Research

Research Groups > Human-Computer Interaction

Human-Computer Interaction

Northeastern's human-computer interaction faculty members are examining the relationship between humans and computers in novel and groundbreaking ways. "The current term being used to describe what a lot of us are doing is human-centered computing," says Harriet Fell. "We're interested in how technology impacts people."

Fell, with a grant from the National Institutes of Health (NIH), works with Joel MacAuslan of Speech Technology and Applied Research and graduate student Jun Gong on the visiBabble Project. This system's underlying software can extract features from infants' babbles. Fell and MacAuslan have developed applications that use these features to provide diagnostic information indicating whether or not a child may be at risk for speech-related problems.

Timothy Bickmore is developing a virtual nurse to educate hospital patients about their health condition and post-discharge self-care. It also aims to help people change their health behavior by encouraging them to exercise more and take their medications as prescribed. With the help of undergraduate and graduate students, and a grant from the NIH, Bickmore has developed animated conversational agents that simulate face-to-face conversations between patients and health professionals.

Peter Tarasewich, with Carole Hafner, is examining how people split their attention among multiple tasks in the mobile environment. In conducting this research, they apply models of human attention to the design of mobile and ubiquitous computing systems. This work seeks to understand the human and machine complexities in designing mobile device interfaces for environments where multiple tasks must be attended to and distractions are frequent (such as driving a car while using an electronic navigation system), with a goal of creating interfaces that are safe and effective.

PhD Student Highlights

• JUN GONG is completing his doctoral thesis with Peter Tarasewich on predictive text entry methods for mobile devices. He has also been active in developing visiBabble with Harriet Fell. He is co-author of nine conference papers, including a full paper at the Conference on Human Factors in Computing Systems on work pertaining to his thesis.

• DANIEL MAUER and his adviser, Timothy Bickmore, have been developing a portable, animated, PDA-based exercise coach in a project funded by the National Library of Medicine. The coach senses user exercise behavior and responds in real time, using an integrated accelerometer.

• DANIEL SCHULMAN and his adviser, Timothy Bickmore, have been developing an automated "virtual nurse" that will educate patients prior to hospital discharge, under a grant from the National Heart, Lung and Blood Institute. Schulman has developed a visual design tool that enables user-agent dialogs to be rapidly constructed and tested.