Joint Learning of Sentence Embeddings for Relevance and Entailment

TL;DR

This paper introduces a joint learning approach for sentence embeddings to improve relevance and entailment detection in natural language questions, demonstrating its effectiveness on new and existing datasets.

Contribution

It proposes a novel joint training method for sentence embeddings that models relevance and entailment simultaneously without explicit supervision.

Findings

Joint training of embeddings is feasible and effective.

The model improves state-of-the-art on multiple-choice question ranking.

Using unrelated task-trained models enhances performance on small datasets.

Abstract

We consider the problem of Recognizing Textual Entailment within an Information Retrieval context, where we must simultaneously determine the relevancy as well as degree of entailment for individual pieces of evidence to determine a yes/no answer to a binary natural language question. We compare several variants of neural networks for sentence embeddings in a setting of decision-making based on evidence of varying relevance. We propose a basic model to integrate evidence for entailment, show that joint training of the sentence embeddings to model relevance and entailment is feasible even with no explicit per-evidence supervision, and show the importance of evaluating strong baselines. We also demonstrate the benefit of carrying over text comprehension model trained on an unrelated task for our small datasets. Our research is motivated primarily by a new open dataset we introduce, consisting of binary questions and news-based evidence snippets. We also apply the proposed relevance-entailment model on a similar task of ranking multiple-choice test answers, evaluating it on a preliminary dataset of school test questions as well as the standard MCTest dataset, where we improve the neural model state-of-art.