Towards Structure-aware Paraphrase Identification with Phrase Alignment Using Sentence Encoders

TL;DR

This paper introduces a structure-aware paraphrase identification method that combines sentence encoders with span alignment, improving sensitivity to sentence structure and interpretability over traditional similarity measures.

Contribution

It proposes a novel approach that decomposes sentence similarity into span alignments, enhancing structural sensitivity and interpretability in paraphrase detection.

Findings

Improved paraphrase identification performance.

Enhanced sensitivity to structural differences.

Better distinction of non-paraphrases with high lexical overlap.

Abstract

Previous works have demonstrated the effectiveness of utilising pre-trained sentence encoders based on their sentence representations for meaning comparison tasks. Though such representations are shown to capture hidden syntax structures, the direct similarity comparison between them exhibits weak sensitivity to word order and structural differences in given sentences. A single similarity score further makes the comparison process hard to interpret. Therefore, we here propose to combine sentence encoders with an alignment component by representing each sentence as a list of predicate-argument spans (where their span representations are derived from sentence encoders), and decomposing the sentence-level meaning comparison into the alignment between their spans for paraphrase identification tasks. Empirical results show that the alignment component brings in both improved performance and interpretability for various sentence encoders. After closer investigation, the proposed approach indicates increased sensitivity to structural difference and enhanced ability to distinguish non-paraphrases with high lexical overlap.