InfoVis:Wiki - User contributions [en]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 3 - Prototype

2007-01-12T17:59:07Z

UE-InfoVis0607 0005416: finally added the images correctly

== Perspective Wall ==
=== Author ===
Matej Durco

=== Project description ===
We set out to implement a specific type of [[Teaching:TUW_-_UE_InfoVis_WS_2006/07_-_Gruppe_08_-_Aufgabe_1_-_Bifocal_Display|bifocal display]] called "Perspective Wall" [MayKinlay et al., 1991] using the comprehensive prefuse toolkit. Prefuse provides extensive methods for visualising data. While its main focus is implementing graphs it is also possible to implement time-based visualisation methods.

Our solution was originally meant to use a person's diary or, more specifically, records of other people the person met as its data source. The rows would represent individual people while the x-axis would represent time units as identified in the data. It should be possible to move backwards and forwards in time while at the same time zooming into a specific area of interest.

The actual implementation used the provided ''GridLayout'' class for positioning the individual data points. Prefuse's ''GridLayout'' ensures that one specific person will always be positioned in the same row.
For the desired interaction prefuse's ''BifocalDistortion'' class was used (with distortion set only to horizontal direction). It handles expanding and compressing the displayed data interactively.

First we tried to use ''LabelRenderer'' for the datapoints, but it turned out useless because with more data the labels were overlapping massively. Therefore we switched to ''ShapeRenderer'' - ie each datapoint is represented by a tiny square (or other shape) - and the desired information (which person is behind the datapoint) shows up as ''ToolTip'' on hover. The datapoints partly still overlap, but one can view the individual information.

=== Screenshot ===
[[Image:PerspectiveWall.png]]

=== Further Development ===
Existing solution is really very rudimentary. There would be a lot to do, to make it better:

==== Labeling ====
[[Image:DailyView.png|thumb|right|250px|Schematic illustration of the GridLayout for the focus-view]]

The solution with ''ToolTip'' is a rather unsatisfying as only information for one datapoint can be viewed at one time and even this access is rather slow (''ToolTip'' does not appear immediately). A natural alternative is employing AxisLabels. There is even a class - ''AxisLabelLayout'' - which seems suited for this. It would have to be extended to:
* allow more rows on the x-Axis (day and year in separate rows)
* position the y-Axis on the edge of the focus-view of the ''BifocalDistortion''

==== Linking ====
Allow direct access to the place in the original text (provided the data is extracted eg from a diary).
This could be implemented as an external link and text displayed in a '''browser''' '''or''' there could be another '''viewport integrated''' in the Visualisation.

==== Aggregation ====
At present only the basic data is displayed, ie every row is visualized as a datapoint. For larger datasets a feature to aggregate over time would be important. We propose an entity ''TimeUnit'' with its property ''aggregationInterval'' defined in '''number of days''' (ie a month is a ''TimeUnit'' with ''aggregationInterval=30''). Datapoints would represent the number of occurrences pro TimeUnit and could be displayed as vertically scaled rectangle.
There are 2 ways to arrange the datapoints vertically:
* leaving them aligned in a grid where every person has its own line (picture)
* ''StackedChart'' - where the scaled rectangles are scaled atop of each other.

''StackedChart'' has the positive side-effect of also visualizing the overall amount of occurrences, but the drawback of difficult labeling. Individual persons would have to be encoded in different colors and the color-assignement put in a legend which means additional lookup.

If we imagine the slice for one ''TimeUnit'' in the ''StackedChart'' very thin (think of integral calculus) we get actually a ''ThemeRiver'' which better expresses the continuity and also allows for almost direct positioning of labels (in or around the chart if necessary linked to the appropriate color-area by an arrow) but would be even more difficult to implement.

[[Image:AggrViewGridStack.png|frame|center|Schematic illustration of a GridLayout and a StackChartLayout for the context-view]]

==== Multiple Views ====
Integrating all the previously mentioned propositions we propose as an '''optimal solution a multiple-view''' visualization with:
# ''StackedChart'' or ''ThemeRiver'' for the aggregated context-view
# ''GridLayout'' for the focus-view
# browser for the source-text
where 1. and 2. could be either arranged
* in a sophisticated form of the ''BifocalDistortion'', which would provide '''different Layouts in context and focus''' or
* in '''independent viewports''' (which wouldn't make a PerspectiveWall anymore, but we are actually primarily not interested in PerspectiveWall, but in visualizing the data)
The focus-view would display one ''TimeUnit'' of the context-view by default (although this could be parametrized).

=== References ===
* [http://www.infovis-wiki.net/index.php/Teaching:TUW_-_UE_InfoVis_WS_2006/07_-_Gruppe_08_-_Aufgabe_1_-_Bifocal_Display Description of Bifocal Display in infovis-wiki]
*[Mackinlay et al., 1991] Mackinlay, J. D., Robertson, G. G., and Card, S. K.: The Perspective Wall: Detail and Context Smoothly Integrated. In Proceedings of ACM Conference on Human Factors in Computing Systems (CHI91), Information Visualization, pages 173–179, 1991. http://ieg.ifs.tuwien.ac.at/~aigner/teaching/infovis_ue/papers/p173-mackinlay_perspective_wall.pdf
*Prefuse Toolkit http://prefuse.org/

File:AggrViewGridStack.png

2007-01-12T17:57:04Z

UE-InfoVis0607 0005416: Schematic illustration of a GridLayout and a StackChartLayout with scaled datapoints employed in the aggregated context-view of the PerspectiveWall-Visualization

== Summary ==
Schematic illustration of a GridLayout and a StackChartLayout with scaled datapoints employed in the aggregated context-view of the PerspectiveWall-Visualization
== Copyright status ==
just use it
== Source ==
handmade

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 3 - Prototype

2007-01-12T17:35:52Z

UE-InfoVis0607 0005416: added internal link

== Perspective Wall ==
=== Author ===
Matej Durco

=== Project description ===
We set out to implement a specific type of [[Teaching:TUW_-_UE_InfoVis_WS_2006/07_-_Gruppe_08_-_Aufgabe_1_-_Bifocal_Display|bifocal display]] called "Perspective Wall" [MayKinlay et al., 1991] using the comprehensive prefuse toolkit. Prefuse provides extensive methods for visualising data. While its main focus is implementing graphs it is also possible to implement time-based visualisation methods.

Our solution was originally meant to use a person's diary or, more specifically, records of other people the person met as its data source. The rows would represent individual people while the x-axis would represent time units as identified in the data. It should be possible to move backwards and forwards in time while at the same time zooming into a specific area of interest.

The actual implementation used the provided ''GridLayout'' class for positioning the individual data points. Prefuse's ''GridLayout'' ensures that one specific person will always be positioned in the same row.
For the desired interaction prefuse's ''BifocalDistortion'' class was used (with distortion set only to horizontal direction). It handles expanding and compressing the displayed data interactively.

First we tried to use ''LabelRenderer'' for the datapoints, but it turned out useless because with more data the labels were overlapping massively. Therefore we switched to ''ShapeRenderer'' - ie each datapoint is represented by a tiny square (or other shape) - and the desired information (which person is behind the datapoint) shows up as ''ToolTip'' on hover. The datapoints partly still overlap, but one can view the individual information.

=== Screenshot ===
[[Image:PerspectiveWall.png]]

=== Further Development ===
Existing solution is really very rudimentary. There would be a lot to do, to make it better:

==== Labeling ====
The solution with ''ToolTip'' is a rather unsatisfying as only information for one datapoint can be viewed at one time and even this access is rather slow (''ToolTip'' does not appear immediately). A natural alternative is employing AxisLabels. There is even a class - ''AxisLabelLayout'' - which seems suited for this. It would have to be extended to:
* allow more rows on the x-Axis (day and year in separate rows)
* position the y-Axis on the edge of the focus-view of the ''BifocalDistortion''
[[Image:DailyView.png|frame|right||Schematic illustration of the GridLayout for the focus-view of the Perspective Wall]]

==== Linking ====
Allow direct access to the place in the original text (provided the data is extracted eg from a diary).
This could be implemented as an external link and text displayed in a '''browser''' '''or''' there could be another '''viewport integrated''' in the Visualisation.

==== Aggregation ====
At present only the basic data is displayed, ie every row is visualized as a datapoint. For larger datasets a feature to aggregate over time would be important. We propose an entity ''TimeUnit'' with its property ''aggregationInterval'' defined in '''number of days''' (ie a month is a ''TimeUnit'' with ''aggregationInterval=30''). Datapoints would represent the number of occurrences pro TimeUnit and could be displayed as vertically scaled rectangle.
There are 2 ways to arrange the datapoints vertically:
* leaving them aligned in a grid where every person has its own line (picture)
* ''StackedChart'' - where the scaled rectangles are scaled atop of each other.

''StackedChart'' has the positive side-effect of also visualizing the overall amount of occurrences, but the drawback of difficult labeling. Individual persons would have to be encoded in different colors and the color-assignement put in a legend which means additional lookup.

