Transparent Semantic Change Detection with Dependency-Based Profiles

Abstract

Most modern computational approaches to lexical semantic change detection (LSC) rely on embedding-based distributional word representations with neural networks.Despite the strong performance on LSC benchmarks, they are often opaque.We investigate an alternative method which relies purely on dependency co-occurrence patterns of words.We demonstrate that it is effective for semantic change detection and even outperforms a number of distributional semantic models.We provide an in-depth quantitative and qualitative analysis of the predictions, showing that they are plausible and interpretable.

Authors 5

1. Bach Phan Tat
2. Kris Heylen Aachen Instituut voor de Nederlandse Taal

   Instituut voor de Nederlandse Taal · KU Leuven

   Affiliation as printed

   Department of Linguistics , KU Leuven

   Instituut voor de Nederlandse Taal

3. Dirk Geeraerts

   KU Leuven

   Affiliation as printed

   Department of Linguistics , KU Leuven

4. Stefano De Pascale

   Vrije Universiteit Brussel · KU Leuven

   Affiliation as printed

   Department of Linguistics , KU Leuven

   Vrije Universiteit Brussel

5. Dirk Speelman

   KU Leuven

   Affiliation as printed

   Department of Linguistics , KU Leuven