Steven Ross

Research Interests:

• Conversational Interaction
• Cognitive Visualization
• Collaborative Reasoning and Decision-Making
• Knowledge Representation
• Automated Reasoning
• Natural Language Processing
• Collaborative Development Environments
• Speech Recognition

Steve is a member of the IBM Research Visual AI Lab. In recent years, has focused on applying generative AI to develop conversational assistants for visualization and software development. Before that his work in the lab focused primarily on visualization and exploration of high dimensional data and on navigation of content repositories.   He was lead developer of the Watson News Explorer and Watson Discovery Visual Insights.   

Prior to joining the Visual AI Lab, he was working in the area of Collaborative Decision Making, attempting to foster collective cognition processes that help groups of people to think together more effectively. This is an outgrowth of his work on Collaborative Reasoning, which aimed to help individuals to discover and benefit from existing knowledge within their organizations in order to avoid duplication of effort and to permit better decisions to be made in a more timely fashion.

In the past, Steve worked on the ActivitySpaces project, developing an Activity Based Collaborative Development Environment, and on the Jazz project, which brought workgroup computing tools to bear on the software development process.

Steve was the chief architect of the Cambridge Speech Initiative, an effort that used speech recognition and synthesis technology to develop a conversational user interface.

Steve joined Research in June of 2000. Prior to that, Steve worked for many years as an architect on the 1-2-3 spreadsheet, where he was responsible for the development of much of the spreadsheet user interface. Steve also led the effort to add speech capabilities to 1-2-3, for which he received the Lotus Fellows Award for Technical Innovation in 1998. Steve joined Lotus in 1991 after several years as a contractor, during which he was responsible for the development of the LPL (predecessor to LotusScript) compiler, debugger, and interpreter.

Prior to joining Lotus, Steve was a founder of Reasonix, Inc., where he worked on a team that developed a highly optimizing Fortran compiler for an array processor architecture, and a mixed initiative expert system. Before Reasonix, Steve was a software engineer at Verbex, where he participated in the development of the first commercial continuous speech recognition product.

Steve has SM and SB degrees in Computer Science from MIT.