An efficient predictive alert system for natural disasters in tropical region using GIS and social media data using sentiment analysis

S. Mohanarangan (), L. Suganthi (), G. Shoba () and D. Karthika ()
Additional contact information
S. Mohanarangan: Arunai Engineering College
L. Suganthi: Kamban College of Arts and Science for Women
G. Shoba: Arunai Engineering College
D. Karthika: Arunai Engineering College

Natural Hazards: Journal of the International Society for the Prevention and Mitigation of Natural Hazards, 2025, vol. 121, issue 12, No 5, 13985-14007

Abstract: Abstract Natural disasters such as earthquakes, floods, Thunderstorms, cyclones, and rainfall pose significant risks to human life and infrastructure. Traditional prediction methods often rely on physical sensors, meteorological data, and historical patterns, which may not always provide timely or accurate warnings. To investigate the feasibility of using sentiment analysis on GIS-based data in social media posts like Twitter, Facebook, Google News and other textual data to predict natural disasters in tropical regions. By analyzing the sentiment of communications related to weather conditions, emergency alerts, and public reactions, the study seeks to identify patterns and correlations that may serve as early indicators of impending disasters. The ultimate goal is to enhance early warning systems and improve disaster preparedness by integrating sentiment analysis with traditional prediction models. Sentiment keyword graph filtering Technique (SKG) is used to remove neural text keyword-based filtering for disaster-related terms and graph-based filtering for central node. A deep neural network (DNN) is used to classify and analyze sentiment from social media posts. Adam optimization algorithm (AOA) is used to optimise model parameters to minimize the loss function, improving prediction accuracy. The evaluation is conducted on Matlab or Python based on parameters such as Adam accelerates convergence, reducing training time and computational resources and for natural disaster prediction are disaster occurrence risk factors, disaster probability, event timelines, geospatial alerts, early warning alerts, and confidence intervals. The result shows that the Proposed Model demonstrates superior performance with consistently high accuracy rates, peaking at 98.8, implemented using Python software. The future scope of this research is vast and promising, with potential expansions including integration with IoT devices, multilingual support, real-time processing, and more, to enhance the system’s capabilities in predicting and alerting tropical region natural disasters through GIS-based sentiment analysis of social media data.

Keywords: Adam optimization algorithm; Sentiment keyword graph filtering technique; Deep neural network; Natural disasters; Social media data; Tropical region (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s11069-025-07341-w Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:nathaz:v:121:y:2025:i:12:d:10.1007_s11069-025-07341-w

Ordering information: This journal article can be ordered from
http://www.springer.com/economics/journal/11069

DOI: 10.1007/s11069-025-07341-w

Access Statistics for this article

Natural Hazards: Journal of the International Society for the Prevention and Mitigation of Natural Hazards is currently edited by Thomas Glade, Tad S. Murty and Vladimír Schenk

More articles in Natural Hazards: Journal of the International Society for the Prevention and Mitigation of Natural Hazards from Springer, International Society for the Prevention and Mitigation of Natural Hazards
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().