Measuring and Predicting Search Engine Users' Satisfaction

Ovidiu Dan and Brian D. Davison
Official ACM Digital Library page for article

Free Authorized ACM version Full article (35 pages, available free)

Search satisfaction is defined as the fulfillment of a user's information need. Characterizing and predicting the satisfaction of search engine users is vital for improving ranking models, increasing user retention rates, and growing market share. This article provides an overview of the research areas related to user satisfaction. First, we show that whenever users choose to defect from one search engine to another they do so mostly due to dissatisfaction with the search results. We also describe several search engine switching prediction methods, which could help search engines retain more users. Second, we discuss research on the difference between good and bad abandonment, which shows that in approximately 30% of all abandoned searches the users are in fact satisfied with the results. Third, we catalog techniques to determine queries and groups of queries that are underperforming in terms of user satisfaction. This can help improve search engines by developing specialized rankers for these query patterns. Fourth, we detail how task difficulty affects user behavior and how task difficulty can be predicted. Fifth, we characterize satisfaction and we compare major satisfaction prediction algorithms.

ACM Computing Surveys, Volume 49, Issue 1, Article 18, 35 pages. July 2016.

Back to Brian Davison's publications

Last modified: 9 December 2020
Brian D. Davison