Code-Mixed English-Indian Languages: Hate Speech Dataset Analysis

In the era of Social Media Networks (SMN) and Online Forum (OF) such as Facebook, Instagram, Blogging Sites, Gaming Platforms etc., users tend to comment significantly in English and Indian Code Mixed (CM) languages. Code-mixing is a common manifestation in linguistically diverse social groups such as India. Many people do not know English, so thousands of people on these platforms write comments in code-mixed languages such as Hindi-English (Hinglish). Researchers have retrieved comments on the CM language from SMN and OF. Many people respond in the form of Hate Speech (HS) and Offensive Language (OL) to insult or hurt others, a person, or specific groups. SMN and OF encounter issues with pinpointing and eradicating HS and disagreeable information in diverse CM Indian languages. OL is straightforwardly or inadvertently used in hate speech posts to demean someone and that can cause psychological distress to persons. So, a specific and effective Code Mixed dataset and classification are required to detect HS and OL. Therefore, in this study, we carried out an in-depth analysis on the retrieval methodology and classification of existing data sets for HS and OL. In this paper, we study and analyze 6 (SIX) datasets that were widely adopted in the literature.