Improving Multi-Party Dialogue Discourse Parsing via Domain Integration

TL;DR

This paper introduces a Transformer-based parser for multi-party dialogue discourse analysis, enhancing cross-domain generalization through domain integration techniques, and demonstrating improved performance on diverse dialogue datasets.

Contribution

The paper proposes a novel Transformer-based parser and three domain integration methods to improve cross-domain dialogue discourse parsing performance.

Findings

Neural parser benefits from domain integration methods.

Improved cross-domain performance on dialogue datasets.

Transformer-based parser outperforms previous models.

Abstract

While multi-party conversations are often less structured than monologues and documents, they are implicitly organized by semantic level correlations across the interactive turns, and dialogue discourse analysis can be applied to predict the dependency structure and relations between the elementary discourse units, and provide feature-rich structural information for downstream tasks. However, the existing corpora with dialogue discourse annotation are collected from specific domains with limited sample sizes, rendering the performance of data-driven approaches poor on incoming dialogues without any domain adaptation. In this paper, we first introduce a Transformer-based parser, and assess its cross-domain performance. We next adopt three methods to gain domain integration from both data and language modeling perspectives to improve the generalization capability. Empirical results show that the neural parser can benefit from our proposed methods, and performs better on cross-domain dialogue samples.