Paper: Automatic Labelling of Topic Models

ACL ID P11-1154
Title Automatic Labelling of Topic Models
Venue Annual Meeting of the Association of Computational Linguistics
Session Main Conference
Year 2011
Authors

We propose a method for automatically la- belling topics learned via LDA topic models. We generate our label candidate set from the top-rankingtopicterms, titlesofWikipediaar- ticles containing the top-ranking topic terms, and sub-phrases extracted from the Wikipedia article titles. We rank the label candidates us- ingacombinationofassociationmeasuresand lexical features, optionally fed into a super- vised ranking model. Our method is shown to performstronglyoverfourindependentsetsof topics, significantly better than a benchmark method.