Disaster Preparedness and Response Education System for Schools and Colleges

Abstract

This research presents a sophisticated AI-driven digital platform designed to overcome the inherent limitations of traditional disaster preparedness training- specifically its lack of interactivity, personalisation, and scalability. By leveraging advanced Natural Language Processing (NLP) techniques, including transformer-based and extractive summarization, the system intelligently synthesis diverse data sources such as PDFs, text files and YouTube transcripts into cohesive, drill-specific summaries. These summaries serves as the foundation for an automated assessment engine that generates unified quizzes, which administrators can further customise with manual questions to ensure comprehensive evaluation. To maximize user engagement and learning retention, the platform features a centralized dashboard where learners can access resources, track performance, and earn badges based on predefined achievement thresholds. A critical component of the system is the “Disaster Assistant”, an intelligent conversational agent that provides real- time, content-aware responses to queries regarding cyclone preparedness, emergency response, and risk mitigation. Built with a modular SQLite3 database schema for secure administrative oversight and performance analytics, the platform demonstrates that the integration of AI-driven summarization, automated scoring and realtime assistance significantly strengthens disaster readiness and fosters a more adaptive, accessible educational experience.