Phrase-BERT: Improved Phrase Embeddings from BERT with an Application to Corpus Exploration

TL;DR

This paper introduces Phrase-BERT, a fine-tuned BERT model that generates more meaningful phrase embeddings, improving phrase similarity tasks and enabling effective phrase-based topic modeling.

Contribution

We propose a contrastive fine-tuning method for BERT using paraphrase datasets, enhancing phrase embeddings for better semantic and compositional understanding.

Findings

Outperforms baseline models on phrase similarity tasks

Increases lexical diversity in embedding space

Enables phrase-based neural topic modeling with improved coherence

Abstract

Phrase representations derived from BERT often do not exhibit complex phrasal compositionality, as the model relies instead on lexical similarity to determine semantic relatedness. In this paper, we propose a contrastive fine-tuning objective that enables BERT to produce more powerful phrase embeddings. Our approach (Phrase-BERT) relies on a dataset of diverse phrasal paraphrases, which is automatically generated using a paraphrase generation model, as well as a large-scale dataset of phrases in context mined from the Books3 corpus. Phrase-BERT outperforms baselines across a variety of phrase-level similarity tasks, while also demonstrating increased lexical diversity between nearest neighbors in the vector space. Finally, as a case study, we show that Phrase-BERT embeddings can be easily integrated with a simple autoencoder to build a phrase-based neural topic model that interprets topics as mixtures of words and phrases by performing a nearest neighbor search in the embedding space. Crowdsourced evaluations demonstrate that this phrase-based topic model produces more coherent and meaningful topics than baseline word and phrase-level topic models, further validating the utility of Phrase-BERT.