Hate speech Detection in Asian Languages:A Survey

L. K. Dhanya, Kannan Balakrishnan

20212021 International Conference on Communication, Control and Information Sciences (ICCISc)17 citationsDOI

Abstract

In this study, we present a language-based survey of hate speech detection in Asian languages. The motivation of this survey is to encourage the development of an automated hate speech detection system for Malayalam. Any message from social media spreading negativity in the society related to sex, caste, religion, politics, race etc. can be called a hateful message. This kind of text is very challenging to detect. Here we have taken only language-specific studies for hate speech detection and analyzed the approaches used in each work. We have used three parameters in this paper to analyze the overall scenario of this problem among Asian languages. This study tries to identify the best classification algorithm for this task and also find the relation between classification approach, type and size of dataset and accuracy. So this survey will become the foundation of future studies in this area and will help to understand the challenges also.

Topics & Concepts

MalayalamComputer scienceCasteTask (project management)Natural language processingVoice activity detectionArtificial intelligenceRelation (database)Speech recognitionSpeech processingLinguisticsData miningPhilosophyManagementEconomicsHate Speech and Cyberbullying DetectionInternet Traffic Analysis and Secure E-votingAdvanced Malware Detection Techniques