科研

Automatic cyberbullying detection is an expert system of growing importance in the fields of natural language processing and machine learning. Conventional monolingual (English) text-based can provide support for an early detection of potential cyberbullying incidents. However, in Hong Kong, both Chinese and English languages are the medium of communication used in social media platforms. Such a bilingual language environment makes the recently developed monolingual early detection system inapplicable. This proposed project will attempt to develop a new expert system for early detection of cyberbullying incidents in bilingual, specifically Chinese-English code-mixed language. The new expert system will address the cyberbullying problems in Hong Kong and nearby regions (such as Mainland China, Taiwan, and Macau) using both Chinese and English languages as medium of communication in social media platforms.