Analyzing Bubble Tea Brand Preferences and Feedback Through Sentiment Analysis in Malaysia

Bubble tea, also known as milk tea or boba tea, is currently a highly popular and sought-after beverage. Its origins and modern adaptation can be traced back to Taiwan, from where it rapidly gained popularity across the Asia Pacific region. The demand for bubble tea has surged as a significant number of consumers have developed a strong affinity for it. This drink has gained favor not only due to its potential health benefits but also due to its distinct and delightful flavor combinations. This research aims to identify the most favored bubble tea brands within the Twitter user community. To accomplish this, an analysis will be conducted on three prominent bubble tea brands: Tealive, Chatime, and Tiger Sugar. The study will utilize sentiment analysis to evaluate customer reviews, while a Naive Bayes classifier will be employed to categorize the dataset. The sentiment analysis of the collected datasets from these three brands yielded the following outcomes: 383 instances were classified as positive, 163 as negative, and a total of 1493 as neutral. The classification process achieved an accuracy rate of 84%, indicating the effectiveness of the selected classifier in handling the dataset. The results clearly indicate that among the three brands, Tealive emerged as the most favored, followed by Chatime and Tiger Sugar. Additionally, the analysis revealed that words like “sedap” “manis” “matcha” “milk tea” and “boba” appeared frequently in the reviews. In conclusion, while this study sheds light on Twitter users' preferences for bubble tea brands, it also highlights limitations associated with using Twitter data for accurate reviews, given its limited user base for discussing bubble tea opinions. Future enhancements to the system should focus on incorporating a more diverse range of datasets to enhance the precision of the results.