Twitter Hate Aspect Extraction Using Association Analysis and Dictionary-Based Approach

Recent research regarding hate speech is in the domain of social sciences and psychology. From these trends, the dissemination of hate speech and antagonistic content in social media has not been extensively studies from the perspective of sentiment analysis. In this paper, the main studies concerned about aspect-based sentiment analysis through twitter as the most popular social media communication in the world as they have 313 million active users worldwide. This paper initiate to address the shortcomings of implied aspects specific for hate crime domain. The expected beneficial hate aspects can be extracted from twitter based on combination of both analysis. The evaluation with researcher's own Hate Crime Twitter Sentiment (HCTS) dataset and also Hate Speech Twitter Datasoft (HSTD) was shown that the proposed approach is effective and produces significantly better results than baselines method.