What do Language Models know about word senses? Zero-Shot WSD with Language Models and Domain Inventories
Oscar Sainz, Oier Lopez de Lacalle, Eneko Agirre, German Rigau
Abstract
Language Models are the core for almost any Natural Language Processing system nowadays. One of their particularities is their contextualized representations, a game changer feature when a disambiguation between word senses is necessary. In this paper we aim to explore to what extent language models are capable of discerning among senses at inference time. We performed this analysis by prompting commonly used Languages Models such as BERT or RoBERTa to perform the task of Word Sense Disambiguation (WSD). We leverage the relation between word senses and domains, and cast WSD as a textual entailment problem, where the different hypothesis refer to the domains of the word senses. Our results show that this approach is indeed effective, close to supervised systems.- Anthology ID:
- 2023.gwc-1.40
- Volume:
- Proceedings of the 12th Global Wordnet Conference
- Month:
- January
- Year:
- 2023
- Address:
- University of the Basque Country, Donostia - San Sebastian, Basque Country
- Editors:
- German Rigau, Francis Bond, Alexandre Rademaker
- Venue:
- GWC
- SIG:
- Publisher:
- Global Wordnet Association
- Note:
- Pages:
- 331–342
- Language:
- URL:
- https://aclanthology.org/2023.gwc-1.40
- DOI:
- Cite (ACL):
- Oscar Sainz, Oier Lopez de Lacalle, Eneko Agirre, and German Rigau. 2023. What do Language Models know about word senses? Zero-Shot WSD with Language Models and Domain Inventories. In Proceedings of the 12th Global Wordnet Conference, pages 331–342, University of the Basque Country, Donostia - San Sebastian, Basque Country. Global Wordnet Association.
- Cite (Informal):
- What do Language Models know about word senses? Zero-Shot WSD with Language Models and Domain Inventories (Sainz et al., GWC 2023)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-1/2023.gwc-1.40.pdf