Learning to Paraphrase for Question Answering

TL;DR

This paper introduces a neural framework that learns paraphrases to improve question answering systems by better capturing the variety of natural language expressions of the same information need.

Contribution

It presents an end-to-end training method using question-answer pairs to learn paraphrases that enhance QA performance across multiple datasets.

Findings

Improves QA accuracy on Freebase and sentence selection tasks

Achieves competitive results with simple QA models

Consistently outperforms baseline methods

Abstract

Question answering (QA) systems are sensitive to the many different ways natural language expresses the same information need. In this paper we turn to paraphrases as a means of capturing this knowledge and present a general framework which learns felicitous paraphrases for various QA tasks. Our method is trained end-to-end using question-answer pairs as a supervision signal. A question and its paraphrases serve as input to a neural scoring model which assigns higher weights to linguistic expressions most likely to yield correct answers. We evaluate our approach on QA over Freebase and answer sentence selection. Experimental results on three datasets show that our framework consistently improves performance, achieving competitive results despite the use of simple QA models.