Capturing and measuring thematic relatedness

Magdalena Kacmajor, John D. Kelleher

Research output: Contribution to journalArticlepeer-review

Abstract

In this paper we explain the difference between twoaspects of semantic relatedness: taxonomic and thematic relations. We notice the lack of evaluation tools for measuring thematic relatedness, identify two datasets that can be recommended as thematic benchmarks, and verify them experimentally. In further experiments, we use these datasets to perform a comprehensive analysis of the performance of an extensive sample of computational models ofsemantic relatedness, classified according to the sources of information they exploit. We report models that are best at each of the two dimensions of semantic relatedness and those that achieve a good balance between the two.

Original languageEnglish
Pages (from-to)645-682
Number of pages38
JournalLanguage Resources and Evaluation
Volume54
Issue number3
DOIs
Publication statusPublished - 1 Sep 2020

Keywords

  • Evaluation datasets
  • Semantic relatedness
  • Thematic relations
  • Word vector representations

Fingerprint

Dive into the research topics of 'Capturing and measuring thematic relatedness'. Together they form a unique fingerprint.

Cite this