Segment-level display time as implicit feedback: a comparison to eye tracking

Georg Buscher, Ludger van Elst, Andreas Dengel

In: SIGIR '09: Proceedings of the 32nd international ACM SIGIR conference on Research and development in information retrieval. ACM SIGIR Conference on Information Retrieval (SIGIR-09) 32nd July 19-23 Boston MA United States Pages 67-74 ISBN 978-1-60558-483-6 ACM 2009.


We examine two basic sources for implicit relevance feed- back on the segment level for search personalization: eye tracking and display time. A controlled study has been conducted where 32 participants had to view documents in front of an eye tracker, query a search engine, and give explicit relevance ratings for the results. We examined the performance of the basic implicit feedback methods with respect to improved ranking and compared their performance to a pseudo relevance feedback baseline on the segment level and the original ranking of a Web search engine. Our results show that feedback based on display time on the segment level is much coarser than feedback from eye tracking. But surprisingly, for re-ranking and query expansion it did work as well as eye-tracking-based feedback. All behavior-based methods performed significantly better than our non-behavior-based baseline and especially improved poor initial rankings of the Web search engine. The study shows that segment-level display time yields comparable results as eye-tracking-based feedback. Thus, it should be considered in future personalization systems as an inexpensive but precise method for implicit feedback.


German Research Center for Artificial Intelligence
Deutsches Forschungszentrum für Künstliche Intelligenz