A Compare Aggregate Transformer for Understanding Document-grounded Dialogue

TL;DR

This paper introduces the Compare Aggregate Transformer (CAT), a novel model that denoises dialogue context and effectively utilizes external documents for improved response generation in document-grounded dialogue systems.

Contribution

The paper proposes a new Compare Aggregate Transformer with comparison mechanisms to reduce noise and enhance document information aggregation for dialogue response generation.

Findings

CAT outperforms state-of-the-art models on CMUDoG dataset

Introduces two metrics for evaluating document utilization efficiency

Effectively reduces noise in dialogue context processing

Abstract

Unstructured documents serving as external knowledge of the dialogues help to generate more informative responses. Previous research focused on knowledge selection (KS) in the document with dialogue. However, dialogue history that is not related to the current dialogue may introduce noise in the KS processing. In this paper, we propose a Compare Aggregate Transformer (CAT) to jointly denoise the dialogue context and aggregate the document information for response generation. We designed two different comparison mechanisms to reduce noise (before and during decoding). In addition, we propose two metrics for evaluating document utilization efficiency based on word overlap. Experimental results on the CMUDoG dataset show that the proposed CAT model outperforms the state-of-the-art approach and strong baselines.