External Knowledge Selection with Weighted Negative Sampling in Knowledge-grounded Task-oriented Dialogue Systems

TL;DR

This paper presents a system for knowledge-grounded task-oriented dialogue on spoken conversations, introducing weighted negative sampling, data construction, post-training, and style transfer to improve response relevance and style.

Contribution

It introduces weighted negative sampling and other techniques for knowledge selection and response generation in spoken dialogue systems, addressing data scarcity and style consistency.

Findings

Weighted negative sampling improves knowledge selection accuracy.

Post-training and style transfer enhance response relevance and style matching.

The system ranked 7th in objective and 6th in human evaluation among 16 teams.

Abstract

Constructing a robust dialogue system on spoken conversations bring more challenge than written conversation. In this respect, DSTC10-Track2-Task2 is proposed, which aims to build a task-oriented dialogue (TOD) system incorporating unstructured external knowledge on a spoken conversation, extending DSTC9-Track1. This paper introduces our system containing four advanced methods: data construction, weighted negative sampling, post-training, and style transfer. We first automatically construct a large training data because DSTC10-Track2 does not release the official training set. For the knowledge selection task, we propose weighted negative sampling to train the model more fine-grained manner. We also employ post-training and style transfer for the response generation task to generate an appropriate response with a similar style to the target response. In the experiment, we investigate the effect of weighted negative sampling, post-training, and style transfer. Our model ranked 7 out of 16 teams in the objective evaluation and 6 in human evaluation.