THE IMPACT OF MIXUE’S HALAL ANNOUNCEMENT ON THE COMPANY’S BRAND IMAGE: A NAÏVE BAYES SENTIMENT ANALYSIS APPROACH

• Main Author: Tandiawan, Gilbert
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/73222
• Institution: Institut Teknologi Bandung

With the rise of beverage and ice cream market entry through franchising, Halal certification is more concerned than ever, especially in a country where Muslims take a huge percentage of the whole populations, like Indonesia. The hinderance of business operation could potentially occur if customers felt dissatisfied and unwelcome toward the market entry, that had not had Halal certification, ultimately affecting the brand image above all. Hence, this research aims to find out whether Halal Announcement is able to turn a brand into having a better brand image, by analyzing comments retrieved through Instagram, as well as to evaluate the prediction model of Naïve Bayes by using TF-IDF machine learning classification. The data for this research was separated into two posts, which have 3861 and 3128 unfiltered comments for occurrences before and after Halal Announcement respectively. These comments then are processed to validate that the company was able to rebound the positive sentiments after the Halal Announcement, in which the hypothesis was accepted. Furthermore, the developed Naïve Bayes model is evaluated on the After Halal Announcement dataset, achieving an average accuracy of 65.99%, showing that model is able to predict the sentiment quite fair with several key takeaways that are noted.