If we imagine the slice for one ''TimeUnit'' in the ''StackedChart'' very thin (think of integral calculus) we get actually a ''ThemeRiver'' which better expresses the continuity and also allows for almost direct positioning of labels (in or around the chart if necessary linked to the appropriate color-area by an arrow) but would be even more difficult to implement.
[[Image:AggrViewGrid_monochrom.png||Schematic illustration of the GridLayout for the context-view]]
[[Image:AggrViewStack_polychrom.png||Schematic illustration of the StackChartLayout for the context-view]]

==== Multiple Views ====
Integrating all the previously mentioned propositions we propose as an '''optimal solution a multiple-view''' visualization with:
# ''StackedChart'' or ''ThemeRiver'' for the aggregated context-view
# ''GridLayout'' for the focus-view
# browser for the source-text
where 1. and 2. could be either arranged
* in a sophisticated form of the ''BifocalDistortion'', which would provide '''different Layouts in context and focus''' or
* in '''independent viewports''' (which wouldn't make a PerspectiveWall anymore, but we are actually primarily not interested in PerspectiveWall, but in visualizing the data)
The focus-view would display one ''TimeUnit'' of the context-view by default (although this could be parametrized).

=== References ===
* [http://www.infovis-wiki.net/index.php/Teaching:TUW_-_UE_InfoVis_WS_2006/07_-_Gruppe_08_-_Aufgabe_1_-_Bifocal_Display Description of Bifocal Display in infovis-wiki]
*[Mackinlay et al., 1991] Mackinlay, J. D., Robertson, G. G., and Card, S. K.: The Perspective Wall: Detail and Context Smoothly Integrated. In Proceedings of ACM Conference on Human Factors in Computing Systems (CHI91), Information Visualization, pages 173–179, 1991. http://ieg.ifs.tuwien.ac.at/~aigner/teaching/infovis_ue/papers/p173-mackinlay_perspective_wall.pdf
*Prefuse Toolkit http://prefuse.org/

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 3 - Prototype

2007-01-12T17:35:08Z

UE-InfoVis0607 0005416: added internal link

== Perspective Wall ==
=== Author ===
Matej Durco

=== Project description ===
We set out to implement a specific type of [[Teaching:TUW_-_UE_InfoVis_WS_2006/07_-_Gruppe_08_-_Aufgabe_1_-_Bifocal_Display|bifocal display]] called "Perspective Wall" [MayKinlay et al., 1991] using the comprehensive prefuse toolkit. Prefuse provides extensive methods for visualising data. While its main focus is implementing graphs it is also possible to implement time-based visualisation methods.

Our solution was originally meant to use a person's diary or, more specifically, records of other people the person met as its data source. The rows would represent individual people while the x-axis would represent time units as identified in the data. It should be possible to move backwards and forwards in time while at the same time zooming into a specific area of interest.

The actual implementation used the provided ''GridLayout'' class for positioning the individual data points. Prefuse's ''GridLayout'' ensures that one specific person will always be positioned in the same row.
For the desired interaction prefuse's ''BifocalDistortion'' class was used (with distortion set only to horizontal direction). It handles expanding and compressing the displayed data interactively.

First we tried to use ''LabelRenderer'' for the datapoints, but it turned out useless because with more data the labels were overlapping massively. Therefore we switched to ''ShapeRenderer'' - ie each datapoint is represented by a tiny square (or other shape) - and the desired information (which person is behind the datapoint) shows up as ''ToolTip'' on hover. The datapoints partly still overlap, but one can view the individual information.

=== Screenshot ===
[[Image:PerspectiveWall.png]]

=== Further Development ===
Existing solution is really very rudimentary. There would be a lot to do, to make it better:

==== Labeling ====
The solution with ''ToolTip'' is a rather unsatisfying as only information for one datapoint can be viewed at one time and even this access is rather slow (''ToolTip'' does not appear immediately). A natural alternative is employing AxisLabels. There is even a class - ''AxisLabelLayout'' - which seems suited for this. It would have to be extended to:
* allow more rows on the x-Axis (day and year in separate rows)
* position the y-Axis on the edge of the focus-view of the ''BifocalDistortion''
[[Image:DailyView.png|frame|right||Schematic illustration of the GridLayout for the focus-view of the Perspective Wall]]

==== Linking ====
Allow direct access to the place in the original text (provided the data is extracted eg from a diary).
This could be implemented as an external link and text displayed in a '''browser''' '''or''' there could be another '''viewport integrated''' in the Visualisation.

==== Aggregation ====
At present only the basic data is displayed, ie every row is visualized as a datapoint. For larger datasets a feature to aggregate over time would be important. We propose an entity ''TimeUnit'' with its property ''aggregationInterval'' defined in '''number of days''' (ie a month is a ''TimeUnit'' with ''aggregationInterval=30''). Datapoints would represent the number of occurrences pro TimeUnit and could be displayed as vertically scaled rectangle.
There are 2 ways to arrange the datapoints vertically:
* leaving them aligned in a grid where every person has its own line (picture)
* ''StackedChart'' - where the scaled rectangles are scaled atop of each other.

''StackedChart'' has the positive side-effect of also visualizing the overall amount of occurrences, but the drawback of difficult labeling. Individual persons would have to be encoded in different colors and the color-assignement put in a legend which means additional lookup.

If we imagine the slice for one ''TimeUnit'' in the ''StackedChart'' very thin (think of integral calculus) we get actually a ''ThemeRiver'' which better expresses the continuity and also allows for almost direct positioning of labels (in or around the chart if necessary linked to the appropriate color-area by an arrow) but would be even more difficult to implement.
[[Image:AggrViewGrid_monochrom.png||Schematic illustration of the GridLayout for the context-view]]
[[Image:AggrViewStack_polychrom.png||Schematic illustration of the StackChartLayout for the context-view]]

==== Multiple Views ====
Integrating all the previously mentioned propositions we propose as an '''optimal solution a multiple-view''' visualization with:
# ''StackedChart'' or ''ThemeRiver'' for the aggregated context-view
# ''GridLayout'' for the focus-view
# browser for the source-text
where 1. and 2. could be either arranged
* in a sophisticated form of the ''BifocalDistortion'', which would provide '''different Layouts in context and focus''' or
* in '''independent viewports''' (which wouldn't make a PerspectiveWall anymore, but we are actually primarily not interested in PerspectiveWall, but in visualizing the data)
The focus-view would display one ''TimeUnit'' of the context-view by default (although this could be parametrized).

=== References ===
* [http://www.infovis-wiki.net/index.php/Teaching:TUW_-_UE_InfoVis_WS_2006/07_-_Gruppe_08_-_Aufgabe_1_-_Bifocal_Display|Description of Bifocal Display in infovis-wiki]
*[Mackinlay et al., 1991] Mackinlay, J. D., Robertson, G. G., and Card, S. K.: The Perspective Wall: Detail and Context Smoothly Integrated. In Proceedings of ACM Conference on Human Factors in Computing Systems (CHI91), Information Visualization, pages 173–179, 1991. http://ieg.ifs.tuwien.ac.at/~aigner/teaching/infovis_ue/papers/p173-mackinlay_perspective_wall.pdf
*Prefuse Toolkit http://prefuse.org/

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 3 - Prototype

2007-01-12T17:34:48Z

UE-InfoVis0607 0005416: added internal link

== Perspective Wall ==
=== Author ===
Matej Durco

=== Project description ===
We set out to implement a specific type of [[Teaching:TUW_-_UE_InfoVis_WS_2006/07_-_Gruppe_08_-_Aufgabe_1_-_Bifocal_Display|bifocal display]] called "Perspective Wall" [MayKinlay et al., 1991] using the comprehensive prefuse toolkit. Prefuse provides extensive methods for visualising data. While its main focus is implementing graphs it is also possible to implement time-based visualisation methods.

Our solution was originally meant to use a person's diary or, more specifically, records of other people the person met as its data source. The rows would represent individual people while the x-axis would represent time units as identified in the data. It should be possible to move backwards and forwards in time while at the same time zooming into a specific area of interest.

The actual implementation used the provided ''GridLayout'' class for positioning the individual data points. Prefuse's ''GridLayout'' ensures that one specific person will always be positioned in the same row.
For the desired interaction prefuse's ''BifocalDistortion'' class was used (with distortion set only to horizontal direction). It handles expanding and compressing the displayed data interactively.

First we tried to use ''LabelRenderer'' for the datapoints, but it turned out useless because with more data the labels were overlapping massively. Therefore we switched to ''ShapeRenderer'' - ie each datapoint is represented by a tiny square (or other shape) - and the desired information (which person is behind the datapoint) shows up as ''ToolTip'' on hover. The datapoints partly still overlap, but one can view the individual information.

=== Screenshot ===
[[Image:PerspectiveWall.png]]

=== Further Development ===
Existing solution is really very rudimentary. There would be a lot to do, to make it better:

==== Labeling ====
The solution with ''ToolTip'' is a rather unsatisfying as only information for one datapoint can be viewed at one time and even this access is rather slow (''ToolTip'' does not appear immediately). A natural alternative is employing AxisLabels. There is even a class - ''AxisLabelLayout'' - which seems suited for this. It would have to be extended to:
* allow more rows on the x-Axis (day and year in separate rows)
* position the y-Axis on the edge of the focus-view of the ''BifocalDistortion''
[[Image:DailyView.png|frame|right||Schematic illustration of the GridLayout for the focus-view of the Perspective Wall]]

==== Linking ====
Allow direct access to the place in the original text (provided the data is extracted eg from a diary).
This could be implemented as an external link and text displayed in a '''browser''' '''or''' there could be another '''viewport integrated''' in the Visualisation.

