Modelling Lexical Ambiguity with Density Matrices

Francois Meyer, Martha Lewis


Abstract
Words can have multiple senses. Compositional distributional models of meaning have been argued to deal well with finer shades of meaning variation known as polysemy, but are not so well equipped to handle word senses that are etymologically unrelated, or homonymy. Moving from vectors to density matrices allows us to encode a probability distribution over different senses of a word, and can also be accommodated within a compositional distributional model of meaning. In this paper we present three new neural models for learning density matrices from a corpus, and test their ability to discriminate between word senses on a range of compositional datasets. When paired with a particular composition method, our best model outperforms existing vector-based compositional models as well as strong sentence encoders.
Anthology ID:
2020.conll-1.21
Volume:
Proceedings of the 24th Conference on Computational Natural Language Learning
Month:
November
Year:
2020
Address:
Online
Venue:
CoNLL
SIG:
SIGNLL
Publisher:
Association for Computational Linguistics
Note:
Pages:
276–290
Language:
URL:
https://aclanthology.org/2020.conll-1.21
DOI:
10.18653/v1/2020.conll-1.21
Bibkey:
Cite (ACL):
Francois Meyer and Martha Lewis. 2020. Modelling Lexical Ambiguity with Density Matrices. In Proceedings of the 24th Conference on Computational Natural Language Learning, pages 276–290, Online. Association for Computational Linguistics.
Cite (Informal):
Modelling Lexical Ambiguity with Density Matrices (Meyer & Lewis, CoNLL 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.conll-1.21.pdf
Optional supplementary material:
 2020.conll-1.21.OptionalSupplementaryMaterial.zip