Identifying and Grouping Crisis-Related Social Media Posts for Disaster Response

Mahalakshmi, S and Bagavathi Lakshmi, R (2026) Identifying and Grouping Crisis-Related Social Media Posts for Disaster Response. In: 2026 International Conference on Electronics and Renewable Systems (ICEARS), Tuticorin, India.

[thumbnail of Identifying and Grouping Crisis-Related Social Media Posts for Disaster Response.pdf] Text
Identifying and Grouping Crisis-Related Social Media Posts for Disaster Response.pdf

Download (1MB)

Abstract

Every year, within a specific timeframe, human society
confronts challenges arising from natural calamities. The ultimate
result of a natural crisis, such as flooding, hurricanes, earthquakes,
and tsunamis, is the disruption of communication with the outside
world. This communication disruption hinders individuals in the
affected regions from accessing essential services, including medical
care, sustenance, and potable water, potentially resulting in
fatalities. Responders can also face difficulties in gathering
information about the affected areas, which may be inaccessible
after a crisis has occurred. Once the situation arose, the first two
barriers were the breakdown of the power supply and
communication to the outside world. Even though mobile
communication is available, information is often obtained orally for
a particular person, and this can lead to failure due to incorrect
communication with the wrong number or the unavailability of the
receiving party. Due to the introduction of smartphones and
Online Social Networks (OSNs), information has been disseminated
faster with the help of posting over available networks with low
signal strength. Many people will see the posting and become aware
of the situation in a particular area. This paper proposes a
framework for identifying crisis-related posts with diverse topics
and grouping them to disseminate information to responders,
thereby facilitating a faster recovery process. A virtual group has
been formed, comprising both requesters and responders of the
same topics, and will exist until the last requester receives a solution
for their need. The performance of the pr

Item Type: Conference or Workshop Item (Paper)
Subjects: Computer Science Engineering > Data Modeling
Domains: Computer Science
Depositing User: Mr IR Admin
Date Deposited: 03 Sep 2026 10:26
Last Modified: 03 Sep 2026 10:26
URI: https://ir.vistas.ac.in/id/eprint/22528

Actions (login required)

View Item
View Item