Detecting Hate Speech Contents Using Embedding Models

• Published in: Computational Data and Social Networks pp. 138 - 146
• Main Authors: Duong, Phuc H., Chung, Cuong C., Vo, Loc T., Nguyen, Hien T., Ngo, Dat
• Format: Book Chapter
• Language: English
• Published: Cham Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Deep learning; Hate speech detection; Natural language processing
• ISBN: 9783030914332; 303091433X
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-030-91434-9_13

The rise of hate speech contents on social network platforms has recently become a topic of interest. There have been a lot of studies to develop systems that can automatically detect hate speech contents. In this paper, we propose a knowledge-rich solution to hate speech detection by incorporating hate speech embeddings to generate a more accurate representation of the given text. To obtain the hate speech embeddings, we construct a hate speech dictionary in a semi-supervised fashion. We conduct experiments on two popular datasets, which show that the combination of word embeddings and hate speech embeddings can produce promising results when compared with the methods that employ large-scale pre-trained language models.