2021/06/02 by George Boateng, Boateng, George, Peter Hilpert +7
Psychology · #Computation and Language (cs.CL) #FOS: Computer and information sciences #J.4 #Mental Health via Writing
paper · pdf · doi:10.48550/arxiv.2106.01526
openalex publication_date 2021/06/02 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
How romantic partners interact with each other during a conflict influences\nhow they feel at the end of the interaction and is predictive of whether the\npartners stay together in the long term. Hence understanding the emotions of\neach partner is important. Yet current approaches that are used include\nself-reports which are burdensome and hence limit the frequency of this data\ncollection. Automatic emotion prediction could address this challenge. Insights\nfrom psychology research indicate that partners' behaviors influence each\nother's emotions in conflict interaction and hence, the behavior of both\npartners could be considered to better predict each partner's emotion. However,\nit is yet to be investigated how doing so compares to only using each partner's\nown behavior in terms of emotion prediction performance. In this work, we used\nBERT to extract linguistic features (i.e., what partners said) and openSMILE to\nextract paralinguistic features (i.e., how they said it) from a data set of 368\nGerman-speaking Swiss couples (N = 736 individuals) who were videotaped during\nan 8-minutes conflict interaction in the laboratory. Based on those features,\nwe trained machine learning models to predict if partners feel positive or\nnegative after the conflict interaction. Our results show that including the\nbehavior of the other partner improves the prediction performance. Furthermore,\nfor men, considering how their female partners spoke is most important and for\nwomen considering what their male partner said is most important in getting\nbetter prediction performance. This work is a step towards automatically\nrecognizing each partners' emotion based on the behavior of both, which would\nenable a better understanding of couples in research, therapy, and the real\nworld.\n