Distilling relation embeddings from pre-trained language models

Ushio, Asahi, Camacho Collados, Jose ORCID: https://orcid.org/0000-0003-1618-7239 and Schockaert, Steven ORCID: https://orcid.org/0000-0002-9256-2881 2021. Distilling relation embeddings from pre-trained language models. Presented at: 2021 Conference on Empirical Methods in Natural Language Processing (EMNLP), Punta Cana, Dominican Republic, 7-11 November 2021. Published in: Moens, M., Huang, X., Specia, L. and Yih, S. eds. Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics, pp. 9044-9062. 10.18653/v1/2021.emnlp-main.712

PDF - Accepted Post-Print Version Download (342kB)

Abstract

Pre-trained language models have been found to capture a surprisingly rich amount of lexical knowledge, ranging from commonsense properties of everyday concepts to detailed factual knowledge about named entities. Among others, this makes it possible to distill high-quality word vectors from pre-trained language models. However, it is currently unclear to what extent it is possible to distill relation embeddings, i.e. vectors that characterize the relationship between two words. Such relation embeddings are appealing because they can, in principle, encode relational knowledge in a more fine-grained way than is possible with knowledge graphs. To obtain relation embeddings from a pre-trained language model, we encode word pairs using a (manually or automatically generated) prompt, and we fine-tune the language model such that relationally similar word pairs yield similar output vectors. We find that the resulting relation embeddings are highly competitive on analogy (unsupervised) and relation classification (supervised) benchmarks, even without any task-specific fine-tuning. Source code to reproduce our experimental results and the model checkpoints are available in the following repository: https://github.com/asahi417/relbert

Item Type:	Conference or Workshop Item (Paper)
Date Type:	Publication
Status:	Published
Schools:	Schools > Computer Science & Informatics
Publisher:	Association for Computational Linguistics
ISBN:	978-195591709-4
Date of First Compliant Deposit:	19 October 2021
Date of Acceptance:	26 August 2021
Last Modified:	30 Jul 2025 13:57
URI:	https://orca.cardiff.ac.uk/id/eprint/144078

Citation Data

Cited 8 times in Scopus. View in Scopus. Powered By Scopus® Data

Actions (repository staff only)

Edit Item