Analyzing Instagram User Sentiment Toward MSME Using Naive Bayes and Logistic Regression

In the contemporary era, the rise of social media has significantly altered how public opinions are formed and consumers behave. This shift has given internet users unprecedented power to shape societal trends and consumer preferences. Micro, small, and medium-sized enterprises (MSMEs) are poised to capitalize on this phenomenon by strategically taking advantage of social media platforms to increase visibility, engage with customers, and ultimately boost sales. The current study seeks to investigate the impact of consumer behavior on businesses operating through social media, specifically Instagram. To achieve this, datasets are created from user comments on MSME Instagram accounts and then processed using Naive Bayes and Logistic Regression classification algorithms. These algorithms are chosen for their effectiveness in text classification and sentiment analysis. The findings reveal that Naive Bayes achieves an accuracy rate of 77.04% pre-optimization and 78.44% post-optimization, while logistic regression yields an accuracy rate of 85.12% preoptimization and 86.72% post-optimization.