We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computation and Language

Title: Always Keep your Target in Mind: Studying Semantics and Improving Performance of Neural Lexical Substitution

Abstract: Lexical substitution, i.e. generation of plausible words that can replace a particular target word in a given context, is an extremely powerful technology that can be used as a backbone of various NLP applications, including word sense induction and disambiguation, lexical relation extraction, data augmentation, etc. In this paper, we present a large-scale comparative study of lexical substitution methods employing both rather old and most recent language and masked language models (LMs and MLMs), such as context2vec, ELMo, BERT, RoBERTa, XLNet. We show that already competitive results achieved by SOTA LMs/MLMs can be further substantially improved if information about the target word is injected properly. Several existing and new target word injection methods are compared for each LM/MLM using both intrinsic evaluation on lexical substitution datasets and extrinsic evaluation on word sense induction (WSI) datasets. On two WSI datasets we obtain new SOTA results. Besides, we analyze the types of semantic relations between target words and their substitutes generated by different models or given by annotators.
Comments: arXiv admin note: text overlap with arXiv:2006.00031
Subjects: Computation and Language (cs.CL)
Journal reference: Proceedings of the 28th International Conference on Computational Linguistics, pages 1242-1255, Barcelona, Spain (Online). International Committee on Computational Linguistics. 2022
DOI: 10.18653/v1/2020.coling-main.107
Cite as: arXiv:2206.11815 [cs.CL]
  (or arXiv:2206.11815v1 [cs.CL] for this version)

Submission history

From: Nikolay Arefyev [view email]
[v1] Tue, 7 Jun 2022 16:16:19 GMT (2735kb,D)

Link back to: arXiv, form interface, contact.