<![CDATA[Accelerating digital innovations have vastly reshaped the methods individuals use to express their perspectives on the gastronomic industry about culinary services through customer reviews on Google Maps. This study seeks to examine the sentiments conveyed in reviews of Bakso GLG to assist management in understanding customer perceptions objectively by employing the Naïve Bayes algorithm only after undergoing rigorous preprocessing phases, including cleaning, case folding, normalization, tokenization, stopword removal, stemming, and the generation of TF-IDF vectors. The classification results yielded an overall accuracy of 77%. The data distribution is dominated by positive sentiment, comprising 226 reviews, followed by 25 neutral reviews and 13 negative reviews. Although the model demonstrated optimal performance in classifying positive sentiment, it encountered difficulties in classifying the minority classes due to the imbalanced dataset. Overall, the system proved effective in processing large-scale review data as a source of strategic evaluation for improving product and service quality in the culinary sector, particularly at Bakso GLG.]]>
Copyrights © 2026
