Publication

Abstract : In the past years, the activity on social media platforms is rising. Social media platforms ease the communication between users; nevertheless, this has led to hate speech proliferation. Hate speech detection has become a hot research topic, this is not only reflected by the hiked media coverage, but also by the political attention, it is receiving. Automatic hate speech detection is quite demanding owing to the non-standard variations in spelling and grammar. The task becomes strenuous for countries with a multilingual and bilingual population due to the CodeMix nature of the hate content. In this paper, we study the third-order Markov chain model for hate speech detection from CodeMix Hindi-English text. The main motive to use the Markov chain model is that it is language-independent, less time-consuming, and does not require any large domain-specific pretrained models. The promising results indicate that the setup put forward can be used for hate speech detection from any language.