Automatic Web Tagging and Person Tagging Using Language Models

Abstract. Social bookmarking has become an important web2.0 application recently, which is concerned with the dual user behavior to search - tagging. Although social bookmarking websites, e.g., Del.icio.us, has been attracting much attentions, many research problems in social tagging has not been well addressed in literature. In this paper, we formally define the problem of social bookmark suggestion, and propose a probabilistic language modeling approach to automatically label the target web documents with meaningful phrases. The probabilistic language models trained from social tagging logs are used to automatically generate tags which capture the semantics of web documents. We also adapt the modeling approach to label internet users. Empirical experiments show that our approach is effective to suggest meaningful tags for web documents as well as web users. Categories and Subject Descriptors: H.3.3 [Information Search and Retrieval]: Text Mining General Terms: Algorithms

• Type: Conference
• Year: 2008
• Where: ADMA
• Authors: Qiaozhu Mei, Yi Zhang
• Keywords: ADMA 2008, Data Mining, Social Bookmarking, Social Tagging, Web Documents