Chat Erasure Detection for Emotional Well, Being

In the digital communication era, emotional expression often occurs through typed messages that are later erased and never sent. These deleted drafts can reflect moments of emotional vulnerability, hesitation, or distress, yet they often go unnoticed. This research introduces a real time emotional well being monitoring system Chat Erasure Detection that leverages keystroke dynamics and natural language processing (NLP) to detect and analyze unsent emotional messages. The system captures typing behavior, such as speed, rhythm, pauses, and deletion patterns, identifying messages that are predominantly erased. When such instances occur, the system applies sentiment analysis to determine the emotional tone of the deleted content. If emotional distress is detected, a secure alert is sent to a guardian interface for potential support intervention. By combining real time behavioral biometrics with emotional context analysis, the proposed system offers a novel, privacy conscious method of identifying emotional cues in unsaid expressions. This approach contributes to mental health support by addressing the silent nature of digital hesitation, enabling timely and compassionate responses in sensitive scenarios. The system architecture is designed with modular components, focusing on accuracy, efficiency, and ethical data handlingIntroduction