How Interlocutors Coordinate with each other within Emotional Segments?

Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers, pages 728–738, Osaka, Japan, December 11-17 2016. How Interlocutors Coordinate with each other within Emotional Segments? Firoj Alam, Shammur Absar Chowdhury, Morena Danieli, Giuseppe Riccardi Department of Information Engineering and Computer Science, University of Trento, Italy {firoj.alam, shammur.chowdhury, morena.danieli, giuseppe.riccardi}@unitn.it Abstract In this paper, we aim to investigate the coordination of interlocutors behavior in different emo- tional segments. Conversational coordination between the interlocutors is the tendency of speak- ers to predict and adjust each other accordingly on an ongoing conversation. In order to find such a coordination, we investigated 1) lexical similarities between the speakers in each emotional seg- ments, 2) correlation between the interlocutors using psycholinguistic features, such as linguistic styles, psychological process, personal concerns among others, and 3) relation of interlocutors turn-taking behaviors such as competitiveness. To study the degree of coordination in different emotional segments, we conducted our experiments using real dyadic conversations collected from call centers in which agent’s emotional state include empathy and customer’s emotional states include anger and frustration. Our findings suggest that the most coordination occurs be- tween the interlocutors inside anger segments, where as, a little coordination was observed when the agent was empathic, even though an increase in the amount of non-competitive overlaps was observed. We found no significant difference between anger and frustration segment in terms of turn-taking behaviors. However, the length of pause significantly decreases in the preceding segment of anger where as it increases in the preceding segment of frustration. 1 Introduction Behavioral and social signal processing are emerging interdisciplinary areas of research, which combine social science, psychology, and computer science. The aim of the research is to design computational models for processing human behavioral aspects, which can facilitate different domain experts while counseling, consulting and (or) providing services (Narayanan and Georgiou, 2013; Vinciarelli et al., 2009; Pantic et al., 2011; Vinciarelli et al., 2012; Stepanov et al., 2015). The idea is to analyze different overt and covert behavioral signals during social interactions and label them with some short and long term functional aspects (i.e., states and traits) in order to quantitatively measure them. The functional aspects include empathy, politeness, agreement, engagement, uncertainty, competitiveness and other typical, atypical, distressed and affective social behaviors. Using these short and long term states and traits, one can design an informative behavioral profile of an individual from the daily-life interactions. The measured behavioral profile can help to predict the next behavioral outcome/consequence and/or actions of an individual. This kind of behavioral profile can help domain experts in different application scenarios such as call center, health-care and teacher-student interactions. In the field of social and psychological science, researchers have been trying to understand these functional aspects for a very long time, however, very recently there are attempts to design automatic computational models for real-world applications. Designing such automatic systems for measuring these behavioral and social functional aspects is still infancy due to many different challenges. One of the important challenges …

Content truncated to 500 words