==== Aggregation ====
At present only the basic data is displayed, ie every row is visualized as a datapoint. For larger datasets a feature to aggregate over time would be important. We propose an entity ''TimeUnit'' with its property ''aggregationInterval'' defined in '''number of days''' (ie a month is a ''TimeUnit'' with ''aggregationInterval=30''). Datapoints would represent the number of occurrences pro TimeUnit and could be displayed as vertically scaled rectangle.
There are 2 ways to arrange the datapoints vertically:
* leaving them aligned in a grid where every person has its own line (picture)
* ''StackedChart'' - where the scaled rectangles are scaled atop of each other.

''StackedChart'' has the positive side-effect of also visualizing the overall amount of occurrences, but the drawback of difficult labeling. Individual persons would have to be encoded in different colors and the color-assignement put in a legend which means additional lookup.

If we imagine the slice for one ''TimeUnit'' in the ''StackedChart'' very thin (think of integral calculus) we get actually a ''ThemeRiver'' which better expresses the continuity and also allows for almost direct positioning of labels (in or around the chart if necessary linked to the appropriate color-area by an arrow) but would be even more difficult to implement.
[[Image:AggrViewGrid_monochrom.png||Schematic illustration of the GridLayout for the context-view]]
[[Image:AggrViewStack_polychrom.png||Schematic illustration of the StackChartLayout for the context-view]]

==== Multiple Views ====
Integrating all the previously mentioned propositions we propose as an '''optimal solution a multiple-view''' visualization with:
# ''StackedChart'' or ''ThemeRiver'' for the aggregated context-view
# ''GridLayout'' for the focus-view
# browser for the source-text
where 1. and 2. could be either arranged
* in a sophisticated form of the ''BifocalDistortion'', which would provide '''different Layouts in context and focus''' or
* in '''independent viewports''' (which wouldn't make a PerspectiveWall anymore, but we are actually primarily not interested in PerspectiveWall, but in visualizing the data)
The focus-view would display one ''TimeUnit'' of the context-view by default (although this could be parametrized).

=== References ===
* [[http://www.infovis-wiki.net/index.php/Teaching:TUW_-_UE_InfoVis_WS_2006/07_-_Gruppe_08_-_Aufgabe_1_-_Bifocal_Display||Description of Bifocal Display in infovis-wiki]]
*[Mackinlay et al., 1991] Mackinlay, J. D., Robertson, G. G., and Card, S. K.: The Perspective Wall: Detail and Context Smoothly Integrated. In Proceedings of ACM Conference on Human Factors in Computing Systems (CHI91), Information Visualization, pages 173–179, 1991. http://ieg.ifs.tuwien.ac.at/~aigner/teaching/infovis_ue/papers/p173-mackinlay_perspective_wall.pdf
*Prefuse Toolkit http://prefuse.org/

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 3 - Prototype

2007-01-12T17:32:39Z

UE-InfoVis0607 0005416:

== Perspective Wall ==
=== Author ===
Matej Durco

=== Project description ===
We set out to implement a specific type of [[Teaching:TUW_-_UE_InfoVis_WS_2006/07_-_Gruppe_08_-_Aufgabe_1_-_Bifocal_Display|bifocal display]] called "Perspective Wall" [MayKinlay et al., 1991] using the comprehensive prefuse toolkit. Prefuse provides extensive methods for visualising data. While its main focus is implementing graphs it is also possible to implement time-based visualisation methods.

Our solution was originally meant to use a person's diary or, more specifically, records of other people the person met as its data source. The rows would represent individual people while the x-axis would represent time units as identified in the data. It should be possible to move backwards and forwards in time while at the same time zooming into a specific area of interest.

The actual implementation used the provided ''GridLayout'' class for positioning the individual data points. Prefuse's ''GridLayout'' ensures that one specific person will always be positioned in the same row.
For the desired interaction prefuse's ''BifocalDistortion'' class was used (with distortion set only to horizontal direction). It handles expanding and compressing the displayed data interactively.

First we tried to use ''LabelRenderer'' for the datapoints, but it turned out useless because with more data the labels were overlapping massively. Therefore we switched to ''ShapeRenderer'' - ie each datapoint is represented by a tiny square (or other shape) - and the desired information (which person is behind the datapoint) shows up as ''ToolTip'' on hover. The datapoints partly still overlap, but one can view the individual information.

=== Screenshot ===
[[Image:PerspectiveWall.png]]

=== Further Development ===
Existing solution is really very rudimentary. There would be a lot to do, to make it better:

==== Labeling ====
The solution with ''ToolTip'' is a rather unsatisfying as only information for one datapoint can be viewed at one time and even this access is rather slow (''ToolTip'' does not appear immediately). A natural alternative is employing AxisLabels. There is even a class - ''AxisLabelLayout'' - which seems suited for this. It would have to be extended to:
* allow more rows on the x-Axis (day and year in separate rows)
* position the y-Axis on the edge of the focus-view of the ''BifocalDistortion''
[[Image:DailyView.png|frame|right||Schematic illustration of the GridLayout for the focus-view of the Perspective Wall]]

==== Linking ====
Allow direct access to the place in the original text (provided the data is extracted eg from a diary).
This could be implemented as an external link and text displayed in a '''browser''' '''or''' there could be another '''viewport integrated''' in the Visualisation.

==== Aggregation ====
At present only the basic data is displayed, ie every row is visualized as a datapoint. For larger datasets a feature to aggregate over time would be important. We propose an entity ''TimeUnit'' with its property ''aggregationInterval'' defined in '''number of days''' (ie a month is a ''TimeUnit'' with ''aggregationInterval=30''). Datapoints would represent the number of occurrences pro TimeUnit and could be displayed as vertically scaled rectangle.
There are 2 ways to arrange the datapoints vertically:
* leaving them aligned in a grid where every person has its own line (picture)
* ''StackedChart'' - where the scaled rectangles are scaled atop of each other.

''StackedChart'' has the positive side-effect of also visualizing the overall amount of occurrences, but the drawback of difficult labeling. Individual persons would have to be encoded in different colors and the color-assignement put in a legend which means additional lookup.

If we imagine the slice for one ''TimeUnit'' in the ''StackedChart'' very thin (think of integral calculus) we get actually a ''ThemeRiver'' which better expresses the continuity and also allows for almost direct positioning of labels (in or around the chart if necessary linked to the appropriate color-area by an arrow) but would be even more difficult to implement.
[[Image:AggrViewGrid_monochrom.png||Schematic illustration of the GridLayout for the context-view]]
[[Image:AggrViewStack_polychrom.png||Schematic illustration of the StackChartLayout for the context-view]]

==== Multiple Views ====
Integrating all the previously mentioned propositions we propose as an '''optimal solution a multiple-view''' visualization with:
# ''StackedChart'' or ''ThemeRiver'' for the aggregated context-view
# ''GridLayout'' for the focus-view
# browser for the source-text
where 1. and 2. could be either arranged
* in a sophisticated form of the ''BifocalDistortion'', which would provide '''different Layouts in context and focus''' or
* in '''independent viewports''' (which wouldn't make a PerspectiveWall anymore, but we are actually primarily not interested in PerspectiveWall, but in visualizing the data)
The focus-view would display one ''TimeUnit'' of the context-view by default (although this could be parametrized).

=== References ===
*[Mackinlay et al., 1991] Mackinlay, J. D., Robertson, G. G., and Card, S. K.: The Perspective Wall: Detail and Context Smoothly Integrated. In Proceedings of ACM Conference on Human Factors in Computing Systems (CHI91), Information Visualization, pages 173–179, 1991. http://ieg.ifs.tuwien.ac.at/~aigner/teaching/infovis_ue/papers/p173-mackinlay_perspective_wall.pdf
*Prefuse Toolkit http://prefuse.org/

File:AggrViewStack polychrom.png

2007-01-12T17:30:04Z

UE-InfoVis0607 0005416: Schematic illustration of a StackChartLayout with scaled datapoints employed in the aggregated context-view of the PerspectiveWall-Visualization

== Summary ==
Schematic illustration of a StackChartLayout with scaled datapoints employed in the aggregated context-view of the PerspectiveWall-Visualization
== Copyright status ==

== Source ==

File:AggrViewGrid monochrom.png

2007-01-12T17:29:39Z

UE-InfoVis0607 0005416: Schematic illustration of a GridLayout with scaled datapoints employed in the aggregated context-view of the PerspectiveWall-Visualization

== Summary ==
Schematic illustration of a GridLayout with scaled datapoints employed in the aggregated context-view of the PerspectiveWall-Visualization
== Copyright status ==
just use it
== Source ==
handmade

File:DailyView.png

2007-01-12T17:01:58Z

UE-InfoVis0607 0005416: Schematic illustration of a GridLayout employed in the focus-view of the PerspectiveWall-Visualization

== Summary ==
Schematic illustration of a GridLayout employed in the focus-view of the PerspectiveWall-Visualization
== Copyright status ==
just use it
== Source ==
handmade

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 3 - Prototype

2007-01-12T16:47:06Z

UE-InfoVis0607 0005416: previous post by me (forgot to login): Added section: Further Development

== Perspective Wall ==
=== Author ===
Matej Durco

=== Project description ===
We set out to implement a specific type of [[Teaching:TUW_-_UE_InfoVis_WS_2006/07_-_Gruppe_08_-_Aufgabe_1_-_Bifocal_Display|bifocal display]] called "Perspective Wall" [MayKinlay et al., 1991] using the comprehensive prefuse toolkit. Prefuse provides extensive methods for visualising data. While its main focus is implementing graphs it is also possible to implement time-based visualisation methods.

Our solution was originally meant to use a person's diary or, more specifically, records of other people the person met as its data source. The rows would represent individual people while the x-axis would represent time units as identified in the data. It should be possible to move backwards and forwards in time while at the same time zooming into a specific area of interest.

