Hyponymy Extraction and Web Search Behavior Analysis Based On Query Reformulation

Authors

Abstract

A web search engine log is a very rich source of semantic knowledge. In this paper we focus on the extraction of hyponymy relations from individual user sessions by examining, search behavior. The results obtained allow us to identify specific reformulation models as ones that more frequently represent hyponymy relations. The extracted relations reflect the knowledge that the user is employing while searching the web. Simultaneously, this study leads to a better understanding of web user search behavior.

Keywords

Semantic Web Usage Mining, Query Reformulation, Hyponymy Extraction, Web User Search Behavior

Subject

Web Intelligence

Conference

11th edition of the Ibero-American Conference on Artificial Intelligence, October 2008