Skip to main content
EuroPython logo

Social Media and Sentimental Analysis: CBN Currency Redesign Policy

Track:
PyData: Machine Learning, Stats
Type:
Poster
Level:
beginner
Room:
Exhibit Hall
Start:
13:00 on 12 July 2024
Duration:
60 minutes

Abstract

The identification and measurement of an online audience through the social media platform capitalise on the tonality of emotions on the social media presence. On October 20, the most populous country and acclaimed Africa’s largest economy, Nigeria announced the plans to redesign 200, 500 and 1000 banknotes in replacement of the existing ones. Nigerian citizens expressed different opinions over social media in support of or understanding of the proposed plan and process. Research has shown that shared sentiments on social media can influence the opinions of others and thus the Central Bank of Nigeria’s currency redesign policy. This study, therefore, aimed to identify and analyse general sentiments towards the process of the currency redesign policy with the purpose of determining the citizen’s attitude towards the policy, based on social media comments. Firstly, sentiment analysis was performed on naira redesign-related posts from a selected social media using lexicon-based and supervised machine learning techniques with the purpose of determining a summarised polarity percentage (i.e. negative or positive). The post was collected between January and February 2023. In addition, the performance of the lexicon-based classifier and seven machine learning-based classifiers was implemented and compared in order to use the best-performing classifier in determining the sentiment polarity of the post. Also, the thematic analysis on both positive and negative posts to further understand and revealed general views about the currency redesign policy. Finally, the analytical findings and the possibility of changing the currency redesign policy was discussed.