die Datensatzanalysen sehr ähnlich. Aber gut, das soll jetzt nicht das große Problem sein. Viel wichtiger ist die inhaltliche Verbesserung eures Konzeptes - da solltet ihr noch einige Überlegungen reinstecken!''''' -- [[User:Iwolf|Wolfgang Aigner]] 08:56, 28 November 2005 (CET)

*Der Gründe für die Fehler in meiner Ausarbeitung liegen in folgenden Punkten:
**Leider hatte ich Probleme im Internet (für mich) schlüssige Unterlagen zu diesem Thema finden.
** Zudem waren die Definitionen, die ich zu den diversen Datentypen überhaupt finden konnte oft widersprüchlich.
** So war es für mich schwierig die Daten eindeutig zuzuordnen.
:'''''--> Datentypen sollten eigentlich Grundwissen z.B. aus Statistik sein. Hier nochmal eine kompakte Zusammenfassung: [[Data Type]].'''''

*Bei den Referenzen unserer Ausarbeitung können sie die Internetseite finden, die ich schlussendlich hauptsächlich als Ausgangsbasis für meine Ausarbeitung verwendet habe. Diese Seite stellt eine der detailliertesten Angaben dar, die ich finden konnte.

*Das sich die Fehler in der Analyse zum Teil überdecken, kann ich mir persönlich nur damit erklären, dass die anderen Gruppen mit ähnlichen Schwierigkeiten zu kämpfen hatten.

*Natürlich werden wir unsere Abgabe nochmals überarbeiten bzw. überdenken. Daher möchten wir gerne wissen, bis wann Sie uns hierfür Zeit geben.
:'''''--> Ich erwarte mir eine komplette Überarbeitung eures Konzepts (inkl. Datensatzanalyse) und vor allem einen verbesserten Designentwurf. Eine Deadline für die Überarbeitung gibt es nicht - es ist in eurem Interesse, das so bald wie möglich zu tun, schliesslich ist die Abgabe des Prototypen in weniger als 4 Wochen.'''''

-- [[User:Iwolf|Wolfgang Aigner]] 12:52, 26 November 2005 (CET)

== überarbeitetes/neues Konzept ==

*ein überarbeitetes bzw. generell neu ausgelegtes Konzept ist jetzt online.

*Wir glauben, mit unserm neuen Konzept die Aufgabenstellung nun wesentlich besser "getroffen" zu haben und bitten sie wieder um ein kurzes Feedback, damit wir mit unserer Implementierung beginnen können.

*Vielen Dank, G4

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-11T21:46:16Z

UE-InfoVis0506 9335259: /* Topic */

== Topic ==
'''MP3 Archive Visualization - "Interpret Analyser"'''

== Specification of the Application Area and the given Dataset ==
=== Application area Analysis ===

Since we have chosen the MP3 Archive Visualization, our job will be the creation of a clearly arranged visualization for a big music archive consisting of several thousand files.

This can be achieved by using the already existing values of the container format ID3, additional attributes of the iTunes library, attributes of the music files themselves, as well as some system values.

Due the fact that these given sources already provide loads of different Information, we will try to create and present additional Information by combining some prior chosen values in a reasonable way.

Thus we have to keep in mind that ID3 for example theoretically indeed supports a huge amount of input values but in most cases only the most common values like Album, Interpret or Year are specified correctly.

Therefore we will only use some of these entries in our prototype.

=== Dataset Analysis ===

The values we will use in our project consist of nominal, discrete and ordinal data types and are for themselves all one-dimensional.

The table below shows a complete listing:

[[Image:tabelle_datentyp_g4.png]]

The complete data set is multi-dimensional and consists of all attributes listed above.

== Analysis of the Target User Group ==

=== Who should use this kind of visualization technique? ===

This visualization technique is mainly meant for the 'end-users', that is someone who collects lots of MP3s. With 'lots of MP3s' we mean quite a few GBs, just more than 30 GBs. Our visualization should help the user to get an overview of his collection and his listening-habits.
This visualization technique could also be interesting for the band and the music industry, if they want to produce a new album. for example: a band (like "Radiohead") who changed their music-style over the years wants to know which style is preferred more. But therefore they have to compare these datas from many users.

=== What are the characteristics of the target group? ===

People of this group are music enthusiasts. They have thousands of mp3s on their hard-disk and love it to collect them. Most of them have lost track of their collection, on the strength of the abundance of their collection. These people mainly receive their mp3s from the internet instead of buying CDs, because they like to see their whole music-collection at a glance.

=== Are there any known / often used Methods / Visualisation Techniques? ===

No, we don't know any similar visualization technique. ITunes only shows textbased info about how often a song was heard.

== Intended Purpose of our Visualization ==

=== What should be achieved with this visualization? ===
A better information representation of the MP3s should be achieved. Our visualization should help the user to get an overview of his collection and his listening-habits. He will get information about the tracks, in reference to a special artist, which he often listens to and to those which he has never heard before. The representation of the data should be expressive, precise and self-explanatory.

=== Which tasks should be solved? ===
By using this visualization technique, the user will get information about a chosen artist and his discography. For example: in his database the user has got the band "Radiohead", who produced albums over 15 years and in this period they changed their music-style from alternative rock to experimental electronic. The visualization will show him from which producing period he has got more MP3s and which period he likes more, by counting the number of listenings of each song. The result could be that he has got more MP3s from their early years, but likes the experimental electronic tracks more.

=== Questions that should be solved with this visualization technique ===

== Proposal of Design ==

=== Kind of Visualization / Visualization Details ===

When the user opens the "Interpret-Analyser" he will be prompted via a text-message in the main-window to click on an artist/band in the right upper window. The artists/bands are sorted alphabetically and the subject of interest can be found by scrolling the window vertically. If the user chose an artist/band in the upper right window, the main window will visualize him following details on the demanded item:

*The x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data
*The y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library

The respective maximum on the y-axis will give a first overview on how many songs the specific library contains per artist/band per year. Though the users' library might not be complete the visualization allows drawing conclusions according to the artist/band-activities over the last years. In any cases we assume that the user applies the "Interpret-Analyser" to artists/bands whereof he collected the whole discography and not only one song.

For each song (= one data point) one horizontal bar is drawn along the y-axis. That means for example if the library contains 34 songs by the band "Queen" with publication-year "1985", 34 bars are drawn at the x-axis value "1985" along the vertical y-axis.

In addition to that each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently).

Further on the user can interact and influence the characteristic of the visualization by using a slider positioned in the lower right in the graphic below. Via the slider a more objective image can be drawn according to the actual point of interest. This slider with a value-range from "0" to "10" represents the counts how often a song was played. It allows setting a threshold. The default value is "3" and means that songs that were played less than 3 times do not appear coloured, but as grey bars vertically above the coloured ones along the y-axis. If for example someone drags the slider to the position with value "10" and only 1 song out of 27 with a special publishing year was played more than 9 times the "Interpret-Analyser" shows 1 coloured and 26 grey bars at the according year. This could for instance help if someone is on the way to filter out his absolute favourites of an artist/band.

As it is mentioned above the "Interpret-Analyser" represents highly interesting visualizations for End-Users but it might also prevent outstanding features for the Music-Industry respectively bands, who work on a Come-Back. This however would assume to arising the data of a rather big audience, what could for example be achieved via a contest.