The actual implementation used the provided ''GridLayout'' class for positioning the individual data points. Prefuse's ''GridLayout'' ensures that one specific person will always be positioned in the same row.
For the desired interaction prefuse's ''BifocalDistortion'' class was used (with distortion set only to horizontal direction). It handles expanding and compressing the displayed data interactively.

First we tried to use ''LabelRenderer'' for the datapoints, but it turned out useless because with more data the labels were overlapping massively. Therefore we switched to ''ShapeRenderer'' - ie each datapoint is represented by a tiny square (or other shape) - and the desired information (which person is behind the datapoint) shows up as ''ToolTip'' on hover. The datapoints partly still overlap, but one can view the individual information.

=== Screenshot ===
[[Image:PerspectiveWall.png]]

=== Further Development ===
Existing solution is really very rudimentary. There would be a lot to do, to make it better:

==== Labeling ====
The solution with ''ToolTip'' is a rather unsatisfying as only information for one datapoint can be viewed at one time and even this access is rather slow (''ToolTip'' does not appear immediately). A natural alternative is employing AxisLabels. There is even a class - ''AxisLabelLayout'' - which seems suited for this. It would have to be extended to:
* allow more rows on the x-Axis (day and year in separate rows)
* position the y-Axis on the edge of the focus-view of the ''BifocalDistortion''

==== Linking ====
Allow direct access to the place in the original text (provided the data is extracted eg from a diary).
This could be implemented as an external link and text displayed in a '''browser''' '''or''' there could be another '''viewport integrated''' in the Visualisation.

==== Aggregation ====
At present only the basic data is displayed, ie every row is visualized as a datapoint. For larger datasets a feature to aggregate over time would be important. We propose an entity ''TimeUnit'' with its property ''aggregationInterval'' defined in '''number of days''' (ie a month is a ''TimeUnit'' with ''aggregationInterval=30''). Datapoints would represent the number of occurrences pro TimeUnit and could be displayed as vertically scaled rectangle.
There are 2 ways to arrange the datapoints vertically:
* leaving them aligned in a grid where every person has its own line (picture)
* ''StackedChart'' - where the scaled rectangles are scaled atop of each other.

''StackedChart'' has the positive side-effect of also visualizing the overall amount of occurrences, but the drawback of difficult labeling. Individual persons would have to be encoded in different colors and the color-assignement put in a legend which means additional lookup.

If we imagine the slice for one ''TimeUnit'' in the ''StackedChart'' very thin (think of integral calculus) we get actually a ''ThemeRiver'' which better expresses the continuity and also allows for almost direct positioning of labels (in or around the chart if necessary linked to the appropriate color-area by an arrow) but would be even more difficult to implement.

==== Multiple Views ====
Integrating all the previously mentioned propositions we propose as an '''optimal solution a multiple-view''' visualization with:
# ''StackedChart'' or ''ThemeRiver'' for the aggregated context-view
# ''GridLayout'' for the focus-view
# browser for the source-text
where 1. and 2. could be either arranged
* in a sophisticated form of the ''BifocalDistortion'', which would provide '''different Layouts in context and focus''' or
* in '''independent viewports''' (which wouldn't make a PerspectiveWall anymore, but we are actually primarily not interested in PerspectiveWall, but in visualizing the data)
The focus-view would display one ''TimeUnit'' of the context-view by default (although this could be parametrized).

=== References ===
*[Mackinlay et al., 1991] Mackinlay, J. D., Robertson, G. G., and Card, S. K.: The Perspective Wall: Detail and Context Smoothly Integrated. In Proceedings of ACM Conference on Human Factors in Computing Systems (CHI91), Information Visualization, pages 173–179, 1991. http://ieg.ifs.tuwien.ac.at/~aigner/teaching/infovis_ue/papers/p173-mackinlay_perspective_wall.pdf
*Prefuse Toolkit http://prefuse.org/

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 3

2006-12-01T18:57:16Z

UE-InfoVis0607 0005416: added image, highlighted important words

== Technique ==
We are interested in visualisation techniques that are especially well suited for data with an important time-dimension. Thus our choice would be Perspective Wall or Theme River.
We are not sure yet, which one is suited better, so we would like to regard both, as long as possible.
Perspective Wall is suited better for a detailed display of the data (with no aggregation),
whereas Theme River requires some kind of integration and would be better suited for an aggregated overview. Moreover the important elements of Perspective Wall (bifocal display with distortion on the off-center viewports) are better supported by prefuse.
[[Image:Perspective_wall_Inxight.jpg|frame|right||Sample illustration of the Perspective Wall visualisation technique]]

=== Data ===
''How can we describe the usage of a technique without defining the data we want to present?''

There are various data structures, for which this display technique could be suitable.
We want to focus on a '''Diary''': Person(s) meeting other persons and being at places.
There are two types of Diary: one for '''planning''', ie noting what one person is going to do,
the other one for '''protocolling''', ie what the person already did.

The data will most likely be available in tabular form with the three dimensions being '''people''', '''time''' and '''place'''. Prefuse provides extensive data input capabilities for tabular data that can be used.

=== Visual Mapping ===
Basic dimension is time. It is mapped to the horizontal axis. The selected/viewed '''time-slice''' is presented in the primary focus. The primary granularity of the ''time-slice'' is planned to be one month, but other levels of granularity would make sense as well.

Other dimensions are places and people.
As long as one person is examined, the ribbon could be sliced vertically,
according to places, where the person is, as one person can be only on one place at one time.

Other persons, ie persons the examined person meets or mentions, should be displayed in separate, parallel horizontal slices (in Perspective Wall).
This makes the gist of the data ('''who met where when?''') visible at one glance.

=== Interaction ===
* Scrolling the time axis (back and forth)
* Navigation to special time-slice
* Optionally zooming the time axis (from whole regarded data set to one day; this partly overlaps with the setting of the granularity of the time-slice, ie size of the main focus)
* Query on places and persons visualized either as restriction (filtering) or through highlighting.

== Prototype ==
''Mockup of possible Perspective Wall design to come''

File:Perspective wall Inxight.jpg

2006-12-01T18:48:27Z

UE-InfoVis0607 0005416: Sample depicting the bifocal visualisation technique "perspective wall"

== Summary ==
Sample depicting the bifocal visualisation technique "perspective wall"
== Copyright status ==

== Source ==
http://ieg.ifs.tuwien.ac.at/~aigner/teaching/ws06/infovis_ue/infovis_ue_aufgabe3-techniken.html

User:UE-InfoVis0607 0005416

2006-12-01T18:46:34Z

UE-InfoVis0607 0005416:

Matej Durco, 0005416, 534, xnrn at gmx dot net<br/>
member of
[[Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08|Gruppe 08 (Durco)]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 3

2006-11-27T13:33:12Z

UE-InfoVis0607 0005416:

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 3

2006-11-27T12:50:34Z

UE-InfoVis0607 0005416:

Teaching:TUW - UE InfoVis WS 2006/07 - Aufgabe 3 - Technikauswahl

2006-11-23T18:59:56Z

UE-InfoVis0607 0005416:

{|
|-
|'''Gruppe: '''
|'''Gewählte Technik:'''
|-
|01
|?
|-
|02
|?
|-
|03
|?
|-
|08
|3 Perspective Wall oder 6 ThemeRiver
|-
|}

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T22:34:44Z

UE-InfoVis0607 0005416: /* Variation */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* '''Chartjunk''': The columns are filled with distracting gradients.
* '''Data-Ink ratio''': Much of the graphic is filled with useless patterns. The actual (overly large) pie charts occupy only a fraction of the graphic.
* '''Data Density''': One pie chart per year seems a bit excessive for the amount of information they provide.
* '''Friendly Data Graphic''': This graphic relies heavily on colours, rendering it all but useless if converted to grayscale or viewed by colour-blind persons.
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* The pie charts are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* There are four graph labels in different fonts with redundant information.
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==
[[Image:LineChart.png]]

=== Improvements ===
* The market share changes are visible over the years.
* Removed confusing and unnecessary backgrounds.
* The data-ink ratio has been increased.
* The labels are clearly visible next to the lines.
* The difference between market shares is easy to determine. One can immediately see that Sotheby's is always ahead of Christie's by roughly the same margin.

=== Variation ===
[[Image:ColumnChart.png]]

This alternative design is useful for gauging how much of the total market can be attributed to each of the two companies. This representation shows the entirety of this particular market instead of just part of it, thus keeping the idea of the original pie charts. The drawback is that it is more difficult to see that Sotheby's is ahead of Christie's every single year by roughly the same margin. Also it violates against the Data-Ink rule.

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T15:01:09Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===

[[Image:LineChart.png|frame|left|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]
[[Image:ColumnChart.png|frame|left|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T15:00:59Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===

[[Image:LineChart.png|frame|left|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]
[[Image:ColumnChart.png|frame|right|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T15:00:14Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.png|frame|left|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.png|frame|left|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T14:59:52Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.png|frame|left|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.png|none|thumb|400px||Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T14:59:40Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.png|frame|left|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.png|none||||Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T14:59:36Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.png|frame|left|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.png|none|||Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T14:59:30Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.png|frame|left|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.png|none||Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T14:59:21Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.png|frame|left|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.png|none|thumb||Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T14:58:39Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.png|frame|left|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.png|frame|left|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T14:58:27Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.png|frame||Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.png|frame||Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T14:57:43Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.png|none||Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.png|none||Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T14:57:05Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.png|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.png|none|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T14:56:30Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.png|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.png|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T14:56:19Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* cakes are arranged diagonally without any meaning (except for the growing years)
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* 4 graph-labels in different font with redundant information
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.png|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.png|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T14:55:56Z

UE-InfoVis0607 0005416: /* Improved Graphic */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Criticism ===
* Gradient-fill of the year-columns - unnecessary ''Chartjunk''.
* Six cakes for one data-row - very low ''Data Density''.
* The individual pie charts are virtually indistinguishable, making pie charts a rather poor choice in this context.
* The pie charts are arranged diagonally without any meaning (except for the growing years).
* The position of the pie charts suggests an increase on the y-axis which doesn't have any function or labels.
* Four graph-labels in different font with redundant information.
* The labels are not anywhere near the pie charts and have to be found to derive any meaning from the charts.

