Application of Cognitive-Computational Models Based on Artificial Intelligence in Predicting Human Emotional Fluctuations: A Step Towards the Development of Human-Machine Co-Regulatory Systems

Application of Cognitive-Computational Models Based on Artificial Intelligence in Predicting Human Emotional Fluctuations: A Step Towards the Development of Human-Machine Co-Regulatory Systems

Nazanin Haghighat Bayan1

1) PhD in Clinical Psychology, Islamic Azad University, Roudehen Branch. (Assistant Professor and Faculty Member, Toloumehr University).

IOI Code : AAHD-ACZACF
Publication : National Conference on Information Technology, Nanotechnology, Artificial Intelligence and Technological Futures Studies(itnaf.ir)
Abstract :
    With the expansion of artificial intelligence technologies and the increase in human interaction with intelligent systems, attention to the ability of machines to understand and predict human emotional states has become an important research topic. Emotions play a decisive role in decision-making, learning, social interactions, and mental health of individuals, and their sudden changes can have wide-ranging individual and social consequences. In recent years, approaches based on cognitive science and artificial intelligence have attempted to move beyond the detection of current emotions and provide the possibility of predicting emotional fluctuations before they manifest. In this regard, the present study aims to investigate the application of cognitive-computational models based on artificial intelligence in predicting human emotional fluctuations and providing a framework for the development of human-machine co-regulation systems. This study is of a review-analytical type and was conducted using library studies and a systematic review of recent research in the fields of artificial intelligence, cognitive science, computational neuroscience, and emotion analysis. In this study, the theoretical foundations of emotion, cognitive models of the mind, machine learning and deep learning technologies, multimodal data analysis methods and their application in emotion prediction were examined. Also, by integrating theoretical findings, a conceptual model for intelligent human-machine co-regulation based on emotion prediction was designed and explained. The findings showed that the simultaneous use of multimodal data including behavioral, speech, textual, facial and biological signals can significantly increase the accuracy of predicting emotional changes. It was also found that cognitive-computational models, by representing components such as attention, memory and cognitive state, are able to simulate mental processes related to emotion formation and provide a suitable platform for predicting future emotional states. Accordingly, the proposed research model was composed of six main layers including multimodal data, feature processing, cognitive model of the mind, emotion prediction engine, co-regulation system and adaptive learning.

Keywords : artificial intelligence cognitive computational models emotion prediction multimodal data human machine co ordination