Publication: Automatic discovery of abusive thai language usages in social networks
Issued Date
2017-01-01
Resource Type
ISSN
16113349
03029743
03029743
Other identifier(s)
2-s2.0-85034109547
Rights
Mahidol University
Rights Holder(s)
SCOPUS
Bibliographic Citation
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol.10647 LNCS, (2017), 267-278
Suggested Citation
Suppawong Tuarob, Jarernsri L. Mitrpanont Automatic discovery of abusive thai language usages in social networks. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol.10647 LNCS, (2017), 267-278. doi:10.1007/978-3-319-70232-2_23 Retrieved from: https://repository.li.mahidol.ac.th/handle/20.500.14594/42341
Research Projects
Organizational Units
Authors
Journal Issue
Thesis
Title
Automatic discovery of abusive thai language usages in social networks
Author(s)
Other Contributor(s)
Abstract
© 2017, Springer International Publishing AG. Social networks have become a standard means of communication that allows a massive amount of users to interact and consume information anywhere and anytime. In Thailand, millions of users have access to social networks, a majority of which include young children. The colloquial nature of social media inherently encourages certain expressions of language that do not conform to the standard, some of which may be considered abusive and offensive. Such ill-mannered language fashion has become increasingly used by a large number of Thai social media users. If these abusive languages are exposed to adolescents without proper guidance, they could compulsorily develop a familiar attitude towards such language styles. To address the issue, we present a set of algorithms based on machine learning, that automatically detect abusive Thai language in social networks. Our best results yield 86% f-measure (88.73% precision and 83.53% recall).