June 27, 2013: People of ACM: Joseph A. Konstan
Thursday, June 27, 2013
Joseph A. Konstan is Distinguished McKnight University Professor and Distinguished University Teaching Professor in the Department of Computer Science and Engineering at the University of Minnesota. His research is in the field of Human-Computer Interaction, with specific work in recommender systems, social computing, and application of computing to health challenges. His work on the GroupLens Recommender Systems was awarded the 2010 ACM Software Systems Award. Prof. Konstan received an A.B. in Computer Science from Harvard in 1987 and M.S. and Ph.D. degrees in Computer Science from the University of California, Berkeley in 1990 and 1993, respectively.
Dr. Konstan is a long-time ACM Volunteer. He is currently serving as the SGB Publications Advisor (which includes serving on the Publications Board and on the SGB Executive Committee), as chair of ACM's task force exploring collaboration with AMIA, as co-chair of an ACM- and NSF-funded workshop to explore how to better serve the diverse Social Computing research community (including exploring a new ACM journal in the area), and as Associate Editor of both ACM Transactions on Computer-Human Interaction and ACM Transactions on the Web. He is about to start a three year term as co-chair of the ACM Publications Board. Previously, he has served as Chair of the SIG Governing Board, as vice-Chair of the Membership Services Board, and previous terms on both Council and the ACM EC. He served as President of SIGCHI from 2003-2006, and as Editor of SIGCHI Bulletin from 1999-2003. He also chaired three ACM conferences — UIST 2003, RecSys 2007, and CHI 2012. He recently received the SIGCHI Lifetime Service Award.
Dr. Konstan has published over 100 peer-reviewed papers in journals and conferences, as well as numerous book chapters, six US patents, and a book. He is listed as the eighth-most cited researcher in human-computer interaction by Google Scholar. He has been elected a Fellow of the ACM, AAAS, and IEEE, and a member of the CHI Academy.
How does your MovieLens system, which collects data to determine individual movie preferences, differ from conventional data mining, which searches through large amounts of computerized data to find useful patterns or trends?
MovieLens is a recommender system based on automated collaborative filtering, which is a technology focused on making personal recommendations rather than of finding patterns or trends. Originally, MovieLens used an algorithm which identified for each user the MovieLens members whose tastes were most like theirs, and then combined the opinions of the "neighborhood" of users to form personal recommendations. Over time, we've adapted the system to use more sophisticated and efficient algorithms, but the core idea remains the same. Today, collaborative filtering algorithms sit alongside other data mining techniques as part of the toolkit a computer scientist can bring to a big data problem.
What motivated you to found and publish the Harvard Computer Review and serve as its president when you were an undergraduate?
There seemed to be both a need and an opportunity. This was in 1984-5, which was a very exciting time with the increasing prevalence of personal computers and the introduction of the original Macintosh. The Harvard Computer Society had regular gatherings of Apple, IBM PC, and other hardware-specific user groups — new product demonstrations would attract hundreds of students. And desktop publishing was just making this all possible. So, with the benefit of an early desktop publishing program (PageMaker 1.0), a phototypesetter we could borrow, and the ability to use the campus newspaper's printing facilities, I went to potential sponsors (folks selling computers, and also pizza and wings) and received enough advertising to put together a 48-page inaugural issue (later ones were smaller, but we made our mark and became a well-recognized campus publication).
Why did you pursue volunteer opportunities with ACM SIGCHI and the ACM SIG Governing Board, and what benefits have they provided as your career developed?
I've always enjoyed volunteering and helping make important things happen. Beyond that, I would volunteer and take on tasks people asked of me. That started with smaller tasks (running the CHI 96 exhibits program), grew larger (editing the SIGCHI Bulletin), and eventually put me in a position to serve as President of SIGCHI, and then as Chair of the SIG Governing Board and now as co-Chair of the ACM Publications Board. I really enjoy both the people I get to work with and the feeling that the work I do makes a difference.
As a computer science major, what advice would you give to students who are considering career opportunities in computer-human interaction?
Computer-human interaction is a broad field with lots of areas for computer scientists to specialize in. While you're still a student, it pays to explore as many of them as you can, from interaction technology to mobile interfaces to social computing and more. Focus on building a solid set of basic skills, because over the course of your career the interesting technology platforms will keep changing. And don't forget to learn a bunch about people (psychology, human factors, sociology), because while technology changed quickly, humans adapt much more slowly.