=== Visual Mapping ===

'''2D Diagram''':

*'''"X-AXIS":''' the x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data

*'''"Y-AXIS":''' the y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library. The height of the vertical bar-column represents the "Song-Occurrence" per year

*'''"Colour":''' each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently). Grey bars represent songs that did not pass the adjusted threshold

=== Specification of used Techniques / applied Principles ===

*'''Bar X Plot:''' In this plot, one vertical bar is drawn for each data point [StatSoft, 2003]

*'''Histograms, 2D:''' 2D histograms present a graphical representation of the frequency distribution of the selected variable(s) in which the columns are drawn over the class intervals and the heights of the columns are proportional to the class frequencies. [StatSoft, 2003]

*'''Colour-Range, Linking & Brushing:''' A colour-Range representing the levels between "not played for a long time" and "recently played" (s. slide 100 of Info_Vis0.pdf handed out in the course 188.305 VO InfoVis)

*'''Scatterplot, 2D:''' The scatterplot visualizes a relation (correlation) between two variables X and Y (e.g., weight and height). Individual data points are represented in two-dimensional space (see below), where axes represent the variables (X on the horizontal axis and Y on the vertical axis). The two coordinates (X and Y) that determine the location of each point correspond to its specific values on the two variables. [StatSoft, 2003]

*'''Dynamic Queries:''' Adjusting the slider generates dynamic queries [StatSoft, 2003]

=== Possibilities of User-Interaction ===

*Select item of interest (artist/band)
**Get artist/band-details

*Adjust the slider to influence the threshold
**Get individual Visualizations according to the users' point of interest

=== Mockup / Fake Screenshot ===

[[Image:Interpret-Analyser.png]]

== References ==

[Wikipedia, 2005a] ID3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[Wikipedia, 2005b] MP3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://en.wikipedia.org/wiki/Mp3

[Id3.org, 2004] ID3v2 frames, Id3.org, Last updated: 28. February, 2004, Retrieved at: November 22, 2005, http://www.id3.org/frames.html

[Montclaire, 2000] Data Types, Department of Science and Mathematics at Montclair State University, Last updated: 3. August, 2000, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[StatSoft, 2003] Graphical Analytic Techniques, Last updated: 2003, Retrieved at: November 24, 2005, http://www.statsoft.com/textbook/stgraph.html

Teaching talk:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-11T21:44:11Z

UE-InfoVis0506 9335259: /* überarbeitetes/neues Konzept */

'''Kommentar zum Konzept:'''

*Die Datentypen für die Attribute sind nicht korrekt - z.B. Title, Interpret, Album sind nominale Datentypen, usw.
*Verwendet bitte bei den Datentypen ein durchgängiges Schema:
** nominal, ordinal, discrete, continuous, binary
** (quantitative kann nämlich discrete oder continuous bedeuten)
*Also entweder haben die Gruppen 9 und 10 von euch diese Tabelle abgekupfert oder umgekehrt? :-/
*Der Datensatz ist als solcher ''multidimensional''
*Subgrouping ist soweit ich weiss bei ID3 tags nicht möglich (ein Genre als Unterpunkt eines anderen) - Woher kommt die Zusatzinformation, die so nicht in den Daten vorkommt?
*IMHO ist die Anzahl verschiedener Genres in einer MP3 Sammlung nicht wirklich "enorm" (ich habe z.B. über 30GB MP3s, aber nur knapp über 100 verschiedene Genres...)
*Für die relativ wenig Information, die in eurer graphischen Repräsentation visualisiert wird, wird sehr viel Platz verwendet. (Ein normaler TreeView mit dazugehängten Balken würde dieselbe Information transportieren, allerdings wesentlich weniger Platz benötigen und damit eine bessere Übersicht ermöglichen.)
*Ihr visualisiert eigentlich nur 2 verschiedene Datenattribute: Genre und Anzahl von Songs innerhalb des Genres. Dafür braucht man eigentlich nicht wirklich besondere, interaktive InfoVis Methoden (die sollten dazu dienen, möglichst viele Parameter übersichtlich und sinnvoll miteinenader zu kombinieren um neues Wissen zu generieren)
*Ihr geht auf die selben Variablen ein wie Gruppe 4 -- der Kopierverdacht erhärtet sich...
*Kurz und gut: überdenkt euer Konzept noch einmal und versucht mehr Information zu integrieren bzw. andere Darstellungsvarianten dafür...

-- [[User:Iwolf|Wolfgang Aigner]] 14:31, 25 November 2005 (CET)

== Kopierverdacht + Korrektur ==

*Bezüglich Ihres Verdachts, dass die Datensatzanalyse abgeschrieben sein könnte, kann ich nur sagen, dass ich diesen Teil selbst erarbeitet habe.
:'''''--> also das halte ich für eine sehr gewagte Theorie und es fällt mir ziemlich schwer das zu glauben... ihr habt die fragliche Tabelle wesentlich später online gestellt und haargenau dieselben Fehler wie die Gruppen 9 und 10 (sogar die Rechtschreibfehler!) - also Zufall hin oder her - 3x die (fast) gleiche Tabelle und identische inhaltliche sowie Rechtschreibfehler...'''''

** --> Rechtschreibfehler konnte ich bei meiner Tabelle keine finden. Zudem haben wir in unserer Angabe als einzige den Datentyp "Quantitative" angegeben. Bei einem Vergleich mit den anderen Gruppen konnte ich ebenfalls Ähnlichkeiten feststellen, die aber wohl eher damit zu begründen sind, dass wir alle das Thema MP3 gewählt haben.

::'''''--> Stimmt, da habt ihr recht mit den Rechtschreibfehlern und Datentypen - sorry. Nichts desto trotz sehen sich die Datensatzanalysen sehr ähnlich. Aber gut, das soll jetzt nicht das große Problem sein. Viel wichtiger ist die inhaltliche Verbesserung eures Konzeptes - da solltet ihr noch einige Überlegungen reinstecken!''''' -- [[User:Iwolf|Wolfgang Aigner]] 08:56, 28 November 2005 (CET)

*Der Gründe für die Fehler in meiner Ausarbeitung liegen in folgenden Punkten:
**Leider hatte ich Probleme im Internet (für mich) schlüssige Unterlagen zu diesem Thema finden.
** Zudem waren die Definitionen, die ich zu den diversen Datentypen überhaupt finden konnte oft widersprüchlich.
** So war es für mich schwierig die Daten eindeutig zuzuordnen.
:'''''--> Datentypen sollten eigentlich Grundwissen z.B. aus Statistik sein. Hier nochmal eine kompakte Zusammenfassung: [[Data Type]].'''''

*Bei den Referenzen unserer Ausarbeitung können sie die Internetseite finden, die ich schlussendlich hauptsächlich als Ausgangsbasis für meine Ausarbeitung verwendet habe. Diese Seite stellt eine der detailliertesten Angaben dar, die ich finden konnte.

*Das sich die Fehler in der Analyse zum Teil überdecken, kann ich mir persönlich nur damit erklären, dass die anderen Gruppen mit ähnlichen Schwierigkeiten zu kämpfen hatten.

*Natürlich werden wir unsere Abgabe nochmals überarbeiten bzw. überdenken. Daher möchten wir gerne wissen, bis wann Sie uns hierfür Zeit geben.
:'''''--> Ich erwarte mir eine komplette Überarbeitung eures Konzepts (inkl. Datensatzanalyse) und vor allem einen verbesserten Designentwurf. Eine Deadline für die Überarbeitung gibt es nicht - es ist in eurem Interesse, das so bald wie möglich zu tun, schliesslich ist die Abgabe des Prototypen in weniger als 4 Wochen.'''''

-- [[User:Iwolf|Wolfgang Aigner]] 12:52, 26 November 2005 (CET)

== überarbeitetes/neues Konzept ==

*ein überarbeitetes bzw. generell neu ausgelegtes Konzept ist jetzt online.

*Wir glauben, mit unserm neuen Konzept die Aufgabenstellung nun wesentlich besser "getroffen" zu haben und bitten sie wieder um ein kurzes Feedback, damit wir mit unserer Implementierung beginnen können.

*Vielen Dank, G4

Teaching talk:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-11T21:43:47Z

UE-InfoVis0506 9335259: /* Überarbeitetes/neues Konzept */

'''Kommentar zum Konzept:'''

*Die Datentypen für die Attribute sind nicht korrekt - z.B. Title, Interpret, Album sind nominale Datentypen, usw.
*Verwendet bitte bei den Datentypen ein durchgängiges Schema:
** nominal, ordinal, discrete, continuous, binary
** (quantitative kann nämlich discrete oder continuous bedeuten)
*Also entweder haben die Gruppen 9 und 10 von euch diese Tabelle abgekupfert oder umgekehrt? :-/
*Der Datensatz ist als solcher ''multidimensional''
*Subgrouping ist soweit ich weiss bei ID3 tags nicht möglich (ein Genre als Unterpunkt eines anderen) - Woher kommt die Zusatzinformation, die so nicht in den Daten vorkommt?
*IMHO ist die Anzahl verschiedener Genres in einer MP3 Sammlung nicht wirklich "enorm" (ich habe z.B. über 30GB MP3s, aber nur knapp über 100 verschiedene Genres...)
*Für die relativ wenig Information, die in eurer graphischen Repräsentation visualisiert wird, wird sehr viel Platz verwendet. (Ein normaler TreeView mit dazugehängten Balken würde dieselbe Information transportieren, allerdings wesentlich weniger Platz benötigen und damit eine bessere Übersicht ermöglichen.)
*Ihr visualisiert eigentlich nur 2 verschiedene Datenattribute: Genre und Anzahl von Songs innerhalb des Genres. Dafür braucht man eigentlich nicht wirklich besondere, interaktive InfoVis Methoden (die sollten dazu dienen, möglichst viele Parameter übersichtlich und sinnvoll miteinenader zu kombinieren um neues Wissen zu generieren)
*Ihr geht auf die selben Variablen ein wie Gruppe 4 -- der Kopierverdacht erhärtet sich...
*Kurz und gut: überdenkt euer Konzept noch einmal und versucht mehr Information zu integrieren bzw. andere Darstellungsvarianten dafür...

