Colloquia / Seminars / Industry Talks

As online stores offer practically an infinite shelf space, recommender systems are playing an increasingly important role in helping users search and discover items that they may want to buy. In this talk, I first start with a brief survey of the rating based social recommender systems and their applications in online industry. I will then spend some time discussing some of the unsolved issues, especially concerning user adoption problems such as the cold start phenomena, users’ acceptance of recommendations, and personalization. The main part of the talk focuses on the technology behind critiquing based recommender (CBR) systems. Even though they may not address all of the user issues, CBR systems offer some effective solutions. They do not require users to leave traces of their interests via behavioral patterns. Instead, they encourage users to express them via the interface. Moreover, since users are completely involved in the preference elicitation process in such systems, users feel more in control of the recommendation process, and as a consequence they are more convinced of the products recommended to them. I will finish the talk by explaining the baggage carousel phenomenon and show you how critiquing based recommender systems enable users find personalized items without expending extra interaction effort. Through the analysis of some of our empirical studies, I hope to reveal to you some insights on the effective design of recommender systems for scalable user adoption.

Dr. Pearl Pu is the director of the Human Computer Interaction Group in the School of Computer and Communication Sciences at the Swiss Federal Institute of Technology in Lausanne (EPFL). Her research interests include decision support, electronic commerce, online consumer decision behavior, product recommender systems, travel planning tools, trust-inspiring interfaces for recommender agent, music recommenders, scalable user experience, and social navigation. She has been recently elected as the general chair for the ACM international conference on Recommender Systems, a field that provides the key technology to most e-commerce sites for up-sale, cross-sale and other revenue increasing strategies. See http://recsys.acm.org/.

A native from Shanghai, China, she moved to the United States shortly after passing the entrance examination to ZheJiang University. She obtained her Master and Ph.D. degrees from University of Pennsylvania in artificial intelligence and computer graphics. She was a visiting scholar at Stanford University in 2001, both in the database and HCI groups. While there, she gave seminars at Xerox PARC's weekly seminar series and Stanford's HCI Design Studio class as a guest lecturer.

She was also co-founder and chairwoman of Iconomic Systems (1997-2001), and invented the any-criteria preference-based search method for travel solutions. The company was successfully sold to i:FAO, Germany.

Her recent publications included papers from Electronic Commerce Research Journal, Journal of Artificial Intelligence Research, Knowledge based Systems, AAAI, ACM ECommerce, AI Communications, International Conferences on Intelligent User Interfaces, ACM CHI, ACM EC, AH, IEEE InfoVis, AVI and the Constraints journal's special issue on Constraint Agents. She serves as committee members for numerous conferences. Details at http://hci.epfl.ch/members/pearl/index.html.