Automated Detection of Hate Speech and Profanity for Multiple and Mixed Languages
Author(s):
Vishwanath V Karki, Kalikidhar Reddy, K S Nitish, B Deepthi
Keywords:
Natural Language Processing, Deep Learning, Code-Mixed, Multilingual, Hinglish, Text Classification
Abstract
Detecting profanity, abuse or hate by building efficient Text Content Moderation systems has become an integral process and a practice for various digital and online media platforms. Chat platforms and community discussion forums are essential customer services that are being provided by many online media business platforms which allow the users which to express their opinions. Many users tend to misuse this service to spread profane and abusive content online.
The aim of this research is to design techniques and create independent models that are able to identify and detect instances of hate, abuse or profane content in English, Hindi and Hinglish. Further discrimination of the inappropriate instances into various classes, namely: (i) HATE (hate speech), (ii) OFFENSIVE (insulting, degrading) and (iii) PROFANE (swear words, cursing) has been performed. This work which aims to solve a prominent problem in the field of Natural Language Processing leverages the Machine Learning and Deep Learning Models to perform the Classification of text.
Article Details
Unique Paper ID: 153363
Publication Volume & Issue: Volume 8, Issue 6
Page(s): 547 - 551
Article Preview & Download
Share This Article
Join our RMS
Conference Alert
NCSEM 2024
National Conference on Sustainable Engineering and Management - 2024