== Improved Graphic ==

[[Image:LineChart.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 1]]

=== Improvements ===
* Market share changes are visible over the years.
* Removed confusing and unnecessary backgrounds.
* Increased data-ink ratio.
* Labels are clearly visible next to the lines.
* Difference between market shares easy to determine.

=== Variation ===
[[Image:ColumnChart.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis - Improved Graphic - Variant 2]]

File:ColumnChart.png

2006-11-13T14:50:53Z

UE-InfoVis0607 0005416: Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2 Improved Graphic, Variant 1 - Column Graph

== Summary ==
Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2 Improved Graphic, Variant 1 - Column Graph
== Copyright status ==

== Source ==

File:LineChart.png

2006-11-13T14:49:01Z

UE-InfoVis0607 0005416: /* Summary */

== Summary ==
Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2
Improved Graphic, Variant 1 - Column Graph

== Copyright status ==

== Source ==

File:LineChart.png

2006-11-13T14:48:50Z

UE-InfoVis0607 0005416: Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2 Improved Graphic, Variant 1 - Line Graph

== Summary ==
Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2
Improved Graphic, Variant 1 - Line Graph
== Copyright status ==

== Source ==

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-13T12:48:21Z

UE-InfoVis0607 0005416: /* Critique */

== Poor Graphic ==

[[Image:Friendly00sothebys.jpg|none|thumb|600px|none|Sotheby's / Christie's Worldwide Sales Market Share Analysis]]

=== Critique ===
* gradient-fill of the year-columns - unnecessary ''Chartjunk''
* six cakes for one data-row - very low ''Data Density''
* the cakes do not visualize clearly the change over the years.
* cakes are arranged diagonally without any meaning (except for the growing years)
* 4 graph-labels in different font with redundant information

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 2

2006-11-12T20:48:30Z

UE-InfoVis0607 0005416:

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 1 - Bifocal Display

2006-10-30T14:48:38Z

UE-InfoVis0607 0005416: /* Definition */

== Definition ==
'''Bifocal display''' is a visualisation technique, whose goal is to represent the current region of interest in a detailed view, maintaining the context visible in a schematic but recognisable way. [Apperley et al., 1982]

== Context ==
The desire to present all the relevant information within its context usually results in an extremely cluttered and confusing display. If only the current region of interest is displayed, one has to arrange for a mechanism to quickly switch to relevant information before and after the region of interest. Even so, much of the surrounding context is lost as it's not as easily accessible. [Spence, 2006]
This is known as the [[Focus-plus-Context|"focus+context"]] problem.

== Original concept ==
[[Image:Bifocal_display_01.gif|frame|right||Bifocal display visualised as a continuous ribbon]]To address this problem, [[Spence, Robert|Spence]] and [[Apperley, Mark|Apperley]] proposed their ''bifocal display'':

A bifocal display can be thought of as a '''continuous ribbon''' of data that is displayed on three seperate viewports or monitors (see image). The centre viewport is used to display the current region of interest while the left and right viewports are used to display the preceding and succeeding context respectively. The region of interest is displayed in great detail while the context in compressed, so as to allow one to identify objects at a cursory glance.

Most importantly, it should be easily possible to "scroll" backwards and forwards, replacing the current region of interest with content from its surrounding context, thus expanding the newly focused region of interest and relegating the former region of interest to its place in the compressed context. [Spence, 2006]

== Usage ==
[[Image:Fisheye_grid.gif|frame|left||Scheme of applying a fisheye view]]
The original concept is limited to visualization problems with one main continuous dimension.
If we add a second dimension we get the concept of fisheye or magnifying glass (see image).
Both are applicable only in situations with continuous dimensions.
This is already reflected in [Apperley et al., 1982]:

Instead of simply compressing the context visually it could prove useful to choose a different representation for the peripheral viewports than is used for the central viewport.

In many situations the context is best displayed by some sort of graph, the most common example being the filesystem tree.

[[Image:Chart timewall3.jpg|frame|right||TimeWall(TM) by Inxight Software, Inc.(TM) - Employment of the metaphor of continuous ribbon in a commercial application]]
Depending on how broad the metaphor of bifocal display is understood, one can see all types of applications with some sort of ''navigation'' and ''detail'' pane as applying the visualisation strategy of ''bifocal display''.

Following the various techniques are summarized:
* '''continuous ribbon''' - "the original" but only applicable for timelines or similar one-dimensional visualisation problems.
* '''fisheye view / magnyfing glass''' - very useful for viewing maps or plans
* '''tree + detail''' - probably the most widespread technique (email client, filesystem browser, etc)
* '''graph + detail''' - eg modeling tools which provide interactive viewports with zoomable views of a graph allowing to simply access the detailed information of its nodes or edges.

== References ==
*[Spence, 2006] Bob Spence, The Acquisition of Insight. Accessed at: Oct 29, 2006, http://www.ee.ic.ac.uk/research/information/www/Bobs.html#Context
*[Apperley et al., 1982] M.D. Apperley, I. Tzavaras, R. Spence. A bifocal display technique for data presentation. ''Eurographics '82'', p. 27. Abstract: http://diglib.eg.org/EG/DL/Conf/EG82/papers/EUROGRAPHICS_82pp027-043.pdf.abstract.pdf

'''Current Applications''' - bifocal display in practice:
*[Inxight, 2006] Inxight VizServer - corporate website, product section. Accessed at: Oct 30, 2006. http://www.inxight.com/products/vizserver/
*[Brain, 2006] TheBrainEKP: Enterprise Knowledge Platform - corporate website, product section. Accessed at: Oct 30, 2006. http://www.thebrain.com/Overview/default.html

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 1 - Bifocal Display

2006-10-30T14:48:01Z

UE-InfoVis0607 0005416: /* Definition */

== Definition ==
The goal of '''bifocal display''' is to represent the current region of interest in a detailed view, maintaining the context visible in a schematic but recognisable way. [Apperley et al., 1982]

== Context ==
The desire to present all the relevant information within its context usually results in an extremely cluttered and confusing display. If only the current region of interest is displayed, one has to arrange for a mechanism to quickly switch to relevant information before and after the region of interest. Even so, much of the surrounding context is lost as it's not as easily accessible. [Spence, 2006]
This is known as the [[Focus-plus-Context|"focus+context"]] problem.

== Original concept ==
[[Image:Bifocal_display_01.gif|frame|right||Bifocal display visualised as a continuous ribbon]]To address this problem, [[Spence, Robert|Spence]] and [[Apperley, Mark|Apperley]] proposed their ''bifocal display'':

A bifocal display can be thought of as a '''continuous ribbon''' of data that is displayed on three seperate viewports or monitors (see image). The centre viewport is used to display the current region of interest while the left and right viewports are used to display the preceding and succeeding context respectively. The region of interest is displayed in great detail while the context in compressed, so as to allow one to identify objects at a cursory glance.

Most importantly, it should be easily possible to "scroll" backwards and forwards, replacing the current region of interest with content from its surrounding context, thus expanding the newly focused region of interest and relegating the former region of interest to its place in the compressed context. [Spence, 2006]

== Usage ==
[[Image:Fisheye_grid.gif|frame|left||Scheme of applying a fisheye view]]
The original concept is limited to visualization problems with one main continuous dimension.
If we add a second dimension we get the concept of fisheye or magnifying glass (see image).
Both are applicable only in situations with continuous dimensions.
This is already reflected in [Apperley et al., 1982]:

Instead of simply compressing the context visually it could prove useful to choose a different representation for the peripheral viewports than is used for the central viewport.

In many situations the context is best displayed by some sort of graph, the most common example being the filesystem tree.

[[Image:Chart timewall3.jpg|frame|right||TimeWall(TM) by Inxight Software, Inc.(TM) - Employment of the metaphor of continuous ribbon in a commercial application]]
Depending on how broad the metaphor of bifocal display is understood, one can see all types of applications with some sort of ''navigation'' and ''detail'' pane as applying the visualisation strategy of ''bifocal display''.

Following the various techniques are summarized:
* '''continuous ribbon''' - "the original" but only applicable for timelines or similar one-dimensional visualisation problems.
* '''fisheye view / magnyfing glass''' - very useful for viewing maps or plans
* '''tree + detail''' - probably the most widespread technique (email client, filesystem browser, etc)
* '''graph + detail''' - eg modeling tools which provide interactive viewports with zoomable views of a graph allowing to simply access the detailed information of its nodes or edges.

