In recent years, a lot of data is being poured on social media. Due to the penetration of social media among people, a lot of people have started posting their sentiments, ideas, etc., on social media. These posts can be facts or personal emotions. In this paper, we introduce the concept of hate speech and discuss how it differs from non-hate speeches. The concept of hate speech is very old; however, posting them on social media needs special attention. We have reviewed several techniques and approaches to identify hate speech from textual data with a focus on micro-blogs. Since the notion of hate speech is quite personal, we feel that better IR systems are required to identify hate speech and delete build the systems that are capable to delete the content automatically from social media.
Keywords
- Hate speech
- Machine learning
- Text mining
- Evaluation metrics