A Dynamic Memory-Based Adaptive Model for Individual Self-Efficacy Improvement

Abstract

This paper presents an adaptive agent with dynamic memory AI system designed to address the challenges faced by elderly people, such as declines in physical health, cognitive function, and social interaction. The system comprised of a Language Model, Advanced Memory Manager, Emotion Detection Module, and Feedback Model, processes elderly users' diary entries to generate empathetic responses, while learning and adapting to their behavioral patterns and emotional states. Through human-like interaction that is cognizant of users' emotions and history, and by integrating user feedback, the system aims to stimulate behaviors that enhance self-efficacy and improve overall well-being among the elderly.