Home // GEOProcessing 2023, The Fifteenth International Conference on Advanced Geographic Information Systems, Applications, and Services // View article


Using a Social Network for Road Accidents Detection, Geolocation and Notification - A Machine Learning Approach

Authors:
Samuel Pereira de Vasconcelos
Cláudio De Souza Baptista
Hugo Feitosa de Figueiredo

Keywords: Road accidents; Machine learning, Natural Language Processing; Geoprocessing; Mobile Computing; Social media.

Abstract:
The road system is the main means of transport used in Brazil. Traffic accidents are quite common in this mode of transport, incurring one of the biggest causes of death in the country. Profiles on social networks of the Federal Highway Police (PRF) and other sources of information, contribute to alert drivers as quickly as possible about road accidents that have occurred, in order to prevent other accidents from occurring. Also, such information can be used by drivers about possible delays, or even deviations in their paths. However, accessing such information via text while driving is illegal and further increases the risk of accidents. Therefore, this paper addresses a study about reliable posts in social networks, in particular Twitter, to create a supervised classification model, which is capable of classifying tweets about the occurrence or not of accidents. The results include the best induction model obtained for classifying tweets, among several analyzed, as well as the construction of a mobile application that can notify through audio drivers about accidents reported on their way, in real time.

Pages: 86 to 91

Copyright: Copyright (c) IARIA, 2023

Publication date: April 24, 2023

Published in: conference

ISSN: 2308-393X

ISBN: 978-1-68558-079-7

Location: Venice, Italy

Dates: from April 24, 2023 to April 28, 2023