== References ==
*[Spence, 2006] Bob Spence, The Acquisition of Insight. Accessed at: Oct 29, 2006, http://www.ee.ic.ac.uk/research/information/www/Bobs.html#Context
*[Apperley et al., 1982] M.D. Apperley, I. Tzavaras, R. Spence. A bifocal display technique for data presentation. ''Eurographics '82'', p. 27. Abstract: http://diglib.eg.org/EG/DL/Conf/EG82/papers/EUROGRAPHICS_82pp027-043.pdf.abstract.pdf

'''Current Applications''' - bifocal display in practice:
*[Inxight, 2006] Inxight VizServer - corporate website, product section. Accessed at: Oct 30, 2006. http://www.inxight.com/products/vizserver/
*[Brain, 2006] TheBrainEKP: Enterprise Knowledge Platform - corporate website, product section. Accessed at: Oct 30, 2006. http://www.thebrain.com/Overview/default.html

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 1 - Bifocal Display

2006-10-30T14:45:04Z

UE-InfoVis0607 0005416: /* Usage */

== Definition ==
A '''bifocal display''' provides a detailed view of the current region of interest while providing a more compressed view of information outside that region of interest.

== Context ==
The desire to present all the relevant information within its context usually results in an extremely cluttered and confusing display. If only the current region of interest is displayed, one has to arrange for a mechanism to quickly switch to relevant information before and after the region of interest. Even so, much of the surrounding context is lost as it's not as easily accessible. [Spence, 2006]
This is known as the [[Focus-plus-Context|"focus+context"]] problem.

== Original concept ==
[[Image:Bifocal_display_01.gif|frame|right||Bifocal display visualised as a continuous ribbon]]To address this problem, [[Spence, Robert|Spence]] and [[Apperley, Mark|Apperley]] proposed their ''bifocal display'':

A bifocal display can be thought of as a '''continuous ribbon''' of data that is displayed on three seperate viewports or monitors (see image). The centre viewport is used to display the current region of interest while the left and right viewports are used to display the preceding and succeeding context respectively. The region of interest is displayed in great detail while the context in compressed, so as to allow one to identify objects at a cursory glance.

Most importantly, it should be easily possible to "scroll" backwards and forwards, replacing the current region of interest with content from its surrounding context, thus expanding the newly focused region of interest and relegating the former region of interest to its place in the compressed context. [Spence, 2006]

== Usage ==
[[Image:Fisheye_grid.gif|frame|left||Scheme of applying a fisheye view]]
The original concept is limited to visualization problems with one main continuous dimension.
If we add a second dimension we get the concept of fisheye or magnifying glass (see image).
Both are applicable only in situations with continuous dimensions.
This is already reflected in [Apperley et al., 1982]:

Instead of simply compressing the context visually it could prove useful to choose a different representation for the peripheral viewports than is used for the central viewport.

In many situations the context is best displayed by some sort of graph, the most common example being the filesystem tree.

[[Image:Chart timewall3.jpg|frame|right||TimeWall(TM) by Inxight Software, Inc.(TM) - Employment of the metaphor of continuous ribbon in a commercial application]]
Depending on how broad the metaphor of bifocal display is understood, one can see all types of applications with some sort of ''navigation'' and ''detail'' pane as applying the visualisation strategy of ''bifocal display''.

Following the various techniques are summarized:
* '''continuous ribbon''' - "the original" but only applicable for timelines or similar one-dimensional visualisation problems.
* '''fisheye view / magnyfing glass''' - very useful for viewing maps or plans
* '''tree + detail''' - probably the most widespread technique (email client, filesystem browser, etc)
* '''graph + detail''' - eg modeling tools which provide interactive viewports with zoomable views of a graph allowing to simply access the detailed information of its nodes or edges.

== References ==
*[Spence, 2006] Bob Spence, The Acquisition of Insight. Accessed at: Oct 29, 2006, http://www.ee.ic.ac.uk/research/information/www/Bobs.html#Context
*[Apperley et al., 1982] M.D. Apperley, I. Tzavaras, R. Spence. A bifocal display technique for data presentation. ''Eurographics '82'', p. 27. Abstract: http://diglib.eg.org/EG/DL/Conf/EG82/papers/EUROGRAPHICS_82pp027-043.pdf.abstract.pdf

'''Current Applications''' - bifocal display in practice:
*[Inxight, 2006] Inxight VizServer - corporate website, product section. Accessed at: Oct 30, 2006. http://www.inxight.com/products/vizserver/
*[Brain, 2006] TheBrainEKP: Enterprise Knowledge Platform - corporate website, product section. Accessed at: Oct 30, 2006.
http://www.thebrain.com/Overview/default.html

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 1 - Bifocal Display

2006-10-30T14:43:28Z

UE-InfoVis0607 0005416: /* References */

== Definition ==
A '''bifocal display''' provides a detailed view of the current region of interest while providing a more compressed view of information outside that region of interest.

== Context ==
The desire to present all the relevant information within its context usually results in an extremely cluttered and confusing display. If only the current region of interest is displayed, one has to arrange for a mechanism to quickly switch to relevant information before and after the region of interest. Even so, much of the surrounding context is lost as it's not as easily accessible. [Spence, 2006]
This is known as the [[Focus-plus-Context|"focus+context"]] problem.

== Original concept ==
[[Image:Bifocal_display_01.gif|frame|right||Bifocal display visualised as a continuous ribbon]]To address this problem, [[Spence, Robert|Spence]] and [[Apperley, Mark|Apperley]] proposed their ''bifocal display'':

A bifocal display can be thought of as a '''continuous ribbon''' of data that is displayed on three seperate viewports or monitors (see image). The centre viewport is used to display the current region of interest while the left and right viewports are used to display the preceding and succeeding context respectively. The region of interest is displayed in great detail while the context in compressed, so as to allow one to identify objects at a cursory glance.

Most importantly, it should be easily possible to "scroll" backwards and forwards, replacing the current region of interest with content from its surrounding context, thus expanding the newly focused region of interest and relegating the former region of interest to its place in the compressed context. [Spence, 2006]

== Usage ==
[[Image:Fisheye_grid.gif|frame|left||Scheme of applying a fish-eye view]]
The original concept is limited to visualization problems with one main continuous dimension.
If we add a second dimension we get the concept of fish-eye or magnifying glass (see image).
Both are applicable only in situations with continuous dimensions.
This is already reflected in [Apperley et al., 1982]:

Instead of simply compressing the context visually it could prove useful to choose a different representation for the peripherial viewports than is used for the central viewport.

In many situations the context is best displayed by some sort of graph, the most common example being the filesystem-tree.

[[Image:Chart timewall3.jpg|frame|right||TimeWall(TM) by Inxight Software, Inc.(TM) - Employment of the metaphor of continuous ribbon in a commercial application]]
Depending on how broad the metapher of bifocal display is understood, one can see all types of applications with some sort of ''navigation'' and ''detail'' pane as applying the visualisation strategy of ''bifocal display''.

Following the various techniques are summarized:
* '''continuous ribbon''' - "the original" but only applicable for timelines or similar one-dimensional visualisation problems.
* '''fisheye view / magnyfing glass''' - very useful for viewing maps or plans
* '''tree + detail''' - probably the most widespread technique (email client, filesystem browser, etc)
* '''graph + detail''' - eg modeling tools which provide interactive viewports with zoomable views of a graph allowing to simply access the detailed information of its nodes or edges.

== References ==
*[Spence, 2006] Bob Spence, The Acquisition of Insight. Accessed at: Oct 29, 2006, http://www.ee.ic.ac.uk/research/information/www/Bobs.html#Context
*[Apperley et al., 1982] M.D. Apperley, I. Tzavaras, R. Spence. A bifocal display technique for data presentation. ''Eurographics '82'', p. 27. Abstract: http://diglib.eg.org/EG/DL/Conf/EG82/papers/EUROGRAPHICS_82pp027-043.pdf.abstract.pdf

'''Current Applications''' - bifocal display in practice:
*[Inxight, 2006] Inxight VizServer - corporate website, product section. Accessed at: Oct 30, 2006. http://www.inxight.com/products/vizserver/
*[Brain, 2006] TheBrainEKP: Enterprise Knowledge Platform - corporate website, product section. Accessed at: Oct 30, 2006.
http://www.thebrain.com/Overview/default.html

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 1 - Bifocal Display

2006-10-30T14:42:04Z

UE-InfoVis0607 0005416: /* Usage */

== Definition ==
A '''bifocal display''' provides a detailed view of the current region of interest while providing a more compressed view of information outside that region of interest.

== Context ==
The desire to present all the relevant information within its context usually results in an extremely cluttered and confusing display. If only the current region of interest is displayed, one has to arrange for a mechanism to quickly switch to relevant information before and after the region of interest. Even so, much of the surrounding context is lost as it's not as easily accessible. [Spence, 2006]
This is known as the [[Focus-plus-Context|"focus+context"]] problem.

== Original concept ==
[[Image:Bifocal_display_01.gif|frame|right||Bifocal display visualised as a continuous ribbon]]To address this problem, [[Spence, Robert|Spence]] and [[Apperley, Mark|Apperley]] proposed their ''bifocal display'':

A bifocal display can be thought of as a '''continuous ribbon''' of data that is displayed on three seperate viewports or monitors (see image). The centre viewport is used to display the current region of interest while the left and right viewports are used to display the preceding and succeeding context respectively. The region of interest is displayed in great detail while the context in compressed, so as to allow one to identify objects at a cursory glance.

