COMPARISON OF RANDOM FOREST AND SUPPORT VECTOR MACHINE METHODS ON TWITTER SENTIMENT ANALYSIS (CASE STUDY: INTERNET SELEBGRAM RACHEL VENNYA ESCAPE FROM QUARANTINE)

Coronavirus (Covid-19) is an infectious disease spreading widely throughout the world. Covid-19 has been declared a pandemic. Transmission of Covid-19 spreads through the Droplet. The Indonesian government has made efforts to prevent the spread of Covid-19, one of which is the implementation of quarantine regulated through Circular Letter Number 8 of 2021 concerning International Travel Health Protocols. The case of Selebgram Rachel Vennya's escape from quarantine had become a trending topic on Twitter. Many Twitter users in Indonesia gave their opinions and comments on this case. Therefore, it is necessary to research public sentiment on the case of the escape of Selebgram Rachel Vennya from quarantine. The data used is taken from netizen comments from social media, namely Twitter, in the form of positive and negative comments; the algorithms used are Random Forest (RF) and Support Vector Machine (SVM). This study aims to compare the classification method to public sentiment regarding the case of the escape of Selebgram Rachel Vennya from quarantine using the Random Forest and SVM methods. The classification results show that the Random Forest algorithm has an accuracy value of 94%. In comparison, the SVM algorithm classification results get an accuracy value of 93%. So it can be concluded, Twitter sentiment analysis in the case study of Rachel Venya's escape from quarantine that the Random Forest algorithm got the best results