Data Science and Intelligent Computing Techniques
LGBT Cyberbullying Detection in Thai Language Utilizing Transformers-Based Algorithms
Authors: Vajratiya Vajrobol, Nitisha Aggarwal, Unmesh Shukla, Sanjeev Singh, Geetika Jain Saxena and Amit Pundir
Publishing Date: 17-10-2023
ISBN: 978-81-955020-2-8
Abstract
Cyberbullying is a growing issue worldwide, particularly among minority groups such as the LGBT community. While research has shown that LGBT individuals are at a higher risk for experiencing cyberbullying, there is a lack of studies focusing on detecting cyberbullying directed towards this group in the context of the Thai language. This research aims to identify incidents of cyberbullying targeting the LGBT community in the Thai language on the social media platform, Twitter by using specific Thai derogatory keywords for data collection. In terms of model development for the detection of LGBTcyberbullying, deep learning-based algorithms, including CNN, LSTM, and transformer models have been applied to the Thai corpus to classify the cyberbullying messages. The results show that CNN performs better than other algorithms with 99.98% accuracy and 99.99% F1 score. The study aims to develop effective tools and strategies for detecting cyberbullying and ultimately contribute towards creating a better society that values diversity and inclusion of LGBT.
Keywords
Cyberbullying, LGBT, Thai language, Low-resource language
Cite as
Vajratiya Vajrobol, Nitisha Aggarwal, Unmesh Shukla, Sanjeev Singh, Geetika Jain Saxena and Amit Pundir, "LGBT Cyberbullying Detection in Thai Language Utilizing Transformers-Based Algorithms", In: Satyasai Jagannath Nanda and Rajendra Prasad Yadav (eds), Data Science and Intelligent Computing Techniques, SCRS, India, 2023, pp. 393-399. https://doi.org/10.56155/978-81-955020-2-8-35