ResQTrack: AI-Driven Crisis Detector and Geo-Mapping

Abstract

Disaster monitoring plays a critical role in reducing the impact of natural disasters and man-made disasters, but we know how the traditional methods work. Though the information is significant, it gets delayed after the disaster has occurred. This doesn’t save lives and earnings. In contrast with that, we know that on social media, the information is shared faster than news channels and other government official websites; However, the information is unstructured, inconsistent, and difficult to analyse directly. The ResQTrack system is developed to address the challenge by integrating real-time data from the social media platform X (Twitter) in addition to the help of APIs such as USGS and GDACS, processing it to identify relevant events, and extracting location details even when precise coordinates are not available. The data is processed and analysed, and then it is displayed through an interactive dashboard and on a global map. Our system also provides analytical insights and severity distribution of the regions, and it also shows the graphical representation of the data and disaster frequency trends. ResQTrack converts the raw data into structured and meaningful analytics. ResQTrack improves situational awareness and supports faster decision-making. The system demonstrates how real-time data and simple analytical techniques can be effectively used.

Introduction

The text presents ResQTrack, an AI-based disaster monitoring and visualization system designed to provide fast, organized, and reliable information about disasters by combining social media data with official disaster information sources.

1. Background

Natural and human-made disasters such as earthquakes, floods, wildfires, cyclones, industrial accidents, conflicts, and wars can cause major damage within a short time. Traditional sources such as government websites and news channels may not provide information quickly enough during emergencies.

Social media, particularly X (formerly Twitter), can provide real-time information because people often post images, videos, and descriptions immediately after experiencing a disaster. However, social media data is usually unstructured, noisy, duplicated, and sometimes unreliable, making it difficult to analyze manually.

ResQTrack addresses this problem by converting raw disaster-related information into structured, useful, and visual information for emergency response and decision-making.

2. Main Data Sources

ResQTrack combines information from multiple sources:

• X/Twitter API: Real-time user-generated disaster reports, posts, images, and videos.

• USGS API: Earthquake and seismic-event information.

• GDACS: Global disaster events and alerts, including floods and cyclones.

Using multiple sources provides broader coverage and reduces dependence on a single information channel.

3. Proposed System

ResQTrack follows a five-stage processing pipeline:

1. Data Collection – Collects disaster information from APIs and social media.

2. Data Preprocessing – Removes irrelevant content, duplicate records, noise, and unnecessary information.

3. Data Analysis – Identifies disaster types and determines their severity.

4. Location Extraction – Extracts geographic information and converts it into coordinates, even when location information is incomplete.

5. Data Visualization – Displays the processed information on an interactive dashboard.

The system classifies events into categories such as natural disasters and man-made disasters. Severity is categorized into high, medium, and low levels.

4. System Implementation

The system uses several technologies:

• Node.js and Express.js: Backend processing and API integration.

• MongoDB: Storage of disaster information such as event type, magnitude, location, time, source, and severity.

• HTML, CSS, and JavaScript: Frontend development.

• Interactive maps: Visualization of disaster locations and severity.

Each disaster is displayed as a map marker, with different colors and sizes representing different characteristics such as severity. Users can select markers to view additional event information.

The system also performs event grouping, combining related events based on location and time to reduce duplicate information and improve the clarity of the dashboard.

5. Handling Social Media Reliability

Because information from social media may not always be trustworthy, ResQTrack performs additional credibility checks. These include considering factors such as:

• User activity duration.

• Number of disaster reports made by a user.

• Consistency of user-generated reports.

This helps the system distinguish potentially useful reports from less reliable information.

6. Results

The system was tested using real-time disaster data streams. The results showed that ResQTrack could:

• Collect disaster information from multiple sources.

• Automatically identify and categorize disaster events.

• Map disaster locations.

• Display disaster type, severity, time, and source.

• Group related events and reduce duplication.

• Present disaster information through maps, charts, and dashboards.

• Identify regions experiencing frequent disaster events.

The visualizations also made it easier to identify areas with frequent seismic activity or higher concentrations of disaster events.

Overall Conclusion

ResQTrack is a real-time disaster monitoring and situational-awareness platform that combines social media and official disaster data with AI-based processing, geographic analysis, and interactive visualization. Its main purpose is to transform large amounts of raw, unstructured disaster information into concise and understandable information.

By integrating X, USGS, and GDACS, the system can provide broader and faster disaster information than relying on a single source. Its combination of data preprocessing, disaster classification, severity assessment, location extraction, event grouping, and dashboard visualization can support authorities and the public in understanding disaster situations and making faster decisions during emergencies.

Conclusion

ResQTrack offers real-time disaster monitoring by showing raw data, computers analysing that data, and graphical representations of that analysis. This helps determine the location, type, and severity of discrete aspects of a disaster. Users can see disaster events by looking through structured data rather than searching unstructured data. Turning disperse and unstructured data into structured data and visual representations is the primary goal of ResQTrack. One primary visualization component is an interactive map. Along with that, the visual representations help show what regions are affected by what kind of disasters. These components helps achieve the goal of the system. The core components and systems of ResQTrack can be altered and improved to provide different types of alerts (SMS for example), incorporate different artificial intelligence and machine learning (analysis systems), and gain data from different systems. Ultimately, ResQTrack is a system that has potential to integrate different visualized and structured data to help with disaster monitoring and situational awareness. Overall, ResQTrack demonstrates the potential of integrating real-time disaster data with structured processing and visualization for improved situational awareness. With future enhancements such as advanced AI analysis and real-time alert mechanisms, the system can be further developed into a more comprehensive disaster-monitoring platform.

References

[1] Awajan, I., Mohamad, M., and Al-Quran, A., Sentiment Analysis Technique and Neutrosophic Set Theory for Mining and Ranking Big Data from Online Reviews, 2021. [2] GDACS, Global Disaster Alert and Coordination System, Available: https://www.gdacs.org [3] Ilias, L., Kazelidis, I. M., and Askounis, D., Multimodal Detection of Bots on X (Twitter) Using Transformers, IEEE Access, 2024. [4] Leaflet, Leaflet.js Interactive Maps Documentation, Available: https://leafletjs.com [5] Momin, K. A., Sadri, A.M., Olofsson, K., Muraleetharan, K. K., and Hugh, Information Switching Patterns of Risk Communication in Social Media During Disasters, 2025. [6] MongoDB Inc., MongoDB Documentation, Available: https://www.mongodb.com [7] Node.js Foundation, Node.js Documentation, Available: https://nodejs.org [8] OpenStreetMap Contributors, OpenStreetMap Data, Available: https://www.openstreetmap.org [9] Sufi, F. K., and Khalil, I., Automated Disaster Monitoring from Social Media Posts Using AI-Based Location Intelligence and Sentiment Analysis, IEEE, 2024. [10] USGS, United States Geological Survey Earthquake API, Available: https://earthquake.usgs.gov [11] Vwarakanath, L., Kamsin, A., Rasheed, R. A., Anandhan, A., and Shuib, L., Automated Machine Learning Approaches for Emergency Response and Coordination via Social Media in the Aftermath of a Disaster: A Review, 2021

Copyright

Copyright © 2026 B. Vamshi, J. Ashritha, M. Roshitha, M. Lasya. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.