Task-Specific Attentive Pooling of Phrase Alignments Contributes to Sentence Matching
It addresses sentence matching for NLP applications, offering a task-aware method that is incremental over prior work.
The paper tackles sentence matching by analyzing how phrase alignment importance varies between textual entailment and answer selection tasks, proposing a GRU-based architecture with task-specific attentive pooling that improves performance on both tasks.
This work studies comparatively two typical sentence matching tasks: textual entailment (TE) and answer selection (AS), observing that weaker phrase alignments are more critical in TE, while stronger phrase alignments deserve more attention in AS. The key to reach this observation lies in phrase detection, phrase representation, phrase alignment, and more importantly how to connect those aligned phrases of different matching degrees with the final classifier. Prior work (i) has limitations in phrase generation and representation, or (ii) conducts alignment at word and phrase levels by handcrafted features or (iii) utilizes a single framework of alignment without considering the characteristics of specific tasks, which limits the framework's effectiveness across tasks. We propose an architecture based on Gated Recurrent Unit that supports (i) representation learning of phrases of arbitrary granularity and (ii) task-specific attentive pooling of phrase alignments between two sentences. Experimental results on TE and AS match our observation and show the effectiveness of our approach.