Most importantly, it should be easily possible to "scroll" backwards and forwards, replacing the current region of interest with content from its surrounding context, thus expanding the newly focused region of interest and relegating the former region of interest to its place in the compressed context. [Spence, 2006]

== Usage ==
[[Image:Fisheye_grid.gif|frame|left||Scheme of applying a fish-eye view]]
The original concept is limited to visualization problems with one main continuous dimension.
If we add a second dimension we get the concept of fish-eye or magnifying glass (see image).
Both are applicable only in situations with continuous dimensions.
This is already reflected in [Apperley et al., 1982]:

Instead of simply compressing the context visually it could prove useful to choose a different representation for the peripherial viewports than is used for the central viewport.

In many situations the context is best displayed by some sort of graph, the most common example being the filesystem-tree.

[[Image:Chart timewall3.jpg|frame|right||TimeWall(TM) by Inxight Software, Inc.(TM) - Employment of the metaphor of continuous ribbon in a commercial application]]
Depending on how broad the metapher of bifocal display is understood, one can see all types of applications with some sort of ''navigation'' and ''detail'' pane as applying the visualisation strategy of ''bifocal display''.

Following the various techniques are summarized:
* '''continuous ribbon''' - "the original" but only applicable for timelines or similar one-dimensional visualisation problems.
* '''fisheye view / magnyfing glass''' - very useful for viewing maps or plans
* '''tree + detail''' - probably the most widespread technique (email client, filesystem browser, etc)
* '''graph + detail''' - eg modeling tools which provide interactive viewports with zoomable views of a graph allowing to simply access the detailed information of its nodes or edges.

== References ==
*[Spence, 2006] Bob Spence, The Acquisition of Insight. Accessed at: Oct 29, 2006, http://www.ee.ic.ac.uk/research/information/www/Bobs.html#Context
*[Apperley et al., 1982] M.D. Apperley, I. Tzavaras, R. Spence. ''A bifocal display technique for data presentation''. Eurographics '82, p. 27. Abstract: http://diglib.eg.org/EG/DL/Conf/EG82/papers/EUROGRAPHICS_82pp027-043.pdf.abstract.pdf

'''Current Applications''' - bifocal display in practice:
*[Inxight, 2006] Inxight VizServer - corporate website, product section. Accessed at: Oct 30, 2006. http://www.inxight.com/products/vizserver/
*[Brain, 2006] TheBrainEKP: Enterprise Knowledge Platform - corporate website, product section. Accessed at: Oct 30, 2006.
http://www.thebrain.com/Overview/default.html

File:Chart timewall3.jpg

2006-10-30T14:40:37Z

UE-InfoVis0607 0005416: /* Summary */

== Summary ==
TimeWall(TM) by Inxight Software, Inc.(TM) - Employment of the metaphor of continuous ribbon in a commercial application

== Copyright status ==

== Source ==
*[Inxight, 2006] Inxight VizServer - corporate web-site, product section. Accessed at: Oct 30, 2006. http://www.inxight.com/products/vizserver/

File:Bifocal display 01.gif

2006-10-30T14:39:14Z

UE-InfoVis0607 0005416: /* Source */

== Summary ==
Bifocal display visualised as a continous ribbon
== Copyright status ==

== Source ==
[Spence, 2006] Bob Spence, The Acquisition of Insight. Accessed at: Oct 29, 2006, http://www.ee.ic.ac.uk/research/information/www/Bobs.html#Context

File:Fisheye grid.gif

2006-10-30T14:39:04Z

UE-InfoVis0607 0005416: /* Source */

== Summary ==
scheme of applying the fish-eye-view
== Copyright status ==

== Source ==
[Spence, 2006] Bob Spence, The Acquisition of Insight. Accessed at: Oct 29, 2006, http://www.ee.ic.ac.uk/research/information/www/Bobs.html

File:Chart timewall3.jpg

2006-10-30T14:38:22Z

UE-InfoVis0607 0005416: /* Source */

== Summary ==
TimeWall(TM) by Inxight Software, Inc.(TM) - Employment of the metapher of continuous ribbon in a commercial application

== Copyright status ==

== Source ==
*[Inxight, 2006] Inxight VizServer - corporate web-site, product section. Accessed at: Oct 30, 2006. http://www.inxight.com/products/vizserver/

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 1 - Bifocal Display

2006-10-30T14:38:01Z

UE-InfoVis0607 0005416: /* References */

== Definition ==
A '''bifocal display''' provides a detailed view of the current region of interest while providing a more compressed view of information outside that region of interest.

== Context ==
The desire to present all the relevant information within its context usually results in an extremely cluttered and confusing display. If only the current region of interest is displayed, one has to arrange for a mechanism to quickly switch to relevant information before and after the region of interest. Even so, much of the surrounding context is lost as it's not as easily accessible. [Spence, 2006]
This is known as the [[Focus-plus-Context|"focus+context"]] problem.

== Original concept ==
[[Image:Bifocal_display_01.gif|frame|right||Bifocal display visualised as a continuous ribbon]]To address this problem, [[Spence, Robert|Spence]] and [[Apperley, Mark|Apperley]] proposed their ''bifocal display'':

A bifocal display can be thought of as a '''continuous ribbon''' of data that is displayed on three seperate viewports or monitors (see image). The centre viewport is used to display the current region of interest while the left and right viewports are used to display the preceding and succeeding context respectively. The region of interest is displayed in great detail while the context in compressed, so as to allow one to identify objects at a cursory glance.

Most importantly, it should be easily possible to "scroll" backwards and forwards, replacing the current region of interest with content from its surrounding context, thus expanding the newly focused region of interest and relegating the former region of interest to its place in the compressed context. [Spence, 2006]

== Usage ==
[[Image:Fisheye_grid.gif|frame|left||Scheme of applying a fish-eye view]]
The original concept is limited to visualization problems with one main continuous dimension.
If we add a second dimension we get the concept of fish-eye or magnyfing glass (see image).
Both are rather applicable only in situations with continuous dimensions.
This is already reflected in [Apperley et al., 1982]:

Instead of simply compressing the context visually it could prove useful to choose a different representation for the peripherial viewports than is used for the central viewport.

In many situations the context is best displayed by some sort of graph, the most common example being the filesystem-tree.

[[Image:Chart timewall3.jpg|frame|right||TimeWall(TM) by Inxight Software, Inc.(TM) - Employment of the metapher of continuous ribbon in a commercial application]]
Depending on how broad the metapher of bifocal display is understood, one can see all types of applications with some sort of ''navigation'' and ''detail'' pane as applying the visualisation strategy of ''bifocal display''.

Following the various techniques are summarized:
* '''continuous ribbon''' - "the original" but only applicable for time-lines or similar one-dimensional visualisation problems.
* '''fisheye view / magnyfing glass''' - very useful for viewing maps or plans
* '''tree + detail''' - probably the most wide-spread technique (email-client, filesystem-browser, etc)
* '''graph + detail''' - eg modeling tools which provide interactive viewports with zoomable views of a graph allowing to simply access the detailed information of its nodes or edges.

== References ==
*[Spence, 2006] Bob Spence, The Acquisition of Insight. Accessed at: Oct 29, 2006, http://www.ee.ic.ac.uk/research/information/www/Bobs.html#Context
*[Apperley et al., 1982] M.D. Apperley, I. Tzavaras, R. Spence. ''A bifocal display technique for data presentation''. Eurographics '82, p. 27. Abstract: http://diglib.eg.org/EG/DL/Conf/EG82/papers/EUROGRAPHICS_82pp027-043.pdf.abstract.pdf

'''Current Applications''' - bifocal display in practice:
*[Inxight, 2006] Inxight VizServer - corporate web-site, product section. Accessed at: Oct 30, 2006. http://www.inxight.com/products/vizserver/
*[Brain, 2006] TheBrainEKP: Enterprise Knowledge Platform - corporate web-site, product section. Accessed at: Oct 30, 2006.
http://www.thebrain.com/Overview/default.html

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 1 - Bifocal Display

2006-10-30T12:44:44Z

UE-InfoVis0607 0005416:

== Definition ==
A '''bifocal display''' provides a detailed view of the current region of interest while providing a more compressed view of information outside that region of interest.

== Context ==
The desire to present all the relevant information within its context usually results in an extremely cluttered and confusing display. If only the current region of interest is displayed, one has to arrange for a mechanism to quickly switch to relevant information before and after the region of interest. Even so, much of the surrounding context is lost as it's not as easily accessible. [Spence, 2006]
This is known as the [[Focus-plus-Context|"focus+context"]] problem.

== Original concept ==
[[Image:Bifocal_display_01.gif|frame|right||Bifocal display visualised as a continuous ribbon]]To address this problem, [[Spence, Robert|Spence]] and [[Apperley, Mark|Apperley]] proposed their ''bifocal display'':

A bifocal display can be thought of as a '''continuous ribbon''' of data that is displayed on three seperate viewports or monitors (see image). The centre viewport is used to display the current region of interest while the left and right viewports are used to display the preceding and succeeding context respectively. The region of interest is displayed in great detail while the context in compressed, so as to allow one to identify objects at a cursory glance.

Most importantly, it should be easily possible to "scroll" backwards and forwards, replacing the current region of interest with content from its surrounding context, thus expanding the newly focused region of interest and relegating the former region of interest to its place in the compressed context. [Spence, 2006]

== Usage ==
[[Image:Fisheye_grid.gif|frame|left||Scheme of applying a fish-eye view]]
The original concept is limited to visualization problems with one main continuous dimension.
If we add a second dimension we get the concept of fish-eye or magnyfing glass (see image).
Both are rather applicable only in situations with continuous dimensions.
This is already reflected in [Apperley et al., 1982]:

Instead of simply compressing the context visually it could prove useful to choose a different representation for the peripherial viewports than is used for the central viewport.

