SENTIMENT ANALYSIS OF TWITTER DATA ON KIP-KULIAH USING TEXTBLOB AND GRADIENT BOOSTING

The Indonesian government aims to position the country among developed nations by 2045, with a primary focus on improving education quality from elementary to higher education levels. One of the key initiatives is the KIP-Kuliah (Indonesia Smart College Card) program, which supports high-achieving students from underprivileged economic backgrounds in accordance with UU No. 12/2012 on Higher Education. This study applies sentiment analysis using TextBlob and the Gradient Boosting algorithm to build a predictive model that identifies public support for the program through Twitter data. The results reveal a significant dominance of negative sentiment, with the model achieving an accuracy of 97%. These findings underscore the importance of sentiment analysis as a feedback tool for policymakers during the implementation of education-related programs. Furthermore, the results suggest that continuous monitoring of public opinion via social media can contribute to more adaptive and responsive policy development. This research highlights the need for future studies to expand the scope of analysis using more advanced natural language processing techniques for deeper understanding and broader coverage of public sentiment.