Assessment of Human State Based on Data from Keyboard Activity

In this study, we present a model developed for assessing human emotional states using data obtained from interactions with a computer keyboard. The model incorporates a feedback mechanism, which significantly enhances the accuracy of emotional state predictions. Through systematic analysis of the dynamics of keyboard usage by the user and monitoring individual environmental factors, our approach not only predicts the user's state but also adapts to changes in user behavior. The efficacy of the model was validated through conducted experiments. The model's ability to enhance adaptability to humans and the personalization of user interfaces improves human interaction with intelligent systems. The developed system demonstrates directions for improvements in human-machine interfaces, providing a foundation for future integration into more complex systems aimed at comprehensive assessment of psychophysiological states.