-- [[User:Iwolf|Wolfgang Aigner]] 14:31, 25 November 2005 (CET)

== Kopierverdacht + Korrektur ==

*Bezüglich Ihres Verdachts, dass die Datensatzanalyse abgeschrieben sein könnte, kann ich nur sagen, dass ich diesen Teil selbst erarbeitet habe.
:'''''--> also das halte ich für eine sehr gewagte Theorie und es fällt mir ziemlich schwer das zu glauben... ihr habt die fragliche Tabelle wesentlich später online gestellt und haargenau dieselben Fehler wie die Gruppen 9 und 10 (sogar die Rechtschreibfehler!) - also Zufall hin oder her - 3x die (fast) gleiche Tabelle und identische inhaltliche sowie Rechtschreibfehler...'''''

** --> Rechtschreibfehler konnte ich bei meiner Tabelle keine finden. Zudem haben wir in unserer Angabe als einzige den Datentyp "Quantitative" angegeben. Bei einem Vergleich mit den anderen Gruppen konnte ich ebenfalls Ähnlichkeiten feststellen, die aber wohl eher damit zu begründen sind, dass wir alle das Thema MP3 gewählt haben.

::'''''--> Stimmt, da habt ihr recht mit den Rechtschreibfehlern und Datentypen - sorry. Nichts desto trotz sehen sich die Datensatzanalysen sehr ähnlich. Aber gut, das soll jetzt nicht das große Problem sein. Viel wichtiger ist die inhaltliche Verbesserung eures Konzeptes - da solltet ihr noch einige Überlegungen reinstecken!''''' -- [[User:Iwolf|Wolfgang Aigner]] 08:56, 28 November 2005 (CET)

*Der Gründe für die Fehler in meiner Ausarbeitung liegen in folgenden Punkten:
**Leider hatte ich Probleme im Internet (für mich) schlüssige Unterlagen zu diesem Thema finden.
** Zudem waren die Definitionen, die ich zu den diversen Datentypen überhaupt finden konnte oft widersprüchlich.
** So war es für mich schwierig die Daten eindeutig zuzuordnen.
:'''''--> Datentypen sollten eigentlich Grundwissen z.B. aus Statistik sein. Hier nochmal eine kompakte Zusammenfassung: [[Data Type]].'''''

*Bei den Referenzen unserer Ausarbeitung können sie die Internetseite finden, die ich schlussendlich hauptsächlich als Ausgangsbasis für meine Ausarbeitung verwendet habe. Diese Seite stellt eine der detailliertesten Angaben dar, die ich finden konnte.

*Das sich die Fehler in der Analyse zum Teil überdecken, kann ich mir persönlich nur damit erklären, dass die anderen Gruppen mit ähnlichen Schwierigkeiten zu kämpfen hatten.

*Natürlich werden wir unsere Abgabe nochmals überarbeiten bzw. überdenken. Daher möchten wir gerne wissen, bis wann Sie uns hierfür Zeit geben.
:'''''--> Ich erwarte mir eine komplette Überarbeitung eures Konzepts (inkl. Datensatzanalyse) und vor allem einen verbesserten Designentwurf. Eine Deadline für die Überarbeitung gibt es nicht - es ist in eurem Interesse, das so bald wie möglich zu tun, schliesslich ist die Abgabe des Prototypen in weniger als 4 Wochen.'''''

-- [[User:Iwolf|Wolfgang Aigner]] 12:52, 26 November 2005 (CET)

== überarbeitetes/neues Konzept ==

*ein überarbeitetes bzw. generell neu ausgelegtes Konzept ist jetzt online.

*Wir glauben, mit unserm neuen Konzept die Aufgabenstellung nun wesentlich besser "getroffen" zu haben und bitten sie wieder um ein kurzes Feedback, damit wir mit unserer Implementierung beginnen können.

**Vielen Dank, G4

Teaching talk:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-11T21:43:26Z

UE-InfoVis0506 9335259: /* Überarbeitetes/neues Konzept */

'''Kommentar zum Konzept:'''

*Die Datentypen für die Attribute sind nicht korrekt - z.B. Title, Interpret, Album sind nominale Datentypen, usw.
*Verwendet bitte bei den Datentypen ein durchgängiges Schema:
** nominal, ordinal, discrete, continuous, binary
** (quantitative kann nämlich discrete oder continuous bedeuten)
*Also entweder haben die Gruppen 9 und 10 von euch diese Tabelle abgekupfert oder umgekehrt? :-/
*Der Datensatz ist als solcher ''multidimensional''
*Subgrouping ist soweit ich weiss bei ID3 tags nicht möglich (ein Genre als Unterpunkt eines anderen) - Woher kommt die Zusatzinformation, die so nicht in den Daten vorkommt?
*IMHO ist die Anzahl verschiedener Genres in einer MP3 Sammlung nicht wirklich "enorm" (ich habe z.B. über 30GB MP3s, aber nur knapp über 100 verschiedene Genres...)
*Für die relativ wenig Information, die in eurer graphischen Repräsentation visualisiert wird, wird sehr viel Platz verwendet. (Ein normaler TreeView mit dazugehängten Balken würde dieselbe Information transportieren, allerdings wesentlich weniger Platz benötigen und damit eine bessere Übersicht ermöglichen.)
*Ihr visualisiert eigentlich nur 2 verschiedene Datenattribute: Genre und Anzahl von Songs innerhalb des Genres. Dafür braucht man eigentlich nicht wirklich besondere, interaktive InfoVis Methoden (die sollten dazu dienen, möglichst viele Parameter übersichtlich und sinnvoll miteinenader zu kombinieren um neues Wissen zu generieren)
*Ihr geht auf die selben Variablen ein wie Gruppe 4 -- der Kopierverdacht erhärtet sich...
*Kurz und gut: überdenkt euer Konzept noch einmal und versucht mehr Information zu integrieren bzw. andere Darstellungsvarianten dafür...

-- [[User:Iwolf|Wolfgang Aigner]] 14:31, 25 November 2005 (CET)

== Kopierverdacht + Korrektur ==

*Bezüglich Ihres Verdachts, dass die Datensatzanalyse abgeschrieben sein könnte, kann ich nur sagen, dass ich diesen Teil selbst erarbeitet habe.
:'''''--> also das halte ich für eine sehr gewagte Theorie und es fällt mir ziemlich schwer das zu glauben... ihr habt die fragliche Tabelle wesentlich später online gestellt und haargenau dieselben Fehler wie die Gruppen 9 und 10 (sogar die Rechtschreibfehler!) - also Zufall hin oder her - 3x die (fast) gleiche Tabelle und identische inhaltliche sowie Rechtschreibfehler...'''''

** --> Rechtschreibfehler konnte ich bei meiner Tabelle keine finden. Zudem haben wir in unserer Angabe als einzige den Datentyp "Quantitative" angegeben. Bei einem Vergleich mit den anderen Gruppen konnte ich ebenfalls Ähnlichkeiten feststellen, die aber wohl eher damit zu begründen sind, dass wir alle das Thema MP3 gewählt haben.

::'''''--> Stimmt, da habt ihr recht mit den Rechtschreibfehlern und Datentypen - sorry. Nichts desto trotz sehen sich die Datensatzanalysen sehr ähnlich. Aber gut, das soll jetzt nicht das große Problem sein. Viel wichtiger ist die inhaltliche Verbesserung eures Konzeptes - da solltet ihr noch einige Überlegungen reinstecken!''''' -- [[User:Iwolf|Wolfgang Aigner]] 08:56, 28 November 2005 (CET)

*Der Gründe für die Fehler in meiner Ausarbeitung liegen in folgenden Punkten:
**Leider hatte ich Probleme im Internet (für mich) schlüssige Unterlagen zu diesem Thema finden.
** Zudem waren die Definitionen, die ich zu den diversen Datentypen überhaupt finden konnte oft widersprüchlich.
** So war es für mich schwierig die Daten eindeutig zuzuordnen.
:'''''--> Datentypen sollten eigentlich Grundwissen z.B. aus Statistik sein. Hier nochmal eine kompakte Zusammenfassung: [[Data Type]].'''''

