Unsupervised Attention-based Sentence-Level Meta-Embeddings from Contextualised Language Models

Takahashi, Keigo and Bollegala, Danushka ORCID: 0000-0003-4476-7003 (2022) Unsupervised Attention-based Sentence-Level Meta-Embeddings from Contextualised Language Models. In: Language Resources and Evaluation Conference (LREC), 2022-6-20 - 2022-6-25, Marseille, France.

Text Meta_Sentence_Embedding__LREC_2022_ (2).pdf - Author Accepted Manuscript Download (240kB) | Preview

Abstract

A variety of contextualised language models have been proposed in the NLP community, which are trained on diverse corpora to produce numerous Neural Language Models (NLMs). However, different NLMs have reported different levels of performances in downstream NLP applications when used as text representations. We propose a sentence-level meta-embedding learning method that takes independently trained contextualised word embedding models and learns a sentence embedding that preserves the complementary strengths of the input source NLMs. Our proposed method is unsupervised and is not tied to a particular downstream task, which makes the learnt meta-embeddings in principle applicable to different tasks that require sentence representations. Specifically, we first project the token-level embeddings obtained by the individual NLMs and learn attention weights that indicate the contributions of source embeddings towards their token-level meta-embeddings. Next, we apply mean and max pooling to produce sentence-level meta-embeddings from token-level meta-embeddings. Experimental results on semantic textual similarity benchmarks show that our proposed unsupervised sentence-level meta-embedding method outperforms previously proposed sentence-level meta-embedding methods as well as a supervised baseline.

Item Type:	Conference or Workshop Item (Unspecified)
Uncontrolled Keywords:	meta-sentence embedding, unsupervised, contextualised language model
Divisions:	Faculty of Science and Engineering > School of Electrical Engineering, Electronics and Computer Science
Depositing User:	Symplectic Admin
Date Deposited:	14 Apr 2022 13:56
Last Modified:	23 Nov 2023 22:56
Related URLs:	Author
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3152960