Analysis of Machine Sensing of Hate Speech on Twitter in Nigeria

Abstract

The Advent Of Social Media Has No Doubt Liberated The Communication Space, Deepened Participatory Democracy And Granted Citizens Platform To Ventilate Their Anger On National And Important Discourse. Twitter Have Gained Prominence In Recent Years Because Of Nature And Quality Of Information Available On Them. Policy Statements That Are Sourced On The Conventional Media Are Replete On Twitter. However, The Platform Has Also Become Breeding Ground For Obfuscation Of Many Absurdities. Therefore, This Study Investigates The Prevalence Of Hate Speech On Twitter In Nigeria From 2015 To 2019. Using Twitter Api As Data Mining Method, The Study Content Analyzed The Mined Tweets By Categorizing The Tweets Under Political, Religious, Ethnic, Region, Gender And Disability Categories. Findings Of The Study Show That Hate Speech On Twitter Is More Frequent During The Electioneering Years. Also, Unverified Handles On Twitter Tend To Tweet More Hate Speech Than The Verified Handles. Findings Show That Twitter Users Tend To ‘Like’ Hate Tweets Than ‘Retweet’ Them. Similarly, Political Hate Speeches Are Dominant On Twitter. The Study Therefore Recommend For Regulatory Legislation On Hate Speech Within The Confinement Of Freedom Of Speech In Order Not To Infringe On The Fundamental Human Right Of The People.