*Bei den Referenzen unserer Ausarbeitung können sie die Internetseite finden, die ich schlussendlich hauptsächlich als Ausgangsbasis für meine Ausarbeitung verwendet habe. Diese Seite stellt eine der detailliertesten Angaben dar, die ich finden konnte.

*Das sich die Fehler in der Analyse zum Teil überdecken, kann ich mir persönlich nur damit erklären, dass die anderen Gruppen mit ähnlichen Schwierigkeiten zu kämpfen hatten.

*Natürlich werden wir unsere Abgabe nochmals überarbeiten bzw. überdenken. Daher möchten wir gerne wissen, bis wann Sie uns hierfür Zeit geben.
:'''''--> Ich erwarte mir eine komplette Überarbeitung eures Konzepts (inkl. Datensatzanalyse) und vor allem einen verbesserten Designentwurf. Eine Deadline für die Überarbeitung gibt es nicht - es ist in eurem Interesse, das so bald wie möglich zu tun, schliesslich ist die Abgabe des Prototypen in weniger als 4 Wochen.'''''

-- [[User:Iwolf|Wolfgang Aigner]] 12:52, 26 November 2005 (CET)

== Überarbeitetes/neues Konzept ==

*Ein überarbeitetes bzw. generell neu ausgelegtes Konzept ist jetzt online.

*Wir glauben, mit unserm neuen Konzept die Aufgabenstellung nun wesentlich besser "getroffen" zu haben und bitten sie wieder um ein kurzes Feedback, damit wir mit unserer Implementierung beginnen können.

**Vielen Dank, G4

Teaching talk:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-11T21:43:02Z

UE-InfoVis0506 9335259: /* Überarbeitetes/neues Konzept */

'''Kommentar zum Konzept:'''

*Die Datentypen für die Attribute sind nicht korrekt - z.B. Title, Interpret, Album sind nominale Datentypen, usw.
*Verwendet bitte bei den Datentypen ein durchgängiges Schema:
** nominal, ordinal, discrete, continuous, binary
** (quantitative kann nämlich discrete oder continuous bedeuten)
*Also entweder haben die Gruppen 9 und 10 von euch diese Tabelle abgekupfert oder umgekehrt? :-/
*Der Datensatz ist als solcher ''multidimensional''
*Subgrouping ist soweit ich weiss bei ID3 tags nicht möglich (ein Genre als Unterpunkt eines anderen) - Woher kommt die Zusatzinformation, die so nicht in den Daten vorkommt?
*IMHO ist die Anzahl verschiedener Genres in einer MP3 Sammlung nicht wirklich "enorm" (ich habe z.B. über 30GB MP3s, aber nur knapp über 100 verschiedene Genres...)
*Für die relativ wenig Information, die in eurer graphischen Repräsentation visualisiert wird, wird sehr viel Platz verwendet. (Ein normaler TreeView mit dazugehängten Balken würde dieselbe Information transportieren, allerdings wesentlich weniger Platz benötigen und damit eine bessere Übersicht ermöglichen.)
*Ihr visualisiert eigentlich nur 2 verschiedene Datenattribute: Genre und Anzahl von Songs innerhalb des Genres. Dafür braucht man eigentlich nicht wirklich besondere, interaktive InfoVis Methoden (die sollten dazu dienen, möglichst viele Parameter übersichtlich und sinnvoll miteinenader zu kombinieren um neues Wissen zu generieren)
*Ihr geht auf die selben Variablen ein wie Gruppe 4 -- der Kopierverdacht erhärtet sich...
*Kurz und gut: überdenkt euer Konzept noch einmal und versucht mehr Information zu integrieren bzw. andere Darstellungsvarianten dafür...

-- [[User:Iwolf|Wolfgang Aigner]] 14:31, 25 November 2005 (CET)

== Kopierverdacht + Korrektur ==

*Bezüglich Ihres Verdachts, dass die Datensatzanalyse abgeschrieben sein könnte, kann ich nur sagen, dass ich diesen Teil selbst erarbeitet habe.
:'''''--> also das halte ich für eine sehr gewagte Theorie und es fällt mir ziemlich schwer das zu glauben... ihr habt die fragliche Tabelle wesentlich später online gestellt und haargenau dieselben Fehler wie die Gruppen 9 und 10 (sogar die Rechtschreibfehler!) - also Zufall hin oder her - 3x die (fast) gleiche Tabelle und identische inhaltliche sowie Rechtschreibfehler...'''''

** --> Rechtschreibfehler konnte ich bei meiner Tabelle keine finden. Zudem haben wir in unserer Angabe als einzige den Datentyp "Quantitative" angegeben. Bei einem Vergleich mit den anderen Gruppen konnte ich ebenfalls Ähnlichkeiten feststellen, die aber wohl eher damit zu begründen sind, dass wir alle das Thema MP3 gewählt haben.

::'''''--> Stimmt, da habt ihr recht mit den Rechtschreibfehlern und Datentypen - sorry. Nichts desto trotz sehen sich die Datensatzanalysen sehr ähnlich. Aber gut, das soll jetzt nicht das große Problem sein. Viel wichtiger ist die inhaltliche Verbesserung eures Konzeptes - da solltet ihr noch einige Überlegungen reinstecken!''''' -- [[User:Iwolf|Wolfgang Aigner]] 08:56, 28 November 2005 (CET)

*Der Gründe für die Fehler in meiner Ausarbeitung liegen in folgenden Punkten:
**Leider hatte ich Probleme im Internet (für mich) schlüssige Unterlagen zu diesem Thema finden.
** Zudem waren die Definitionen, die ich zu den diversen Datentypen überhaupt finden konnte oft widersprüchlich.
** So war es für mich schwierig die Daten eindeutig zuzuordnen.
:'''''--> Datentypen sollten eigentlich Grundwissen z.B. aus Statistik sein. Hier nochmal eine kompakte Zusammenfassung: [[Data Type]].'''''

*Bei den Referenzen unserer Ausarbeitung können sie die Internetseite finden, die ich schlussendlich hauptsächlich als Ausgangsbasis für meine Ausarbeitung verwendet habe. Diese Seite stellt eine der detailliertesten Angaben dar, die ich finden konnte.

*Das sich die Fehler in der Analyse zum Teil überdecken, kann ich mir persönlich nur damit erklären, dass die anderen Gruppen mit ähnlichen Schwierigkeiten zu kämpfen hatten.

*Natürlich werden wir unsere Abgabe nochmals überarbeiten bzw. überdenken. Daher möchten wir gerne wissen, bis wann Sie uns hierfür Zeit geben.
:'''''--> Ich erwarte mir eine komplette Überarbeitung eures Konzepts (inkl. Datensatzanalyse) und vor allem einen verbesserten Designentwurf. Eine Deadline für die Überarbeitung gibt es nicht - es ist in eurem Interesse, das so bald wie möglich zu tun, schliesslich ist die Abgabe des Prototypen in weniger als 4 Wochen.'''''

-- [[User:Iwolf|Wolfgang Aigner]] 12:52, 26 November 2005 (CET)

== Überarbeitetes/neues Konzept ==

*Ein überarbeitetes bzw. generell neu ausgelegtes Konzept ist jetzt online.

*Wir glauben, mit unserm neuen Konzept die Aufgabenstellung nun wesentlich
besser "getroffen" zu haben und bitten sie wieder um ein kurzes Feedback,
damit wir mit unserer Implementierung beginnen können.

