Probabilistic author-topic models for information discovery

Mark Steyvers, Padhraic Smyth, Michal Rosen-Zvi, Thomas Griffiths

Research output: Chapter in Book/Report/Conference proceedingConference contribution

482 Scopus citations

Abstract

We propose a new unsupervised learning technique for extracting information from large text collections. We model documents as if they were generated by a two-stage stochastic process. Each author is represented by a probability distribution over topics, and each topic is represented as a probability distribution over words for that topic. The words in a multi-author paper are assumed to be the result of a mixture of each authors' topic mixture. The topic-word and author-topic distributions are learned from data in an unsupervised manner using a Markov chain Monte Carlo algorithm. We apply the methodology to a large corpus of 160,000 abstracts and 85,000 authors from the well-known CiteSeer digital library, and learn a model with 300 topics. We discuss in detail the interpretation of the results discovered by the system including specific topic and author models, ranking of authors by topic and topics by author, significant trends in the computer science literature between 1990 and 2002, parsing of abstracts by topics and authors and detection of unusual papers by specific authors. An online query interface to the model is also discussed that allows interactive exploration of author-topic models for corpora such as CiteSeer.

Original languageEnglish (US)
Title of host publicationKDD-2004 - Proceedings of the Tenth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
PublisherAssociation for Computing Machinery
Pages306-315
Number of pages10
ISBN (Print)1581138881, 9781581138887
DOIs
StatePublished - 2004
Externally publishedYes
EventKDD-2004 - Proceedings of the Tenth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - Seattle, WA, United States
Duration: Aug 22 2004Aug 25 2004

Publication series

NameKDD-2004 - Proceedings of the Tenth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

Other

OtherKDD-2004 - Proceedings of the Tenth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
Country/TerritoryUnited States
CitySeattle, WA
Period8/22/048/25/04

All Science Journal Classification (ASJC) codes

  • General Engineering

Keywords

  • Gibbs sampling
  • Text modeling
  • Unsupervised learning

Fingerprint

Dive into the research topics of 'Probabilistic author-topic models for information discovery'. Together they form a unique fingerprint.

Cite this