Hate Speech Detection in Roman Urdu

被引:35
作者
Khan, Muhammad Moin [1 ]
Shahzad, Khurram [1 ]
Malik, Muhammad Kamran [1 ]
机构
[1] Univ Punjab, Punjab Univ Coll Informat Technol, Lahore, Pakistan
关键词
Low-resource languages; Hate speech detection; Roman Urdu; South Asian Languages;
D O I
10.1145/3414524
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Hate speech is a specific type of controversial content that is widely legislated as a crime that must be identified and blocked. However, due to the sheer volume and velocity of the Twitter data stream, hate speech detection cannot be performed manually. To address this issue, several studies have been conducted for hate speech detection in European languages, whereas little attention has been paid to low-resource South Asian languages, making the social media vulnerable for millions of users. In particular, to the best of our knowledge, no study has been conducted for hate speech detection in Roman Urdu text, which is widely used in the sub-continent. In this study, we have scrapped more than 90,000 tweets and manually parsed them to identify 5,000 Roman Urdu tweets. Subsequently, we have employed an iterative approach to develop guidelines and used them for generating the Hate Speech Roman Urdu 2020 corpus. The tweets in the this corpus are classified at three levels: Neutral-Hostile, Simple-Complex, and Offensive-Hate speech. As another contribution, we have used five supervised learning techniques, including a deep learning technique, to evaluate and compare their effectiveness for hate speech detection. The results show that Logistic Regression outperformed all other techniques, including deep learning techniques for the two levels of classification, by achieved an F1 score of 0.906 for distinguishing between Neutral-Hostile tweets, and 0.756 for distinguishing between Offensive-Hate speech tweets.
引用
收藏
页数:19
相关论文
共 32 条
[1]  
[Anonymous], 2017, P RANLP
[2]  
[Anonymous], 2016, P 1 WORKSH NLP COMP, P138, DOI [DOI 10.18653/V1/W16-5618, 10.18653/v1/W16-5618]
[3]  
[Anonymous], 2005, PHOTONIC DEVICES ALG
[4]  
Artstein R., 2017, Handbook of linguistic annotation, DOI 10.1007/978-94-024-0881-211
[5]  
Ashraf J, 2010, LECT NOTES COMPUT SC, V6177, P140, DOI 10.1007/978-3-642-13881-2_14
[6]   Deep Learning for Hate Speech Detection in Tweets [J].
Badjatiya, Pinkesh ;
Gupta, Shashank ;
Gupta, Manish ;
Varma, Vasudeva .
WWW'17 COMPANION: PROCEEDINGS OF THE 26TH INTERNATIONAL CONFERENCE ON WORLD WIDE WEB, 2017, :759-760
[7]  
Clement J., 2020, Statista
[8]   Urdu language processing: a survey [J].
Daud, Ali ;
Khan, Wahab ;
Che, Dunren .
ARTIFICIAL INTELLIGENCE REVIEW, 2017, 47 (03) :279-311
[9]  
Daud M., 2015, arXiv preprint arXiv:150101386
[10]  
Davidson T., 2017, P INT AAAI C WEB SOC, P512