**Danke, G4

Teaching talk:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-11T21:42:43Z

UE-InfoVis0506 9335259: /* Überarbeitetes/neues Konzept */

'''Kommentar zum Konzept:'''

*Die Datentypen für die Attribute sind nicht korrekt - z.B. Title, Interpret, Album sind nominale Datentypen, usw.
*Verwendet bitte bei den Datentypen ein durchgängiges Schema:
** nominal, ordinal, discrete, continuous, binary
** (quantitative kann nämlich discrete oder continuous bedeuten)
*Also entweder haben die Gruppen 9 und 10 von euch diese Tabelle abgekupfert oder umgekehrt? :-/
*Der Datensatz ist als solcher ''multidimensional''
*Subgrouping ist soweit ich weiss bei ID3 tags nicht möglich (ein Genre als Unterpunkt eines anderen) - Woher kommt die Zusatzinformation, die so nicht in den Daten vorkommt?
*IMHO ist die Anzahl verschiedener Genres in einer MP3 Sammlung nicht wirklich "enorm" (ich habe z.B. über 30GB MP3s, aber nur knapp über 100 verschiedene Genres...)
*Für die relativ wenig Information, die in eurer graphischen Repräsentation visualisiert wird, wird sehr viel Platz verwendet. (Ein normaler TreeView mit dazugehängten Balken würde dieselbe Information transportieren, allerdings wesentlich weniger Platz benötigen und damit eine bessere Übersicht ermöglichen.)
*Ihr visualisiert eigentlich nur 2 verschiedene Datenattribute: Genre und Anzahl von Songs innerhalb des Genres. Dafür braucht man eigentlich nicht wirklich besondere, interaktive InfoVis Methoden (die sollten dazu dienen, möglichst viele Parameter übersichtlich und sinnvoll miteinenader zu kombinieren um neues Wissen zu generieren)
*Ihr geht auf die selben Variablen ein wie Gruppe 4 -- der Kopierverdacht erhärtet sich...
*Kurz und gut: überdenkt euer Konzept noch einmal und versucht mehr Information zu integrieren bzw. andere Darstellungsvarianten dafür...

-- [[User:Iwolf|Wolfgang Aigner]] 14:31, 25 November 2005 (CET)

== Kopierverdacht + Korrektur ==

*Bezüglich Ihres Verdachts, dass die Datensatzanalyse abgeschrieben sein könnte, kann ich nur sagen, dass ich diesen Teil selbst erarbeitet habe.
:'''''--> also das halte ich für eine sehr gewagte Theorie und es fällt mir ziemlich schwer das zu glauben... ihr habt die fragliche Tabelle wesentlich später online gestellt und haargenau dieselben Fehler wie die Gruppen 9 und 10 (sogar die Rechtschreibfehler!) - also Zufall hin oder her - 3x die (fast) gleiche Tabelle und identische inhaltliche sowie Rechtschreibfehler...'''''

** --> Rechtschreibfehler konnte ich bei meiner Tabelle keine finden. Zudem haben wir in unserer Angabe als einzige den Datentyp "Quantitative" angegeben. Bei einem Vergleich mit den anderen Gruppen konnte ich ebenfalls Ähnlichkeiten feststellen, die aber wohl eher damit zu begründen sind, dass wir alle das Thema MP3 gewählt haben.

::'''''--> Stimmt, da habt ihr recht mit den Rechtschreibfehlern und Datentypen - sorry. Nichts desto trotz sehen sich die Datensatzanalysen sehr ähnlich. Aber gut, das soll jetzt nicht das große Problem sein. Viel wichtiger ist die inhaltliche Verbesserung eures Konzeptes - da solltet ihr noch einige Überlegungen reinstecken!''''' -- [[User:Iwolf|Wolfgang Aigner]] 08:56, 28 November 2005 (CET)

*Der Gründe für die Fehler in meiner Ausarbeitung liegen in folgenden Punkten:
**Leider hatte ich Probleme im Internet (für mich) schlüssige Unterlagen zu diesem Thema finden.
** Zudem waren die Definitionen, die ich zu den diversen Datentypen überhaupt finden konnte oft widersprüchlich.
** So war es für mich schwierig die Daten eindeutig zuzuordnen.
:'''''--> Datentypen sollten eigentlich Grundwissen z.B. aus Statistik sein. Hier nochmal eine kompakte Zusammenfassung: [[Data Type]].'''''

*Bei den Referenzen unserer Ausarbeitung können sie die Internetseite finden, die ich schlussendlich hauptsächlich als Ausgangsbasis für meine Ausarbeitung verwendet habe. Diese Seite stellt eine der detailliertesten Angaben dar, die ich finden konnte.

*Das sich die Fehler in der Analyse zum Teil überdecken, kann ich mir persönlich nur damit erklären, dass die anderen Gruppen mit ähnlichen Schwierigkeiten zu kämpfen hatten.

*Natürlich werden wir unsere Abgabe nochmals überarbeiten bzw. überdenken. Daher möchten wir gerne wissen, bis wann Sie uns hierfür Zeit geben.
:'''''--> Ich erwarte mir eine komplette Überarbeitung eures Konzepts (inkl. Datensatzanalyse) und vor allem einen verbesserten Designentwurf. Eine Deadline für die Überarbeitung gibt es nicht - es ist in eurem Interesse, das so bald wie möglich zu tun, schliesslich ist die Abgabe des Prototypen in weniger als 4 Wochen.'''''

-- [[User:Iwolf|Wolfgang Aigner]] 12:52, 26 November 2005 (CET)

== Überarbeitetes/neues Konzept ==

EIn überarbeitetes bzw. generell neu ausgelegtes Konzept ist jetzt online.

Wir glauben, mit unserm neuen Konzept die Aufgabenstellung nun wesentlich
besser "getroffen" zu haben und bitten sie wieder um ein kurzes Feedback,
damit wir mit unserer Implementierung beginnen können.

Danke, G4

Teaching talk:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-11T21:38:04Z

UE-InfoVis0506 9335259:

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-11T21:35:00Z

UE-InfoVis0506 9335259: /* Who should use this kind of visualization technique? */

== Topic ==
'''MP3 Archive Visualization'''
== Specification of the Application Area and the given Dataset ==
=== Application area Analysis ===

Since we have chosen the MP3 Archive Visualization, our job will be the creation of a clearly arranged visualization for a big music archive consisting of several thousand files.

This can be achieved by using the already existing values of the container format ID3, additional attributes of the iTunes library, attributes of the music files themselves, as well as some system values.

Due the fact that these given sources already provide loads of different Information, we will try to create and present additional Information by combining some prior chosen values in a reasonable way.

Thus we have to keep in mind that ID3 for example theoretically indeed supports a huge amount of input values but in most cases only the most common values like Album, Interpret or Year are specified correctly.

Therefore we will only use some of these entries in our prototype.

=== Dataset Analysis ===

The values we will use in our project consist of nominal, discrete and ordinal data types and are for themselves all one-dimensional.

The table below shows a complete listing:

[[Image:tabelle_datentyp_g4.png]]

The complete data set is multi-dimensional and consists of all attributes listed above.

== Analysis of the Target User Group ==

=== Who should use this kind of visualization technique? ===

This visualization technique is mainly meant for the 'end-users', that is someone who collects lots of MP3s. With 'lots of MP3s' we mean quite a few GBs, just more than 30 GBs. Our visualization should help the user to get an overview of his collection and his listening-habits.
This visualization technique could also be interesting for the band and the music industry, if they want to produce a new album. for example: a band (like "Radiohead") who changed their music-style over the years wants to know which style is preferred more. But therefore they have to compare these datas from many users.

=== What are the characteristics of the target group? ===

People of this group are music enthusiasts. They have thousands of mp3s on their hard-disk and love it to collect them. Most of them have lost track of their collection, on the strength of the abundance of their collection. These people mainly receive their mp3s from the internet instead of buying CDs, because they like to see their whole music-collection at a glance.

=== Are there any known / often used Methods / Visualisation Techniques? ===

No, we don't know any similar visualization technique. ITunes only shows textbased info about how often a song was heard.

== Intended Purpose of our Visualization ==

=== What should be achieved with this visualization? ===
A better information representation of the MP3s should be achieved. Our visualization should help the user to get an overview of his collection and his listening-habits. He will get information about the tracks, in reference to a special artist, which he often listens to and to those which he has never heard before. The representation of the data should be expressive, precise and self-explanatory.

=== Which tasks should be solved? ===
By using this visualization technique, the user will get information about a chosen artist and his discography. For example: in his database the user has got the band "Radiohead", who produced albums over 15 years and in this period they changed their music-style from alternative rock to experimental electronic. The visualization will show him from which producing period he has got more MP3s and which period he likes more, by counting the number of listenings of each song. The result could be that he has got more MP3s from their early years, but likes the experimental electronic tracks more.

=== Questions that should be solved with this visualization technique ===

== Proposal of Design ==

=== Kind of Visualization / Visualization Details ===

When the user opens the "Interpret-Analyser" he will be prompted via a text-message in the main-window to click on an artist/band in the right upper window. The artists/bands are sorted alphabetically and the subject of interest can be found by scrolling the window vertically. If the user chose an artist/band in the upper right window, the main window will visualize him following details on the demanded item:

*The x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data
*The y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library

The respective maximum on the y-axis will give a first overview on how many songs the specific library contains per artist/band per year. Though the users' library might not be complete the visualization allows drawing conclusions according to the artist/band-activities over the last years. In any cases we assume that the user applies the "Interpret-Analyser" to artists/bands whereof he collected the whole discography and not only one song.

