SENTIMENT MINING ON SOCIAL MEDIA USING NAIVE BAYES: A TOOL FOR ENHANCING ACADEMIC PROGRAM DECISIONS
Alphie P. Lavarias, Christian Ernes A Caranto, Junard S. Secretario, Queen Amchell B. Papa, Romulo L. Olalia, Maynard Gel F. Carse · International Journal of Computer Applications · 2025
This study investigates the application of sentiment analysis to social media posts related to academic programs, utilizing datasets composed of both Filipino and English texts.Employing a Naive Bayes classifier, the system achieved an overall classification accuracy of 78.66%, effectively distinguishing positive, negative, and neutral sentiments within the feedback.The data preprocessing pipeline included thorough cleaning, normalization, tokenization, stopword removal, and lemmatization, all of which contributed to enhanced model performance.These findings demonstrate the practical utility of sentiment analysis as an analytical tool for academic institutions seeking to gauge stakeholder opinions and feedback.By identifying trends in sentiment, educational administrators can make informed decisions to improve program quality and engagement.