Header

UZH-Logo

Maintenance Infos

Towards the usefulness of user-generated content to understand traffic events


Das, Rahul Deb; Purves, Ross S (2018). Towards the usefulness of user-generated content to understand traffic events. In: GIScience 2018: 10th International Conference on Geographic Information Science, Melbourne (Australia), 28 August 2018 - 31 August 2018, 25:1-25:7.

Abstract

This paper explores the usefulness of Twitter data to detect traffic events and their geographical locations in India through machine learning and NLP. We develop a classification module that can identify tweets relevant for traffic authorities with 0.80 recall accuracy using a Naive Bayes classifier. The proposed model also handles vernacular geographical aspects while retrieving place information from unstructured texts using a multi-layered georeferencing module. This work shows Mumbai has a wide spread use of Twitter for traffic information dissemination with substantial geographical information contributed by the users.

Abstract

This paper explores the usefulness of Twitter data to detect traffic events and their geographical locations in India through machine learning and NLP. We develop a classification module that can identify tweets relevant for traffic authorities with 0.80 recall accuracy using a Naive Bayes classifier. The proposed model also handles vernacular geographical aspects while retrieving place information from unstructured texts using a multi-layered georeferencing module. This work shows Mumbai has a wide spread use of Twitter for traffic information dissemination with substantial geographical information contributed by the users.

Statistics

Altmetrics

Downloads

3 downloads since deposited on 18 Jan 2019
3 downloads since 12 months
Detailed statistics

Additional indexing

Item Type:Conference or Workshop Item (Paper), not_refereed, original work
Communities & Collections:07 Faculty of Science > Institute of Geography
Dewey Decimal Classification:910 Geography & travel
Language:English
Event End Date:31 August 2018
Deposited On:18 Jan 2019 09:07
Last Modified:19 Jan 2019 08:33
Publisher:Schloss Dagstuhl -- Leibniz-Zentrum fuer Informatik
Series Name:LIPIcs : Leibniz International Proceedings in Informatics
Number:114
ISSN:1868-8969
ISBN:978-3-95977-083-5
OA Status:Gold
Free access at:Publisher DOI. An embargo period may apply.
Official URL:http://drops.dagstuhl.de/opus/volltexte/2018/9353/pdf/LIPIcs-GISCIENCE-2018-25.pdf

Download

Download PDF  'Towards the usefulness of user-generated content to understand traffic events'.
Preview
Content: Published Version
Language: English
Filetype: PDF
Size: 356kB
Licence: Creative Commons: Attribution 3.0 Unported (CC BY 3.0)