An Improved Method for Assessing Psycho-Emotional State Based on Analysis of Body's Functional Systems

Article examines an algorithm for studying the possibilities of analyzing and processing human body signals in response to viewing media content in an automatic mode. This process aims to identify the features of human interaction with various types of information presented in a multimedia format. The emotional state of the content is automatically assessed through text and image analysis using recurrent neural networks. Such networks allow for the analysis of sequential data and the identification of complex patterns that may be associated with the emotional characteristics of the content. The evaluation of body response signals is performed automatically by applying a matrix singular value decomposition algorithm to electrocardiogram signals. This method makes it possible to extract key data components that can be interpreted as physiological markers of emotional reactions. It is expected that the application of this approach will enable the identification of emotionally significant responses, as well as the assessment of their depth within the analyzed content. The prospects lie in the use of such algorithms in data mining systems and big data processing algorithms. This could contribute to the development of personalized approaches to creating media content, as well as improve technologies related to analyzing user behavior and their emotional reactions.