Investigating Behavioral Variability in Web Search
Ryen White (Microsoft Research)
Steven M. Drucker (Microsoft LiveLabs)
In this paper we describe a longitudinal log-based study that investigated variability in peoples' interaction behavior when engaged in search-related activities on the World Wide Web. We analyze the consistency of interaction patterns for more than two thousand volunteer users over a period of five months. Findings of our analysis suggest that there are dramatic differences in variability in key aspects of the interaction within and between queries, and within and between users. Our findings also indicate the existence of at least two distinct classes of user -- navigators and explorers -- who exhibit large differences in their search behavior. These findings have implications for the design of tools to support more effective search-related interactions on the Web.
New Brunswick, Wednesday, May 9, 2007, 1:30pm to 3:00pm.