TUMS: twitter-based user modeling service

  • Authors:
  • Ke Tao;Fabian Abel;Qi Gao;Geert-Jan Houben

  • Affiliations:
  • Web Information Systems, Delft University of Technology, The Netherlands;Web Information Systems, Delft University of Technology, The Netherlands;Web Information Systems, Delft University of Technology, The Netherlands;Web Information Systems, Delft University of Technology, The Netherlands

  • Venue:
  • ESWC'11 Proceedings of the 8th international conference on The Semantic Web
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Twitter is today's most popular micro-blogging service on the Social Web. As people discuss various fresh topics, Twitter messages (tweets) can tell much about the current interests and concerns of a user. In this paper, we introduce TUMS, a Twitter-based User Modeling Service, that infers semantic user profiles from the messages people post on Twitter. It features topic detection and entity extraction for tweets and allows for further enrichment by linking tweets to news articles that describe the context of the tweets. TUMS is made publicly available as a Web application. It allows end-users to overview Twitter-based profiles in a structured way and allows them to see in which topics or entities a user was interested at a specific point in time. Furthermore, it provides Twitter-based user profiles in RDF format and allows applications to incorporate these profiles in order to adapt their functionality to the current interests of a user. TUMS is available via: http://wis.ewi.tudelft.nl/tums/