Sentiment Analysis of Students' Tweets on Unilorin CBT Examinations

Abstract

Feedback is essential for any system or service to continue serving the purpose for which it was put in place. Students opinions about a service rendered by their institution can contribute to the improvement of such service. Most studies carried out on students opinions about different issues have been quantitative-based while those based on qualitative feedback have been processed manually, however, this study focused on performing sentiment analysis on students opinion expressed in tweets on their various Twitter account using a hashtag relating to the university of Ilorin CBT examinations such as Unilorin CBT, CBT Unilorin. A lexicon-based sentiment analysis approach was used to derive the polarity of students tweets by classifying the tweets into positive, negative or neutral. The following polarity level was found from analyzed data: 40.54% negative, 34.17% positive and 25.29% others. The results revealed that many students hold a negative opinion about the CBT examination due to challenges faced when writing the examination. To curtail this trend, the study suggested that more CBT halls should be constructed to cater to the high increase in student intake and examination time should be scheduled according to faculty to prevent a situation where the entire students for a particular exam are queuing at the same time.