For each song (= one data point) one horizontal bar is drawn along the y-axis. That means for example if the library contains 34 songs by the band "Queen" with publication-year "1985", 34 bars are drawn at the x-axis value "1985" along the vertical y-axis.

In addition to that each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently).

Further on the user can interact and influence the characteristic of the visualization by using a slider positioned in the lower right in the graphic below. Via the slider a more objective image can be drawn according to the actual point of interest. This slider with a value-range from "0" to "10" represents the counts how often a song was played. It allows setting a threshold. The default value is "3" and means that songs that were played less than 3 times do not appear coloured, but as grey bars vertically above the coloured ones along the y-axis. If for example someone drags the slider to the position with value "10" and only 1 song out of 27 with a special publishing year was played more than 9 times the "Interpret-Analyser" shows 1 coloured and 26 grey bars at the according year. This could for instance help if someone is on the way to filter out his absolute favourites of an artist/band.

As it is mentioned above the "Interpret-Analyser" represents highly interesting visualizations for End-Users but it might also prevent outstanding features for the Music-Industry respectively bands, who work on a Come-Back. This however would assume to arising the data of a rather big audience, what could for example be achieved via a contest.

=== Visual Mapping ===

'''2D Diagram''':

*'''"X-AXIS":''' the x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data

*'''"Y-AXIS":''' the y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library. The height of the vertical bar-column represents the "Song-Occurrence" per year

*'''"Colour":''' each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently). Grey bars represent songs that did not pass the adjusted threshold

=== Specification of used Techniques / applied Principles ===

*'''Bar X Plot:''' In this plot, one vertical bar is drawn for each data point [StatSoft, 2003]

*'''Histograms, 2D:''' 2D histograms present a graphical representation of the frequency distribution of the selected variable(s) in which the columns are drawn over the class intervals and the heights of the columns are proportional to the class frequencies. [StatSoft, 2003]

*'''Colour-Range, Linking & Brushing:''' A colour-Range representing the levels between "not played for a long time" and "recently played" (s. slide 100 of Info_Vis0.pdf handed out in the course 188.305 VO InfoVis)

*'''Scatterplot, 2D:''' The scatterplot visualizes a relation (correlation) between two variables X and Y (e.g., weight and height). Individual data points are represented in two-dimensional space (see below), where axes represent the variables (X on the horizontal axis and Y on the vertical axis). The two coordinates (X and Y) that determine the location of each point correspond to its specific values on the two variables. [StatSoft, 2003]

*'''Dynamic Queries:''' Adjusting the slider generates dynamic queries [StatSoft, 2003]

=== Possibilities of User-Interaction ===

*Select item of interest (artist/band)
**Get artist/band-details

*Adjust the slider to influence the threshold
**Get individual Visualizations according to the users' point of interest

=== Mockup / Fake Screenshot ===

[[Image:Interpret-Analyser.png]]

== References ==

[Wikipedia, 2005a] ID3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[Wikipedia, 2005b] MP3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://en.wikipedia.org/wiki/Mp3

[Id3.org, 2004] ID3v2 frames, Id3.org, Last updated: 28. February, 2004, Retrieved at: November 22, 2005, http://www.id3.org/frames.html

[Montclaire, 2000] Data Types, Department of Science and Mathematics at Montclair State University, Last updated: 3. August, 2000, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[StatSoft, 2003] Graphical Analytic Techniques, Last updated: 2003, Retrieved at: November 24, 2005, http://www.statsoft.com/textbook/stgraph.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-11T21:33:46Z

UE-InfoVis0506 9335259: /* What are the characteristics of the target group? */

== Topic ==
'''MP3 Archive Visualization'''
== Specification of the Application Area and the given Dataset ==
=== Application area Analysis ===

Since we have chosen the MP3 Archive Visualization, our job will be the creation of a clearly arranged visualization for a big music archive consisting of several thousand files.

This can be achieved by using the already existing values of the container format ID3, additional attributes of the iTunes library, attributes of the music files themselves, as well as some system values.

Due the fact that these given sources already provide loads of different Information, we will try to create and present additional Information by combining some prior chosen values in a reasonable way.

Thus we have to keep in mind that ID3 for example theoretically indeed supports a huge amount of input values but in most cases only the most common values like Album, Interpret or Year are specified correctly.

Therefore we will only use some of these entries in our prototype.

=== Dataset Analysis ===

The values we will use in our project consist of nominal, discrete and ordinal data types and are for themselves all one-dimensional.

The table below shows a complete listing:

[[Image:tabelle_datentyp_g4.png]]

The complete data set is multi-dimensional and consists of all attributes listed above.

== Analysis of the Target User Group ==

=== Who should use this kind of visualization technique? ===

This visualization technique is mainly meant for the 'end-users', that is someone who collects lots of MP3s. With 'lots of MP3s' we mean quite a few GBs, just more than 30 GBs. Our visualization should help the user to get an overview of his collection and his listening-habits.
This visualization technique could also be interesting for the band and the music industry, if they want to produce a new album. for example: a band (like radiohead) who changed their musicstyle over the years and wants to know which style is preferd more. But therefor they have to compare these dates from many users.

=== What are the characteristics of the target group? ===

People of this group are music enthusiasts. They have thousands of mp3s on their hard-disk and love it to collect them. Most of them have lost track of their collection, on the strength of the abundance of their collection. These people mainly receive their mp3s from the internet instead of buying CDs, because they like to see their whole music-collection at a glance.

=== Are there any known / often used Methods / Visualisation Techniques? ===

No, we don't know any similar visualization technique. ITunes only shows textbased info about how often a song was heard.

== Intended Purpose of our Visualization ==

=== What should be achieved with this visualization? ===
A better information representation of the MP3s should be achieved. Our visualization should help the user to get an overview of his collection and his listening-habits. He will get information about the tracks, in reference to a special artist, which he often listens to and to those which he has never heard before. The representation of the data should be expressive, precise and self-explanatory.

=== Which tasks should be solved? ===
By using this visualization technique, the user will get information about a chosen artist and his discography. For example: in his database the user has got the band "Radiohead", who produced albums over 15 years and in this period they changed their music-style from alternative rock to experimental electronic. The visualization will show him from which producing period he has got more MP3s and which period he likes more, by counting the number of listenings of each song. The result could be that he has got more MP3s from their early years, but likes the experimental electronic tracks more.

=== Questions that should be solved with this visualization technique ===

== Proposal of Design ==

=== Kind of Visualization / Visualization Details ===

When the user opens the "Interpret-Analyser" he will be prompted via a text-message in the main-window to click on an artist/band in the right upper window. The artists/bands are sorted alphabetically and the subject of interest can be found by scrolling the window vertically. If the user chose an artist/band in the upper right window, the main window will visualize him following details on the demanded item:

*The x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data
*The y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library

The respective maximum on the y-axis will give a first overview on how many songs the specific library contains per artist/band per year. Though the users' library might not be complete the visualization allows drawing conclusions according to the artist/band-activities over the last years. In any cases we assume that the user applies the "Interpret-Analyser" to artists/bands whereof he collected the whole discography and not only one song.

For each song (= one data point) one horizontal bar is drawn along the y-axis. That means for example if the library contains 34 songs by the band "Queen" with publication-year "1985", 34 bars are drawn at the x-axis value "1985" along the vertical y-axis.

In addition to that each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently).

Further on the user can interact and influence the characteristic of the visualization by using a slider positioned in the lower right in the graphic below. Via the slider a more objective image can be drawn according to the actual point of interest. This slider with a value-range from "0" to "10" represents the counts how often a song was played. It allows setting a threshold. The default value is "3" and means that songs that were played less than 3 times do not appear coloured, but as grey bars vertically above the coloured ones along the y-axis. If for example someone drags the slider to the position with value "10" and only 1 song out of 27 with a special publishing year was played more than 9 times the "Interpret-Analyser" shows 1 coloured and 26 grey bars at the according year. This could for instance help if someone is on the way to filter out his absolute favourites of an artist/band.

As it is mentioned above the "Interpret-Analyser" represents highly interesting visualizations for End-Users but it might also prevent outstanding features for the Music-Industry respectively bands, who work on a Come-Back. This however would assume to arising the data of a rather big audience, what could for example be achieved via a contest.

