Evaluating unsupervised dimensionality reduction methods for pretrained sentence embeddings

Zhang, Gaifan, Zhou, Yi ORCID: https://orcid.org/0000-0001-7009-8515 and Bollegala, Danushka 2024. Evaluating unsupervised dimensionality reduction methods for pretrained sentence embeddings. Presented at: The 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), Turin, Italy, 20-25 May 2024. Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024). Turin, Italy: ELRA and ICCL, pp. 6530-6543.

Preview

PDF - Published Version Available under License Creative Commons Attribution Non-commercial. Download (2MB) | Preview

Abstract

Sentence embeddings produced by Pretrained Language Models (PLMs) have received wide attention from the NLP community due to their superior performance when representing texts in numerous downstream applications. However, the high dimensionality of the sentence embeddings produced by PLMs is problematic when representing large numbers of sentences in memory- or compute-constrained devices. As a solution, we evaluate unsupervised dimensionality reduction methods to reduce the dimensionality of sentence embeddings produced by PLMs. Our experimental results show that simple methods such as Principal Component Analysis (PCA) can reduce the dimensionality of sentence embeddings by almost 50%, without incurring a significant loss in performance in multiple downstream tasks. Surprisingly, reducing the dimensionality further improves performance over the original high dimensional versions for the sentence embeddings produced by some PLMs in some tasks.

Item Type:	Conference or Workshop Item (Paper)
Date Type:	Publication
Status:	Published
Schools:	Computer Science & Informatics
Publisher:	ELRA and ICCL
Date of First Compliant Deposit:	20 November 2024
Date of Acceptance:	2024
Last Modified:	26 Nov 2024 16:16
URI:	https://orca.cardiff.ac.uk/id/eprint/173670

Actions (repository staff only)

Edit Item