Semantic Enrichment of Twitter Posts for User Profile Construction

Presented at: 8th Extended Semantic Web Conference (ESWC2011)

by Qi Gao, Fabian Abel, Ke Tao, Geert-Jan Houben

As the most popular microblogging platform, the vast amount of content on Twitter is constantly growing so that the retrieval of relevant information (streams) is becoming more and more difficult every day. Representing the semantics of individual Twitter activities and modeling the interests of Twitter users would allow for personalization and therewith countervail the information overload. Given the variety of topics people discuss on Twitter, semantic user profiles generated from Twitter posts moreover promise to be beneficial also for other applications on the Social Web. However, automatically inferring the semantic meaning of tweets is a non-trivial problem. In this paper we investigate semantic user modeling based on Twitter activities. We introduce and analyze methods for linking Twitter posts with related news articles in order to contextualize Twitter activities. We then propose and compare strategies that exploit the semantics extracted from both tweets and related news articles to represent individual Twitter activities in a semantically meaningful way. A large-scale evaluation validates the applicability of our approach and shows that our methods relate tweets to news articles with high precision, enrich the semantics of tweets clearly and have tremendous impact on the construction of semantic user profiles for the Social Web.

Keywords: news, semantic enrichment, social web, twitter, user profile construction

Resource URI on the dog food server:

Explore this resource elsewhere: