Implicit Feedback in Social Filetering

Doug Oard
University of Maryland College Park

2:00pm Friday October 15 , 1999
Lecture Hall V, ECS

Collaborative filtering offers the potential to augment content-based techniques by incorporating evidence regarding factors such as authority and accuracy into the feature set that describes each document. Present experimental collaborative filtering systems typically depend upon explicit user-assigned ratings, an approach that does not scale easily to large-scale applications. We have been exploring an alternative approach - inference of "implicit" ratings that are derived from observing user behavior. In this talk I will present a framework for thinking about implicit ratings, explain what is known already, and describe the results of an experiment we have recently completed in which we explored the interaction between reading time and retention behavior. The talk will conclude with a summary of some important open issues that merit further work.

Douglas Oard is an Assistant Professor in the College of Library and Information Services at the University of Maryland. His research interests center around the use of emerging technologies to support information seeking by end users, with present projects investigating audio retrieval, cross-language text retrieval, and the exchange of ratings by networked users. Additional information is available at http://www.glue.umd.edu/~oard/.