Detecting Speech Interruptions for Automatic Conflict Detection - Conflict and Multimodal Communication: Social Research and Machine Intelligence - page 387

gave a UAR of 85.3 %. These results show an improvement of 4.5 % compared to
the results of the baseline system of the Conflict Challenge (UAR of 80.8 %).
These experiments have shown the capability of a multi-expert architecture to
integrate a piece of conflict knowledge. Other knowledge that is related to the turn-
taking patterns, such as the modeling of the moderator role (Vinciarelli 2007 ), or
that is related to the nonverbal interactions, such as the movements of the body, the
head, and the arms, could be integrated into the conflict detector.
Acknowledgments Many thanks to Björn Schuller (TUM, Germany), Stefan Steidl (FAU
Erlangen-Nuremberg, Germany), and Anton Batliner (TUM, Germany) for the organization of the
Interspeech 2013 Conflict Challenge and special thanks to Alessandro Vinciarelli (University of
Glasgow, UK) for the SSPNet Conflict Corpus.
References
Atkinson JM, Drew P (1979) Order in court: the organisation of verbal interaction in judicial
settings. Humanities Press, Atlantic Highlands, NJ
Barzilay R, Collins M, Hirschberg J, Whittaker S (2000) The rules behind the roles: identifying
speaker roles in radio broadcasts. Paper presented at 17th National conference on artificial
intelligence, Austin, USA, 30 July-3 Aug, pp 679-684
Beattie GW (1982) Turn-taking and interruption in political interviews: Margaret Thatcher and Jim
Callaghan compared and contrasted. Semiotica 39(1-2):93-114
Boakye K, Trueba-Hornero B, Vinyals O, Friedland G (2008) Overlapped speech detection for
improved diarization in multi-party meetings. Paper presented at ICASSP Conference, Las
Vegas, USA, 31 Mar-4 Apr, pp 4353-4356
Boden D (1994) The business of talk. Organizations in action. Polity Press, London
Brinson SL, Winn JE (1997) Talk shows' representations of interpersonal conflicts. J Broadcast
Electron Media 41(1):25-39
Chen Z, Feng TJ, Houkes Z (2000) Incorporating a priori knowledge into initialized weights for
neural classifier. Paper presented at international joint conference on neural networks (IJCNN),
Como, Italy, 24-27 July, pp 291-296
Clancy PM, Thompson SA, Suzuki R, Tao H (1996) The conversational use of reactive tokens in
English, Japanese and Mandarin. J Pragmat 26:355-387
Daly TM, Lee JA, Soutar GN, Rasmi S (2010) Conflict-handling style measurement: a best-worst
scaling application. Int J Confl Manag 21(3):281-308
De Ruiter JP, Mitterer H, Enfield NJ (2006) Projecting the end of a speaker's turn: a cognitive
cornerstone of conversation. Language 82(3):515-535
Decoste D, Scholkopf B (2002) Training invariant support vector machines. Mach Learn
46(1-3):161-190
Eyben F, Wöllmer M, Schuller B (2010) openSMILE the Munich versatile and fast open-source
audio feature extractor. Paper presented at the ACM multimedia conference (MM), Florence,
Italy, 25-29 Oct, pp 1459-1462
Garcia A (1991) Dispute resolution without disputing: how the interactional organization of
mediation hearings minimizes argumentative talk. Am Sociol Rev 56:818-835
Gravano A, Hirschberg J (2011) Turn-taking cues in task oriented dialogue. Comput Speech Lang
25(3):601-634
Grèzes F, Richards J, Rosenberg A (2013) Let me finish: automatic conflict detection using speaker
overlap. Paper presented at the Interspeech conference, Lyon, France, 25-29 Aug, 5 pages