Capturing Global Structural Information in Long Document Question Answering with Compressive Graph Selector Network

TL;DR

This paper introduces the Compressive Graph Selector Network (CGSN), a novel model that captures global and local structural information in long documents to improve question answering performance.

Contribution

The paper presents CGSN, which integrates local and global graph networks with an evidence memory module for better long-range reasoning in long document QA.

Findings

Outperforms previous methods on two datasets

Effectively captures long-range dependencies

Reduces redundancy in evidence selection

Abstract

Long document question answering is a challenging task due to its demands for complex reasoning over long text. Previous works usually take long documents as non-structured flat texts or only consider the local structure in long documents. However, these methods usually ignore the global structure of the long document, which is essential for long-range understanding. To tackle this problem, we propose Compressive Graph Selector Network (CGSN) to capture the global structure in a compressive and iterative manner. The proposed model mainly focuses on the evidence selection phase of long document question answering. Specifically, it consists of three modules: local graph network, global graph network and evidence memory network. Firstly, the local graph network builds the graph structure of the chunked segment in token, sentence, paragraph and segment levels to capture the short-term dependency of the text. Secondly, the global graph network selectively receives the information of each level from the local graph, compresses them into the global graph nodes and applies graph attention to the global graph nodes to build the long-range reasoning over the entire text in an iterative way. Thirdly, the evidence memory network is designed to alleviate the redundancy problem in the evidence selection by saving the selected result in the previous steps. Extensive experiments show that the proposed model outperforms previous methods on two datasets.