Topical Relevance and Information Quality in Cognitive Models of Web Search Behavior: Introducing Epistemic Scent into Information Foraging Theory


Current cognitive models of Web navigation (e.g., Information Foraging Theory, IFT, Pirolli, 2007) are based on the assumption that users’ behavior is guided by evaluating the topical relevance of information encountered on the Web. This “information scent” has been successfully used to model Web search behavior. In this paper, however, we claim that topicality-oriented theories like IFT need to additionally consider the evaluation of information quality in order to address a broader class of realistic search tasks. For instance, when search tasks are complex and the quality of available Web information is highly variable, Web navigation will also depend on evaluating information quality, in addition to evaluating topical relevance. In this paper we first provide a theoretical framework of quality evaluation during Web search. Second, we review two experimental studies to substantiate this theoretical framework. Finally, we propose an extension of IFT using the concept of epistemic scent to incorporate evaluations of quality into the theory.

Back to Table of Contents