[3P6-OS-9b] OS-9

[3P6-OS-9b-01] Estimating Parameters in Cognitive Architectures Related to Emotional States

〇Kohei Shimbori¹, Masaru Shirasuna¹, Junya Morita¹ (1. Shizuoka University)

Keywords:Emotion, Cognitive Model, ACT-R

Recent attempts have been made to estimate the parameters of cognitive models based on human behavioral data. Such efforts have practical significance in predicting performance on various cognitive tasks through personalized models and are also useful for understanding human internal mechanisms. However, previous studies have primarily focused on estimating parameters, which exhibit little intra-individual variability, while estimating parameters linked to emotions has been rare. This study estimated memory-related parameters in the cognitive architecture ACT-R based on performance in a simple memory task involving mood inductions. Additionally, we examined correlations between the estimated parameters and collected emotional states, as well as anxiety traits that influence emotions. The results showed correlations between state anxiety and parameters related to "the matching penalty for recalling unintentional item," as well as "the time cost for memory retrieval." The former also correlated with subjective positive emotions. Further refinement will help specify emotion-related parameters in cognitive architectures.