FUZZY INFERENCE SYSTEM FORMULATION FOR DISASTER CONTROL SYSTEMS: A CLUSTERING MECHANISM

Abstract

In the digital age, social media platforms have become vital sources for real-time information, especially during natural and human-made disasters. Emojis, often used to convey emotions and situations quickly, serve as powerful yet underutilized indicators of public sentiment and emerging crises. This research proposes a novel disaster control system framework that leverages social media emojis through clustering algorithms and a fuzzy inference system (FIS). The methodology involves extracting emoji data from platforms like Twitter, classifying it based on disaster relevance, and grouping it using fuzzy C-means clustering to identify patterns corresponding to specific disaster types. A fuzzy rule-based inference engine is then used to assess the severity and type of disaster based on input parameters such as emoji density, type, sentiment, and geolocation. The system was validated using real-time data collected during major disaster events, and it demonstrated improved detection speed and accuracy compared to traditional text-only systems. This study introduces a scalable and emotion-aware model for enhancing situational awareness and response effectiveness in disaster management. The use of fuzzy logic enables better handling of imprecise information, while clustering aids in classifying different types of disasters for more accurate assessment and response. Together, these techniques contribute to a more efficient and intelligent disaster management solution.