作者: Fabian Abel , Qi Gao , Geert-Jan Houben , Ke Tao
DOI: 10.1007/978-3-642-21064-8_26
关键词:
摘要: As the most popular microblogging platform, vast amount of content on Twitter is constantly growing so that retrieval relevant information (streams) becoming more and difficult every day. Representing semantics individual activities modeling interests users would allow for personalization therewith countervail overload. Given variety recency topics people discuss Twitter, semantic user profiles generated from posts moreover promise to be beneficial other applications Social Web as well. However, automatically inferring meaning a non-trivial problem. In this paper we investigate based posts. We introduce analyze methods linking with related news articles in order contextualize activities. then propose compare strategies exploit extracted both tweets represent semantically meaningful way. A large-scale evaluation validates benefits our approach shows relate high precision coverage, enrich clearly have strong impact construction Web.