=== Visual Mapping ===

'''2D Diagram''':

*'''"X-AXIS":''' the x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data

*'''"Y-AXIS":''' the y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library. The height of the vertical bar-column represents the "Song-Occurrence" per year

*'''"Colour":''' each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently). Grey bars represent songs that did not pass the adjusted threshold

=== Specification of used Techniques / applied Principles ===

*'''Bar X Plot:''' In this plot, one vertical bar is drawn for each data point [StatSoft, 2003]

*'''Histograms, 2D:''' 2D histograms present a graphical representation of the frequency distribution of the selected variable(s) in which the columns are drawn over the class intervals and the heights of the columns are proportional to the class frequencies. [StatSoft, 2003]

*'''Colour-Range, Linking & Brushing:''' A colour-Range representing the levels between "not played for a long time" and "recently played" (s. slide 100 of Info_Vis0.pdf handed out in the course 188.305 VO InfoVis)

*'''Scatterplot, 2D:''' The scatterplot visualizes a relation (correlation) between two variables X and Y (e.g., weight and height). Individual data points are represented in two-dimensional space (see below), where axes represent the variables (X on the horizontal axis and Y on the vertical axis). The two coordinates (X and Y) that determine the location of each point correspond to its specific values on the two variables. [StatSoft, 2003]

*'''Dynamic Queries:''' Adjusting the slider generates dynamic queries [StatSoft, 2003]

=== Possibilities of User-Interaction ===

*Select item of interest (artist/band)
**Get artist/band-details

*Adjust the slider to influence the threshold
**Get individual Visualizations according to the users' point of interest

=== Mockup / Fake Screenshot ===

[[Image:Interpret-Analyser.png]]

== References ==

[Wikipedia, 2005a] ID3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[Wikipedia, 2005b] MP3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://en.wikipedia.org/wiki/Mp3

[Id3.org, 2004] ID3v2 frames, Id3.org, Last updated: 28. February, 2004, Retrieved at: November 22, 2005, http://www.id3.org/frames.html

[Montclaire, 2000] Data Types, Department of Science and Mathematics at Montclair State University, Last updated: 3. August, 2000, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[StatSoft, 2003] Graphical Analytic Techniques, Last updated: 2003, Retrieved at: November 24, 2005, http://www.statsoft.com/textbook/stgraph.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-11T21:33:14Z

UE-InfoVis0506 9335259: /* Are there any known / often used Methods / Visualisation Techniques? */

== Topic ==
'''MP3 Archive Visualization'''
== Specification of the Application Area and the given Dataset ==
=== Application area Analysis ===

Since we have chosen the MP3 Archive Visualization, our job will be the creation of a clearly arranged visualization for a big music archive consisting of several thousand files.

This can be achieved by using the already existing values of the container format ID3, additional attributes of the iTunes library, attributes of the music files themselves, as well as some system values.

Due the fact that these given sources already provide loads of different Information, we will try to create and present additional Information by combining some prior chosen values in a reasonable way.

Thus we have to keep in mind that ID3 for example theoretically indeed supports a huge amount of input values but in most cases only the most common values like Album, Interpret or Year are specified correctly.

Therefore we will only use some of these entries in our prototype.

=== Dataset Analysis ===

The values we will use in our project consist of nominal, discrete and ordinal data types and are for themselves all one-dimensional.

The table below shows a complete listing:

[[Image:tabelle_datentyp_g4.png]]

The complete data set is multi-dimensional and consists of all attributes listed above.

== Analysis of the Target User Group ==

=== Who should use this kind of visualization technique? ===

This visualization technique is mainly meant for the 'end-users', that is someone who collects lots of MP3s. With 'lots of MP3s' we mean quite a few GBs, just more than 30 GBs. Our visualization should help the user to get an overview of his collection and his listening-habits.
This visualization technique could also be interesting for the band and the music industry, if they want to produce a new album. for example: a band (like radiohead) who changed their musicstyle over the years and wants to know which style is preferd more. But therefor they have to compare these dates from many users.

=== What are the characteristics of the target group? ===

People of this group are music enthusiasts. They have thousands of mp3s on their harddisk and love it to collect them. Most of them have lost track of their collection, on the strength of the abundance of their collection. These people mainly receive their mp3s from the internet instead of buying CDs, because they like to see their whole musiccollection at a glance.

=== Are there any known / often used Methods / Visualisation Techniques? ===

No, we don't know any similar visualization technique. ITunes only shows textbased info about how often a song was heard.

== Intended Purpose of our Visualization ==

=== What should be achieved with this visualization? ===
A better information representation of the MP3s should be achieved. Our visualization should help the user to get an overview of his collection and his listening-habits. He will get information about the tracks, in reference to a special artist, which he often listens to and to those which he has never heard before. The representation of the data should be expressive, precise and self-explanatory.

=== Which tasks should be solved? ===
By using this visualization technique, the user will get information about a chosen artist and his discography. For example: in his database the user has got the band "Radiohead", who produced albums over 15 years and in this period they changed their music-style from alternative rock to experimental electronic. The visualization will show him from which producing period he has got more MP3s and which period he likes more, by counting the number of listenings of each song. The result could be that he has got more MP3s from their early years, but likes the experimental electronic tracks more.

=== Questions that should be solved with this visualization technique ===

== Proposal of Design ==

=== Kind of Visualization / Visualization Details ===

When the user opens the "Interpret-Analyser" he will be prompted via a text-message in the main-window to click on an artist/band in the right upper window. The artists/bands are sorted alphabetically and the subject of interest can be found by scrolling the window vertically. If the user chose an artist/band in the upper right window, the main window will visualize him following details on the demanded item:

*The x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data
*The y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library

The respective maximum on the y-axis will give a first overview on how many songs the specific library contains per artist/band per year. Though the users' library might not be complete the visualization allows drawing conclusions according to the artist/band-activities over the last years. In any cases we assume that the user applies the "Interpret-Analyser" to artists/bands whereof he collected the whole discography and not only one song.

For each song (= one data point) one horizontal bar is drawn along the y-axis. That means for example if the library contains 34 songs by the band "Queen" with publication-year "1985", 34 bars are drawn at the x-axis value "1985" along the vertical y-axis.

In addition to that each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently).

Further on the user can interact and influence the characteristic of the visualization by using a slider positioned in the lower right in the graphic below. Via the slider a more objective image can be drawn according to the actual point of interest. This slider with a value-range from "0" to "10" represents the counts how often a song was played. It allows setting a threshold. The default value is "3" and means that songs that were played less than 3 times do not appear coloured, but as grey bars vertically above the coloured ones along the y-axis. If for example someone drags the slider to the position with value "10" and only 1 song out of 27 with a special publishing year was played more than 9 times the "Interpret-Analyser" shows 1 coloured and 26 grey bars at the according year. This could for instance help if someone is on the way to filter out his absolute favourites of an artist/band.

As it is mentioned above the "Interpret-Analyser" represents highly interesting visualizations for End-Users but it might also prevent outstanding features for the Music-Industry respectively bands, who work on a Come-Back. This however would assume to arising the data of a rather big audience, what could for example be achieved via a contest.

=== Visual Mapping ===

'''2D Diagram''':

*'''"X-AXIS":''' the x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data

*'''"Y-AXIS":''' the y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library. The height of the vertical bar-column represents the "Song-Occurrence" per year

*'''"Colour":''' each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently). Grey bars represent songs that did not pass the adjusted threshold

=== Specification of used Techniques / applied Principles ===

*'''Bar X Plot:''' In this plot, one vertical bar is drawn for each data point [StatSoft, 2003]

*'''Histograms, 2D:''' 2D histograms present a graphical representation of the frequency distribution of the selected variable(s) in which the columns are drawn over the class intervals and the heights of the columns are proportional to the class frequencies. [StatSoft, 2003]

*'''Colour-Range, Linking & Brushing:''' A colour-Range representing the levels between "not played for a long time" and "recently played" (s. slide 100 of Info_Vis0.pdf handed out in the course 188.305 VO InfoVis)

*'''Scatterplot, 2D:''' The scatterplot visualizes a relation (correlation) between two variables X and Y (e.g., weight and height). Individual data points are represented in two-dimensional space (see below), where axes represent the variables (X on the horizontal axis and Y on the vertical axis). The two coordinates (X and Y) that determine the location of each point correspond to its specific values on the two variables. [StatSoft, 2003]

*'''Dynamic Queries:''' Adjusting the slider generates dynamic queries [StatSoft, 2003]

=== Possibilities of User-Interaction ===

*Select item of interest (artist/band)
**Get artist/band-details

*Adjust the slider to influence the threshold
**Get individual Visualizations according to the users' point of interest

