Learn to Resolve Conversational Dependency: A Consistency Training Framework for Conversational Question Answering

TL;DR

This paper introduces ExCorD, a framework that improves conversational question answering by explicitly training models to resolve dependencies like anaphora and ellipsis through generating self-contained questions and using consistency regularization.

Contribution

The paper presents a novel framework, ExCorD, that explicitly trains QA models to resolve conversational dependencies, significantly enhancing their understanding of dialogue context.

Findings

ExCorD improves F1 scores by up to 1.2 on QuAC.

ExCorD improves F1 scores by up to 5.2 on CANARD.

The approach effectively addresses limitations of existing methods.

Abstract

One of the main challenges in conversational question answering (CQA) is to resolve the conversational dependency, such as anaphora and ellipsis. However, existing approaches do not explicitly train QA models on how to resolve the dependency, and thus these models are limited in understanding human dialogues. In this paper, we propose a novel framework, ExCorD (Explicit guidance on how to resolve Conversational Dependency) to enhance the abilities of QA models in comprehending conversational context. ExCorD first generates self-contained questions that can be understood without the conversation history, then trains a QA model with the pairs of original and self-contained questions using a consistency-based regularizer. In our experiments, we demonstrate that ExCorD significantly improves the QA models' performance by up to 1.2 F1 on QuAC, and 5.2 F1 on CANARD, while addressing the limitations of the existing approaches.