Detecting Speech Interruptions for Automatic Conflict Detection - Conflict and Multimodal Communication: Social Research and Machine Intelligence - page 365

Chapter 18
Detecting Speech Interruptions for Automatic
Conflict Detection

Marie-José Caraty and Claude Montacié

18.1 Introduction

Research in organization and management has investigated phenomena, such as
the causes, effects, and handling of interpersonal or intergroup conflicts (Korabik
et al. 1993 ; Macintosh and Stevens 2008 ; Thomas et al. 2008 ). Data on these social
and psychological phenomena are collected from people who are involved in the
conflict, witnesses of the conflict, or, by extension, looking at a recording of the
conflict escalation between the protagonists. A large quantity of audio and/or video
metadata can be extracted from these recordings, such as the conversation, face, and
gesture interactions. In this chapter, the conversational interactions during political
debates have been studied to develop an automatic conflict detector from voice
analysis. A reliable detector of conflict would be useful for many applications, such
as security in public places, the quality of customer services, and the deployment
of intelligent agents. The development of such a system requires modeling of the
conversational interactions as well as the search for specific interactions in relation
to a given measure of conflict handling (Rahim 1983 ;Dalyetal. 2010 ).