Users Tagging Behavior and the Effect of Recommendation

Abstract

Users' tagging behavior has evolved. Formerly, the majority of users employed tags to label explicit content presented in resources. Now many users assign tags to express more than mere content description. The primary goal of this paper is to investigate how recommender systems change users' tagging behavior. We focused our investigation in the tagging task, modeling tags from a semiotic point of view. The results of a user experiment we conducted show that recommender systems can drive users to change the tag structure adopted.