SENTIMENT ANALYSIS OF INSTAGRAM COMMENTS: A TOOL FOR EVALUATING POLITICAL POPULARITY IN POST COVID 19 ERA

Abstract

At the fourth level of Maslow’s hierarchy of needs is the esteems need. This ego-driven need compels individuals to consistently engage in self-evaluation using different parameters; one of which is comparing one’s self to other members of their class. An empirical approach to this may involveengaging in popularity contest usually performed through elections or opinion polls. Rather than count ballotsspecifically casted for the purpose of such surveys, this paper took a different approach of popularity assessment supported by modern literatures to examine the sentimental bias of citizens towards political leadersthrough probing thesentiments of members of the public expressed as free comments on Instagram posts. Using the sentiment analysis feature of Natural Language Understanding (NLU) an approach supported by modern literatures, the researchers examined the sentimental bias of citizens towards President Mohammed Buhari compared to three past leaders of the country (Nnamdi Azikiwe, Obafemi Awolowo and Amino Kanu) in a bid to prove the hypothesis that ‘Buhari is more popular than the formal leaders’ as claimed in a statement credited to the Special Adviser to the President on Media and Publicity Femi Adesina published on October 15th, 2021 as contained in Channels Television and Punch Newspaper Instagram handles. The paper used IBM’s NLU service to analyze about 2,558 comments generated by readers of the Instagram post. The study rejects the hypothesis as the result reveals that majority of the commenters shared a negative sentiment towards the person of Buhari with a value of -0.673087 and -0.865672 respectively. While that of ‘Azikiwe’, ‘Awolowo’ and ‘Amino Kano’ all had positives sentiments of 0.510153, 0.452324, 0.324832 respectively showing their preference.