作者: Junpeng Chen , Juan Liu , Bo Guo
DOI: 10.1007/978-3-642-34500-5_46
关键词:
摘要: Social tagging system, also noted as folksonomies, is an important way for users to describe resources on the Web. Because of continually changing and informal definition, semantics these social tags are ambiguous hard adopt web applications. In this paper, we propose a method attach semantic topic sense tags. The non-negative matrix factorization (NMF) performed find hidden topics in folksonomy. A novel automatic evaluation proposed measure our approach. Our shows that induction folksonomy allows precise complete search, which one key functionalities systems.