The significant increase in the number of users has caused public opinion on the Ruang Guru application to be widely spread through social media, especially Twitter. From 15,000 twitter data taken with the keyword Ruang Guru, a total of 2,358 datasets were obtained through the process of handling duplicates. In this study, sentiment analysis was carried out using the Support Vector Machine (SVM) algorithm which was optimized with Particle Swarm Optimization (PSO) then tested using the 10-Fold Cross Validation method which resulted in the highest accuracy rate of 89.20%, while the Support Vector Machine algorithm (SVM) only produces the highest accuracy rate of 88.56%. There is an increase of 0.64% with Particle Swarm Optimization optimization. Sentiment analysis results are positive, with positive results as much as 1463 data or 62.04% and 895 or 37.96% negative sentiment. From the results of this study, it is expected to be a material consideration for Ruang Guru to improve the quality of the service sector found on social media, especially Twitter.
Copyrights © 2022