Application of Naïve Bayes Algorithm for Sentiment Classification of Subtle Sandwich Feedback from Google Forms

Sentiment analysis is a powerful method for understanding customers' opinions and feelings towards a product or service. This research aims to analyze customer sentiments about Subtle Sandwich products using Naïve Bayes algorithm, with the goal of understanding customer perceptions of Subtle Sandwich. The Naïve Bayes algorithm classifies customer sentiment into three categories, including positive, negative, and neutral. Sentiment data were collected through a Google Form questionnaire distributed to customers. Data processing was done using RapidMiner application along with CRISP-DM data mining design and preparation method approach. The model achieved an accuracy of 78.07 percent, with recall rates of 76.96 percent for positive, 81.37 percent for negative, and 79.07 percent for neutral sentiments. Using the Naïve Bayes algorithm and WordCloud operator, we were able to identify frequently used keywords in the dataset to provide deeper insights for actionable recommendations for improving product quality, consistency, and business strategies.