Analisis Sentimen Terhadap Program Kampus Merdeka Menggunakan Algoritma Naïve Bayes Classifier  Dan Support Vector Machine Di Twitter

Kampus Merdeka is one of the policies initiated by the Minister of Education and Culture in 2020. Since it was first launched, this program has received many pros and cons from the public, one of which is from the social media Twitter. The aim of this research is to determine positive, negative and neutral sentiment in the dataset regarding the Merdeka Campus and to determine the optimal accuracy of the comparison between the SVM and NBC methods for the Merdeka Campus on Twitter. The data used in this research amounted to 1000 data based on the most recent comments when the data was taken, via the APIFY website from Twitter post comments through a crawling process. Support Vector Machine is the best algorithm for analyzing sentiment towards the Independent Campus program on Twitter with the highest level of accuracy at a data comparison of 90:10, namely 87%, for precision, recall and f1-score values for negative sentiment, namely 93%, 95% , and 94%, neutral sentiment was 76%, 84%, and 80%, and positive sentiment was 90%, 82%, and 86%. Meanwhile, the Naïve Bayes algorithm obtained the highest level of accuracy in the 90:10 data comparison, namely 81% and obtained precision, recall and f1-score values for negative sentiment, namely 73%, 100% and 85%, neutral sentiment was 78%, 66%, and 71%, and positive sentiment 93%, 76%, and 84%. Based on the highest accuracy value, namely SVM with a data sharing proportion of 90:10, the sentiment results can be visualized, namely that the public's response to the independent campus program tends to be positive with a percentage of 36.5%, while for negative sentiment the percentage is 32.8%. , and neutral sentiment gets a percentage of 30.7%.