Sentiment-Driven Digital Strategy using Naïve Bayes on Coffee X Kecap Bango

Abstract

This research used Naïve Bayes classifier to analyze consumer sentiment towards the co-branded product Coffee X Kecap Bango to gain sentiment-driven insight on how such unusual combination is perceived. The comments were processed with unigram, bigram, and trigram models and the performance was compared. The unigram model achieved the best performance in classifying sentiment. Analysis of commonly used words showed a mixed reaction, reflecting both enthusiasm and skepticism among consumers towards the product. These findings shows the utility of sentiment analysis in evaluating co-branding effectiveness. Given the results, strategies such as co-branding wth local coffee shops and utilizing social media-based community marketing are adviced to address diverse consumer perception and highlight Kecap Bango as a versatile ingredient.