<![CDATA[Social media platforms such as Instagram and Twitter serve as major channels for football fans to share opinions and respond to club-related dynamics, including Manchester United. Beyond fan interaction, these platforms play an important role in business, marketing, and information exchange, making efficient text classification essential. This study applies the Naïve Bayes to analyze sentiment toward Manchester United’s performance based on 2,500 Instagram comments and 2,500 Twitter comments. The research process included data cleaning, sentiment labeling, and preprocessing steps. An imbalance in positive, negative, and neutral comments was managed using data balancing techniques to enhance model reliability. Results show that balancing significantly improved performance, with accuracy reaching 83.87% for Instagram and 82.48% for Twitter. Improvements in precision, recall, and F1-score further confirmed Naïve Bayes’ capability to handle complex, noisy, and diverse social media language. The study highlights how dataset size, effective preprocessing, and accurate labeling contributed to performance gains. Overall, Naïve Bayes proved effective for sentiment classification, offering insights into public perception of Manchester United. These findings emphasize its potential for large-scale social media analysis, supporting both academic research and practical applications in digital marketing and fan engagement strategies.]]>
Copyrights © 2026
