Cyberbullying Detection Using Data Mining and Sentiment Analysis: Repository Building in Bruneian Context

Cyberbullying is a major concern especially for children and teens who own technology devices such as personal computers, mobile phones or tablets and have access to social media networks. Owning a technological device normally has positive impacts, especially in terms of educational knowledge where valuable information is just at one’s fingertips away. Unfortunately, these young individuals are potentially subjected to cyberbullying through social media networks that could have detrimental effects on their mental and emotional health. In Brunei context, words are often mixed with English, standard Malay, and Brunei Malay which are somehow unique in a way they are made, constructed or sounded. Brunei Malay is quite different from the standard Malay language or Bahasa Melayu in terms of spelling, pronunciation and vocabulary. This paper attempted to detect and analyse cyberbullying signals in Brunei Malay on one of the most popular online social media platforms, Twitter. Textual messages or tweets posted on Twitter were data mined based on seed words identified as ‘cyberbullying words’ in Brunei Malay and analysed through sentiment analysis. Words or lexical items used in detecting cyberbullying amongst Bruneians were identified and collected to develop a repository of cyberbullying lexical terms in Brunei context.