=== Mockup / Fake Screenshot ===

[[Image:Interpret-Analyser.png]]

== References ==

[Wikipedia, 2005a] ID3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[Wikipedia, 2005b] MP3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://en.wikipedia.org/wiki/Mp3

[Id3.org, 2004] ID3v2 frames, Id3.org, Last updated: 28. February, 2004, Retrieved at: November 22, 2005, http://www.id3.org/frames.html

[Montclaire, 2000] Data Types, Department of Science and Mathematics at Montclair State University, Last updated: 3. August, 2000, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[StatSoft, 2003] Graphical Analytic Techniques, Last updated: 2003, Retrieved at: November 24, 2005, http://www.statsoft.com/textbook/stgraph.html

Teaching:TUW - UE InfoVis WS 2005/06 - Gruppe G4 - Aufgabe 3 - Design

2005-12-11T21:32:43Z

UE-InfoVis0506 9335259: /* What should be achieved with this visualization? */

== Topic ==
'''MP3 Archive Visualization'''
== Specification of the Application Area and the given Dataset ==
=== Application area Analysis ===

Since we have chosen the MP3 Archive Visualization, our job will be the creation of a clearly arranged visualization for a big music archive consisting of several thousand files.

This can be achieved by using the already existing values of the container format ID3, additional attributes of the iTunes library, attributes of the music files themselves, as well as some system values.

Due the fact that these given sources already provide loads of different Information, we will try to create and present additional Information by combining some prior chosen values in a reasonable way.

Thus we have to keep in mind that ID3 for example theoretically indeed supports a huge amount of input values but in most cases only the most common values like Album, Interpret or Year are specified correctly.

Therefore we will only use some of these entries in our prototype.

=== Dataset Analysis ===

The values we will use in our project consist of nominal, discrete and ordinal data types and are for themselves all one-dimensional.

The table below shows a complete listing:

[[Image:tabelle_datentyp_g4.png]]

The complete data set is multi-dimensional and consists of all attributes listed above.

== Analysis of the Target User Group ==

=== Who should use this kind of visualization technique? ===

This visualization technique is mainly meant for the 'end-users', that is someone who collects lots of MP3s. With 'lots of MP3s' we mean quite a few GBs, just more than 30 GBs. Our visualization should help the user to get an overview of his collection and his listening-habits.
This visualization technique could also be interesting for the band and the music industry, if they want to produce a new album. for example: a band (like radiohead) who changed their musicstyle over the years and wants to know which style is preferd more. But therefor they have to compare these dates from many users.

=== What are the characteristics of the target group? ===

People of this group are music enthusiasts. They have thousands of mp3s on their harddisk and love it to collect them. Most of them have lost track of their collection, on the strength of the abundance of their collection. These people mainly receive their mp3s from the internet instead of buying CDs, because they like to see their whole musiccollection at a glance.

=== Are there any known / often used Methods / Visualisation Techniques? ===

No, we don't know any similar visualization technique. ITunes only shows textbasedly how often a song was heard.

== Intended Purpose of our Visualization ==

=== What should be achieved with this visualization? ===
A better information representation of the MP3s should be achieved. Our visualization should help the user to get an overview of his collection and his listening-habits. He will get information about the tracks, in reference to a special artist, which he often listens to and to those which he has never heard before. The representation of the data should be expressive, precise and self-explanatory.

=== Which tasks should be solved? ===
By using this visualization technique, the user will get information about a chosen artist and his discography. For example: in his database the user has got the band "Radiohead", who produced albums over 15 years and in this period they changed their music-style from alternative rock to experimental electronic. The visualization will show him from which producing period he has got more MP3s and which period he likes more, by counting the number of listenings of each song. The result could be that he has got more MP3s from their early years, but likes the experimental electronic tracks more.

=== Questions that should be solved with this visualization technique ===

== Proposal of Design ==

=== Kind of Visualization / Visualization Details ===

When the user opens the "Interpret-Analyser" he will be prompted via a text-message in the main-window to click on an artist/band in the right upper window. The artists/bands are sorted alphabetically and the subject of interest can be found by scrolling the window vertically. If the user chose an artist/band in the upper right window, the main window will visualize him following details on the demanded item:

*The x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data
*The y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library

The respective maximum on the y-axis will give a first overview on how many songs the specific library contains per artist/band per year. Though the users' library might not be complete the visualization allows drawing conclusions according to the artist/band-activities over the last years. In any cases we assume that the user applies the "Interpret-Analyser" to artists/bands whereof he collected the whole discography and not only one song.

For each song (= one data point) one horizontal bar is drawn along the y-axis. That means for example if the library contains 34 songs by the band "Queen" with publication-year "1985", 34 bars are drawn at the x-axis value "1985" along the vertical y-axis.

In addition to that each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently).

Further on the user can interact and influence the characteristic of the visualization by using a slider positioned in the lower right in the graphic below. Via the slider a more objective image can be drawn according to the actual point of interest. This slider with a value-range from "0" to "10" represents the counts how often a song was played. It allows setting a threshold. The default value is "3" and means that songs that were played less than 3 times do not appear coloured, but as grey bars vertically above the coloured ones along the y-axis. If for example someone drags the slider to the position with value "10" and only 1 song out of 27 with a special publishing year was played more than 9 times the "Interpret-Analyser" shows 1 coloured and 26 grey bars at the according year. This could for instance help if someone is on the way to filter out his absolute favourites of an artist/band.

As it is mentioned above the "Interpret-Analyser" represents highly interesting visualizations for End-Users but it might also prevent outstanding features for the Music-Industry respectively bands, who work on a Come-Back. This however would assume to arising the data of a rather big audience, what could for example be achieved via a contest.

=== Visual Mapping ===

'''2D Diagram''':

*'''"X-AXIS":''' the x-axis shows the songs listed vertically by-publication-year generated out of the ID3-Data

*'''"Y-AXIS":''' the y-axis shows the number of songs published per year according to the specific number in the users' iTunes-library. The height of the vertical bar-column represents the "Song-Occurrence" per year

*'''"Colour":''' each bar drawn vertically has a specific colour, representing the date when he was last played. As it is shown in our Mock-Up below the range goes from blue (representing songs that have not been played for a long time) to red (representing songs that have been played recently). Grey bars represent songs that did not pass the adjusted threshold

=== Specification of used Techniques / applied Principles ===

*'''Bar X Plot:''' In this plot, one vertical bar is drawn for each data point [StatSoft, 2003]

*'''Histograms, 2D:''' 2D histograms present a graphical representation of the frequency distribution of the selected variable(s) in which the columns are drawn over the class intervals and the heights of the columns are proportional to the class frequencies. [StatSoft, 2003]

*'''Colour-Range, Linking & Brushing:''' A colour-Range representing the levels between "not played for a long time" and "recently played" (s. slide 100 of Info_Vis0.pdf handed out in the course 188.305 VO InfoVis)

*'''Scatterplot, 2D:''' The scatterplot visualizes a relation (correlation) between two variables X and Y (e.g., weight and height). Individual data points are represented in two-dimensional space (see below), where axes represent the variables (X on the horizontal axis and Y on the vertical axis). The two coordinates (X and Y) that determine the location of each point correspond to its specific values on the two variables. [StatSoft, 2003]

*'''Dynamic Queries:''' Adjusting the slider generates dynamic queries [StatSoft, 2003]

=== Possibilities of User-Interaction ===

*Select item of interest (artist/band)
**Get artist/band-details

*Adjust the slider to influence the threshold
**Get individual Visualizations according to the users' point of interest

=== Mockup / Fake Screenshot ===

[[Image:Interpret-Analyser.png]]

== References ==

[Wikipedia, 2005a] ID3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[Wikipedia, 2005b] MP3, Wikipedia, Last updated: 21 November, 2005, Retrieved at: November 22, 2005, http://en.wikipedia.org/wiki/Mp3

[Id3.org, 2004] ID3v2 frames, Id3.org, Last updated: 28. February, 2004, Retrieved at: November 22, 2005, http://www.id3.org/frames.html

[Montclaire, 2000] Data Types, Department of Science and Mathematics at Montclair State University, Last updated: 3. August, 2000, Retrieved at: November 22, 2005, http://www.csam.montclair.edu/~mcdougal/SCP/D_types.htm

[StatSoft, 2003] Graphical Analytic Techniques, Last updated: 2003, Retrieved at: November 24, 2005, http://www.statsoft.com/textbook/stgraph.html