2006-08-01: Job openings at IBM T, J. Watson: Research Staff Member

Research Staff Member (RSM) (JOB ID: B019410)

The Intelligent Multimedia Interaction (IMI) group at IBM T. J. Watson Research is seeking applicants for RSM positions. The IMI group conducts research exploring intelligent multimodal interfaces for information seeking and information analysis. We have developed a number of algorithms and technologies aimed at enabling a full-fledged, mixed-initiative multimodal/multimedia human-computer interaction for information search and analysis.

We are currently seeking qualified researchers whose research interests overlap with our efforts in the following areas:

Context-sensitive information retrieval and analysis: design and develop algorithms and software components that can help users to search and analyze information in context. We are particularly interested in candidates who have research experience in using user interaction context and data characteristics (e.g., meta data and data semantics) to help improve information retrieval (e.g., the ranking of retrieved documents), information integration (e.g., combining retrieved database data and text documents), and information presentation (e.g., visualizing relationships of retrieved data).

Visual analytics: design and develop interactive, visual illustrative and visual query metaphors and tools to aid in fast information exploration and information analysis. In this endeavor, we are exploring the use of context-aware visual interfaces that combine the human reasoning ability with the modern computer power to speed data exploration and knowledge discovery.

Adaptive multimedia information presentation and summary: design and develop algorithms and tools to compose effective visualization of heterogeneous information (e.g., structured data, text, and images) on the fly for dynamic information analysis. We are currently focusing on adaptive, user-system cooperative design of visual presentations. In this effort, both users and the system can recommend and compose suitable visualization formalisms, and the system can learn users? preferences over time.

Adaptive multimodal interpretation and recommendation: design and implement algorithms to robustly interpret a user?s multimodal requests specified in natural language, GUI, or gesture. We are exploring the use of machine learning approaches to learn a user?s new multimodal expressions on the fly. We also explore context-sensitive recommendation strategies to interactively help users to author complex requests (e.g., information analysis requests) and recover from interpretation problems.

We are especially seeking candidates with research interests in information retrieval, information visualization/graphics (e.g., non-photorealistic rendering for information depiction), natural language processing, machine learning, human-computer interaction, and intelligent user interfaces.

Our group is multi-disciplinary, with experts from fields ranging from intelligent user interaction and information visualization, to natural language processing and generation. The complimentary nature of our skills creates an exciting environment for research and exploration.

Qualifications

Candidates for this position must have an excellent research track record in one or more above areas. The candidates must also have extensive experience with designing, developing, and testing large object-oriented software systems (e.g., Java and C++) and programming in heterogeneous networking environment (e.g., Linux and Windows).

Skills

• 5+ year C/C++ experience
• 5+ years Java experience
• 5+ years programming experience with TCP/IP and thorough knowledge of communication protocols
• 1+ year web programming experience

System administration experience and skills strongly preferred

Educational level

Ph.D. in CS or relevant field required. Preference will be given to candidates who have experience in conducting or willing to engage in interdisciplinary and collaborative research.

For more information, or to apply, please contact us imi(at)us.ibm.com.

via Dbworld mailing list by Michelle Zhou