Coarse-grain Fine-grain Coattention Network for Multi-evidence Question Answering

TL;DR

This paper introduces the Coarse-grain Fine-grain Coattention Network (CFC), a novel model that effectively combines evidence from multiple documents for multi-evidence question answering, achieving state-of-the-art results on WikiHop.

Contribution

The paper presents a new neural architecture that integrates coarse and fine-grain coattention modules to improve multi-evidence question answering without pretrained encoders.

Findings

Achieved 70.6% accuracy on WikiHop blind test set.

Outperformed previous best by 3% accuracy.

Demonstrated effectiveness of hierarchical coattention in multi-document QA.

Abstract

End-to-end neural models have made significant progress in question answering, however recent studies show that these models implicitly assume that the answer and evidence appear close together in a single document. In this work, we propose the Coarse-grain Fine-grain Coattention Network (CFC), a new question answering model that combines information from evidence across multiple documents. The CFC consists of a coarse-grain module that interprets documents with respect to the query then finds a relevant answer, and a fine-grain module which scores each candidate answer by comparing its occurrences across all of the documents with the query. We design these modules using hierarchies of coattention and self-attention, which learn to emphasize different parts of the input. On the Qangaroo WikiHop multi-evidence question answering task, the CFC obtains a new state-of-the-art result of 70.6% on the blind test set, outperforming the previous best by 3% accuracy despite not using pretrained contextual encoders.