Team OS's System for Dialogue Robot Competition 2022

TL;DR

This paper presents OSbot, a dialogue robot system designed for the 2022 Dialogue Robot Competition, utilizing manual state transitions, keyword extraction, and sentiment analysis for effective dialogue management.

Contribution

The paper introduces a flexible dialogue system with manually editable state transitions based on keyword and sentiment analysis, tailored for competitive dialogue robots.

Findings

Achieved third place in the preliminary round of the competition.

Implemented a dialogue flow management system that is easy to view and edit.

Utilized SVM-based sentiment analysis trained on a multimodal corpus.

Abstract

This paper describes our dialogue robot system, OSbot, developed for Dialogue Robot Competition 2022. The dialogue flow is based on state transitions described manually and the transition conditions use the results of keyword extraction and sentiment analysis. The transitions can be easily viewed and edited by managing them on a spreadsheet. The keyword extraction is based on named entity extraction and our predefined keyword set. The sentiment analysis is text-based and uses SVM, which was trained with the multimodal dialogue corpus Hazumi. We quickly checked and edited a dialogue flow by using a logging function. In the competition's preliminary round, our system ended up in third place.