Joseph Konstan

Biography
Joseph A. Konstan is Associate Professor of Computer Science and Engineering at the University of Minnesota where he co-directs the GroupLens research group. His interests span a diverse set of topics within Human-Computer Interaction, including recommender systems, design issues in online communities, and the use of online interactions for health promotion. His current projects include investigations into the economics of participation in online communities, the technology and design factors for digital library recommenders, and on-line HIV-prevention studies.

Prof. Konstan earned his Ph.D. in 1993 from the University of California, Berkeley. He is an active volunteer in the professional community, currently serving as President of ACM SIGCHI and as vice-Chair of ACM’s Membership Services Board and ACM’s SIG Governing Board. He is an ACM Distinguished Lecturer and an IEEE Distinguished Visitor.

Position Paper
I am active in three projects directly related to social computing, and will summarize each briefly.

First, I am working in collaboration with an economist (Yan Chen, from the University of Michigan) and our students to explore the economics of participation in online communities. This particular series of studies (which is in its second year and will likely extend for another two or three years) has developed economic models of participation in recommender communities (in our case, the MovieLens system), considering the costs of different types of participation and the benefits that users get from their use of the system. We are currently designing experiments to test several theories of user behavior to understand what cues lead people to contribute effort to the welfare of others (e.g., contributing a review) vs. only toward their own welfare. As far as we know, this is the only study using techniques from experimental economics to explore user behavior in an online community. The first results from this effort will be published this Summer at the 10th International User Modeling conference.

Second, my students and I are working on the design of recommender tools to better support the use of digital libraries. While some of these tools are not outwardly social, others are quite explicitly so. Particular scenarios of interest include the creation and support of reading groups, and the collective creation and ownership of bibliographies. We first demonstrated the feasibility of recommending research papers in 2002, and since then have been expanding our efforts by studying the nature of recommendations desired and by developing the infrastructure for a large-scale experiment. We are now working with ACM’s Digital Library data to develop prototypes that can support a wide range of individual and social interactions that we hope will define the next generation of digital library.

Third, I have been working with colleagues from our medical school, and from across the university, on a project exploring ways to assess communitywide HIV risk online, to understand the relationship between online partner-seeking and unsafe sex, and to develop and test online interventions to improve sexual health and reduce sexual risk taking. This project has already yielded some interesting results, including demonstrations of the higher total risk exposure of US-resident Latino men through Internet-mediated liaisons compared with conventional ones. We are now entering the most exciting stage of the project. We are developing the online “equivalent” of a weekend-long group retreat that has been shown to be effective in increasing sexual health and reducing STD risk.

Back to Social Computing Symposium 2005