Effective Incorporation of Speaker Information in Utterance Encoding in Dialog

TL;DR

This paper introduces a relative speaker modeling method that improves dialog encoding by effectively incorporating speaker information, leading to better performance in dialog act recognition and response generation.

Contribution

A novel relative speaker modeling approach that addresses inconsistencies in speaker annotations and enhances dialog encoding effectiveness.

Findings

Improved dialog act recognition accuracy

Enhanced response generation quality

More consistent performance across dialogs

Abstract

In dialog studies, we often encode a dialog using a hierarchical encoder where each utterance is converted into an utterance vector, and then a sequence of utterance vectors is converted into a dialog vector. Since knowing who produced which utterance is essential to understanding a dialog, conventional methods tried integrating speaker labels into utterance vectors. We found the method problematic in some cases where speaker annotations are inconsistent among different dialogs. A relative speaker modeling method is proposed to address the problem. Experimental evaluations on dialog act recognition and response generation show that the proposed method yields superior and more consistent performances.