Predicting vehicle recalls with user-generated contents: A text mining approach

Xuan Zhang, Shuo Niu, Da Zhang, G. Alan Wang, Weiguo Fan

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Vehicle safety issues and component defects result in property losses and fatalities. Our study proposes a new method to predict vehicle recalls based on user generated contents in online discussion forums. Vehicle defects can cause bodily injuries and sometimes deadly consequences. However, vehicle recalls will not be issued until damage has occurred. Online vehicle discussion forums usually contain traits of vehicle defects long before manufacturers and government agencies take investigative actions. We find overlapping components in user generated contents and official recall notices. Our proposed recall prediction method can correctly predict vehicle recalls once in every two recall events. It is our hope that our proposed technique can be used to monitor online vehicle discussion forums and prompt the manufacturers and government agencies to issue recalls before catastrophic accidents occur. Our research has significant practical implications to vehicle and transportation safety.

Original languageEnglish (US)
Title of host publicationIntelligence and Security Informatics - Pacific Asia Workshop, PAISI 2015, Proceedings
EditorsHsinchun Chen, Michael Chau, G. Alan Wang
PublisherSpringer
Pages41-50
Number of pages10
ISBN (Print)9783319184548
DOIs
StatePublished - Jan 1 2015
Externally publishedYes
Event10th Pacific Asia Workshop on Intelligence and Security Informatics, PAISI 2015 in Conjunction with Pacific-Asia Conference on Knowledge Discovery and Data Mining , PAKDD 2015 - Ho Chi Minh City, Viet Nam
Duration: May 19 2015May 19 2015

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9074
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference10th Pacific Asia Workshop on Intelligence and Security Informatics, PAISI 2015 in Conjunction with Pacific-Asia Conference on Knowledge Discovery and Data Mining , PAKDD 2015
Country/TerritoryViet Nam
CityHo Chi Minh City
Period5/19/155/19/15

Keywords

  • Text classification
  • Transportation safety
  • User generated contents
  • Vehicle defect

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'Predicting vehicle recalls with user-generated contents: A text mining approach'. Together they form a unique fingerprint.

Cite this