In many situations the context is best displayed by some sort of graph, the most common example being the filesystem-tree.

[[Image:Chart timewall3.jpg|frame|right||TimeWall(TM) by Inxight Software, Inc.(TM) - Employment of the metapher of continuous ribbon in a commercial application]]
Depending on how broad the metapher of bifocal display is understood, one can see all types of applications with some sort of ''navigation'' and ''detail'' pane as applying the visualisation strategy of ''bifocal display''.

Following the various techniques are summarized:
* '''continuous ribbon''' - "the original" but only applicable for time-lines or similar one-dimensional visualisation problems.
* '''fisheye view / magnyfing glass''' - very useful for viewing maps or plans
* '''tree + detail''' - probably the most wide-spread technique (email-client, filesystem-browser, etc)
* '''graph + detail''' - eg modeling tools which provide interactive viewports with zoomable views of a graph allowing to simply access the detailed information of its nodes or edges.

== References ==
*[Spence, 2006] Bob Spence, The Acquisition of Insight. Accessed at: Oct 29, 2006, http://www.ee.ic.ac.uk/research/information/www/Bobs.html#Context
*[Apperley et al., 1982] M.D. Apperley, I. Tzavaras, R. Spence. ''A bifocal display technique for data presentation''. Eurographics '82, p. 27. Abstract: http://diglib.eg.org/EG/DL/Conf/EG82/papers/EUROGRAPHICS_82pp027-043.pdf.abstract.pdf
Current Applications - bifocal display in practice/business:
*[InxightVizServer] Inxight VizServer - corporate web-site, product section. Accessed at: Oct 30, 2006. http://www.inxight.com/products/vizserver/
*[BrainEKP] TheBrainEKP: Enterprise Knowledge Platform - corporate web-site, product section. Accessed at: Oct 30, 2006.
http://www.thebrain.com/Overview/default.html

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 1 - Bifocal Display

2006-10-30T12:43:31Z

UE-InfoVis0607 0005416:

== Definition ==
A '''bifocal display''' provides a detailed view of the current region of interest while providing a more compressed view of information outside that region of interest.

== Context ==
The desire to present all the relevant information within its context usually results in an extremely cluttered and confusing display. If only the current region of interest is displayed, one has to arrange for a mechanism to quickly switch to relevant information before and after the region of interest. Even so, much of the surrounding context is lost as it's not as easily accessible. [Spence, 2006]
This is known as the [[Focus-plus-Context|"focus+context"]] problem.

== Original concept ==
[[Image:Bifocal_display_01.gif|frame|right||Bifocal display visualised as a continuous ribbon]]To address this problem, [[Spence, Robert|Spence]] and [[Apperley, Mark|Apperley]] proposed their ''bifocal display'':

A bifocal display can be thought of as a '''continuous ribbon''' of data that is displayed on three seperate viewports or monitors (see image). The centre viewport is used to display the current region of interest while the left and right viewports are used to display the preceding and succeeding context respectively. The region of interest is displayed in great detail while the context in compressed, so as to allow one to identify objects at a cursory glance.

Most importantly, it should be easily possible to "scroll" backwards and forwards, replacing the current region of interest with content from its surrounding context, thus expanding the newly focused region of interest and relegating the former region of interest to its place in the compressed context. [Spence, 2006]

== Usage ==
[[Image:Fisheye_grid.gif|frame|left||Scheme of applying a fish-eye view]]
The original concept is limited to visualization problems with one main continuous dimension.
If we add a second dimension we get the concept of fish-eye or magnyfing glass (see image).
Both are rather applicable only in situations with continuous dimensions.
This is already reflected in [Apperley et al., 1982]:

Instead of simply compressing the context visually it could prove useful to choose a different representation for the peripherial viewports than is used for the central viewport.

[[Image:Chart timewall3.jpg|frame|right||TimeWall(TM) by Inxight Software, Inc.(TM) - Employment of the metapher of continuous ribbon in a commercial application]]
In many situations the context is best displayed by some sort of graph, the most common example being the filesystem-tree.

Depending on how broad the metapher of bifocal display is understood, one can see all types of applications with some sort of ''navigation'' and ''detail'' pane as applying the visualisation strategy of ''bifocal display''.

Following the various techniques are summarized:
* '''continuous ribbon''' - "the original" but only applicable for time-lines or similar one-dimensional visualisation problems.
* '''fisheye view / magnyfing glass''' - very useful for viewing maps or plans
* '''tree + detail''' - probably the most wide-spread technique (email-client, filesystem-browser, etc)
* '''graph + detail''' - eg modeling tools which provide interactive viewports with zoomable views of a graph allowing to simply access the detailed information of its nodes or edges.

== References ==
*[Spence, 2006] Bob Spence, The Acquisition of Insight. Accessed at: Oct 29, 2006, http://www.ee.ic.ac.uk/research/information/www/Bobs.html#Context
*[Apperley et al., 1982] M.D. Apperley, I. Tzavaras, R. Spence. ''A bifocal display technique for data presentation''. Eurographics '82, p. 27. Abstract: http://diglib.eg.org/EG/DL/Conf/EG82/papers/EUROGRAPHICS_82pp027-043.pdf.abstract.pdf
Current Applications - bifocal display in practice/business:
*[InxightVizServer] Inxight VizServer - corporate web-site, product section. Accessed at: Oct 30, 2006. http://www.inxight.com/products/vizserver/
*[BrainEKP] TheBrainEKP: Enterprise Knowledge Platform - corporate web-site, product section. Accessed at: Oct 30, 2006.
http://www.thebrain.com/Overview/default.html

Teaching:TUW - UE InfoVis WS 2006/07 - Gruppe 08 - Aufgabe 1 - Bifocal Display

2006-10-30T12:41:23Z

UE-InfoVis0607 0005416: added Image:Fisheye_grid.gif

== Definition ==
A '''bifocal display''' provides a detailed view of the current region of interest while providing a more compressed view of information outside that region of interest.

== Context ==
The desire to present all the relevant information within its context usually results in an extremely cluttered and confusing display. If only the current region of interest is displayed, one has to arrange for a mechanism to quickly switch to relevant information before and after the region of interest. Even so, much of the surrounding context is lost as it's not as easily accessible. [Spence, 2006]
This is known as the [[Focus-plus-Context|"focus+context"]] problem.

== Original concept ==
[[Image:Bifocal_display_01.gif|frame|right||Bifocal display visualised as a continuous ribbon]]To address this problem, [[Spence, Robert|Spence]] and [[Apperley, Mark|Apperley]] proposed their ''bifocal display'':

A bifocal display can be thought of as a '''continuous ribbon''' of data that is displayed on three seperate viewports or monitors (see image). The centre viewport is used to display the current region of interest while the left and right viewports are used to display the preceding and succeeding context respectively. The region of interest is displayed in great detail while the context in compressed, so as to allow one to identify objects at a cursory glance.

Most importantly, it should be easily possible to "scroll" backwards and forwards, replacing the current region of interest with content from its surrounding context, thus expanding the newly focused region of interest and relegating the former region of interest to its place in the compressed context. [Spence, 2006]

== Usage ==
[[Image:Fisheye_grid.gif|frame|right||Scheme of applying a fish-eye view]]
The original concept is limited to visualization problems with one main continuous dimension.
If we add a second dimension we get the concept of fish-eye or magnyfing glass (see image).
Both are rather applicable only in situations with continuous dimensions.
This is already reflected in [Apperley et al., 1982]:

Instead of simply compressing the context visually it could prove useful to choose a different representation for the peripherial viewports than is used for the central viewport.

In many situations the context is best displayed by some sort of graph, the most common example being the filesystem-tree.

Depending on how broad the metapher of bifocal display is understood, one can see all types of applications with some sort of ''navigation'' and ''detail'' pane as applying the visualisation strategy of ''bifocal display''.

Following the various techniques are summarized:
* '''continuous ribbon''' - "the original" but only applicable for time-lines or similar one-dimensional visualisation problems.
* '''fisheye view / magnyfing glass''' - very useful for viewing maps or plans
* '''tree + detail''' - probably the most wide-spread technique (email-client, filesystem-browser, etc)
* '''graph + detail''' - eg modeling tools which provide interactive viewports with zoomable views of a graph allowing to simply access the detailed information of its nodes or edges.

=== Current Examples ===
[[Image:Chart timewall3.jpg|frame||TimeWall(TM) by Inxight Software, Inc.(TM) - Employment of the metapher of continuous ribbon in a commercial application]]

== References ==
*[Spence, 2006] Bob Spence, The Acquisition of Insight. Accessed at: Oct 29, 2006, http://www.ee.ic.ac.uk/research/information/www/Bobs.html#Context
*[Apperley et al., 1982] M.D. Apperley, I. Tzavaras, R. Spence. ''A bifocal display technique for data presentation''. Eurographics '82, p. 27. Abstract: http://diglib.eg.org/EG/DL/Conf/EG82/papers/EUROGRAPHICS_82pp027-043.pdf.abstract.pdf
Current Applications - bifocal display in practice/business:
*[InxightVizServer] Inxight VizServer - corporate web-site, product section. Accessed at: Oct 30, 2006. http://www.inxight.com/products/vizserver/
*[BrainEKP] TheBrainEKP: Enterprise Knowledge Platform - corporate web-site, product section. Accessed at: Oct 30, 2006.
http://www.thebrain.com/Overview/default.html