Contextual Emotional Classifier: An Advanced AI-Powered Emotional Health Ecosystem for Women Utilizing Edge Devices

Emotion, a complex interplay of feelings and thoughts, represents an individual's mental state and is a crucial semantic component in identifying various emotions. Contemporary digital wearable devices, available in diverse forms, are designed to gather emotional data, primarily focusing on monitoring emotions correlated with physical fitness. Digital assets such as voice assistant devices, accessories, smartwatches, and smartphones, typically used by women, are classified as edge devices. These devices form a connected ecosystem, designed to understand women's emotional states in relation to their behavior. In this work, we propose a Contextual Emotional Classifier (CEC) model that leverages AI learning on edge devices to train on these emotional data. The CEC model gathers emotional data from all AI edge devices, correlating and coordinating the information through contextual computation. This model analyzes text and voice data to propose solutions to emotional thoughts. Classification metrics are used to calculate various epochs of emotional audio data from edge assistant devices. The average weighted accuracy for audio is found to be 1080, while the text accuracy stands at 4000. The output of the contextual computation offers real-time emotional control alerts for dynamic mood swings, warnings about upcoming unplanned activities, and task management. These alerts can activate music, videos, or provide mentoring guidance. The system enhances privacy, security, latency, and avoidance of false data through serverless/cloudless data transfer combined with AI learning. This represents a significant advancement in the development of an emotional health ecosystem for women.