Predicting users’ emotional states during interaction is a long-standing goal of affective computing. However, traditional methods based on sensory data alone fall short due to the interplay between users’ latent cognitive states and emotional responses. To address this, we introduce a computational cognitive model that simulates emotion as a continuous process, rather than a static state, during interactive episodes. This model integrates cognitive-emotional appraisal mechanisms with computational rationality, utilizing value predictions from reinforcement learning. Experiments with human participants demonstrate the model’s ability to predict and explain the emergence of emotions such as happiness, boredom, and irritation during interactions. Our approach opens the possibility of designing interactive systems that adapt to users’ emotional states, thereby improving user experience and engagement. This work also deepens our understanding of the potential of modeling the relationship between reward processing, reinforcement learning, goal-directed behavior, and appraisal.

Large language models, in particular generative pre-trained transformers (GPTs), show impressive results on a wide variety of language-related tasks. In this paper, we explore ChatGPT’s zero-shot ability to perform affective computing tasks using prompting alone. We show that ChatGPT a) performs meaningful sentiment analysis in the Valence, Arousal and Dominance dimensions, b) has meaningful emotion representations in terms of emotion categories and these affective dimensions, and c) can perform basic appraisal-based emotion elicitation of situations based on a prompt-based computational implementation of the OCC appraisal model. These findings are highly relevant: First, they show that the ability to solve complex affect processing tasks emerges from language-based token prediction trained on extensive data sets. Second, they show the potential of large language models for simulating, processing and analyzing human emotions, which has important implications for various applications such as sentiment analysis, socially interactive agents, and social robotics.

Self-awareness is a critical factor in social interaction. Teachers being aware of their own emotions and thoughts during class may enable reflection and behavioral change. While inducing self-awareness through mirrors or video is common in face-to-face training, it has been scarcely examined in digital training with virtual avatars. This paper examines the relationship between avatar visual similarity and inducing self-awareness in digital training environments. We developed a theory-based methodology to reliably manipulate perceptually relevant facial features of digital avatars based on human-human identification and emotional predisposition. Manipulating these features allows to create personalized versions of digital avatars with varying degrees of visual similarity.

In the last decade, empirical sciences have faced a tremendous change in the way of conducting research. As a broad interdisciplinary field, research in Affective Computing often employs empirical user studies. The current paper analyzes research practices in Affective Computing and deduces recommendations for improving the quality of methods and reporting. We extracted a total of k = 65 empirical studies from the two most recent International Conferences on Affective Computing & Intelligent Interaction (ACII) ’17 and ’19. Three raters summarized characteristics of studies (e.g., number of experimental studies) and how much methodological (e.g., participant characteristics) and statistical information (e.g., degrees of freedom) were missing. Also, we conducted a p-curve analysis to test the overall evidential value of findings. Results showed that 1. in at least half of the studies, one important information about statistical results was missing, and 2. those k = 31 studies that had reported all necessary information to be included into the p-curve showed evidential value. In general, all criteria were never met in one single study. We provide concrete recommendations on how to implement open research practices for empirical studies in Affective Computing.

This paper presents a study that examines whether social agents can elicit the social emotion shame as humans do. For that, we use job interviews, which are highly evaluative situations per se. We vary the interview style (shame-eliciting vs. neutral) and the job interviewer (human vs. social agent). Our dependent variables include observational data regarding the social signals of shame and shame regulation as well as self-assessment questionnaires regarding the felt uneasiness and discomfort in the situation. Our results indicate that social agents can elicit shame to the same amount as humans. This gives insights about the impact of social agents on users and the emotional connection between them.