ANALISIS SENTIMEN REVIEW PADA APLIKASI MEDIA SOSIAL TIKTOK MENGGUNAKAN ALGORITMA K-NN DAN SVM BERBASIS PSO

Abstract

Review sentiment analysis on social media applications is one of the methods used to analyze opinions and feelings (sentiment) of social media users towards a particular product, service or topic. Tiktok social media users are the second most in the world. The Tiktok app is the leading social media platform and the ultimate destination for short-form videos. Music, dance, education, beauty, passion, or talent show. This research uses data from Tiktok application reviews based on positive and negative sentiments to compare the K-Nearest Neighbor (K-NN) and Particle Swarm Optimization (PSO)-based Support Vector Machine (SVM) algorithms. To test the results of the PSO-based K-NN and SVM algorithms using the Cross Validation method from the test results that the PSO optimization SVM algorithm has the best accuracy compared to the KNN algorithm. Where the accuracy value of SVM is 86.40% and AUC is 0.908. The PSO optimization SVM has an accuracy of 88.20% and an AUC of 0.91. While the K-NN algorithm has an accuracy of 83.40% and an AUC of 0.903 then the accuracy value of the K-NN optimization PSO gets an accuracy of 69.20% and an AUC of 0.77. This means that the use of the PSO optimization SVM algorithm has the highest level of accuracy.