Perbandingan Kinerja Algoritma Support Vector Machine dan K-Nearest Neighbor Terhadap Analisis Sentimen Kebijakan New Normal

Abstract

Twitter is one of the popular microblogging sites among internet users, so that many people use Twitter to convey their positive and negative sentiments towards the new normal policy. The pandemic period raises much public sentiment towards the policy of adapting to the new normal. This study aims to classify sentiment tweets into positive and negative classes. The classification algorithms used are k-NN and SVM. The test results show that the k-NN algorithm is better than SVM in solving this sentiment case with an accuracy of 72.96%.