ANALISIS SENTIMEN OMNIBUS LAW PADA UMKM MENGGUNAKAN SUPPORT VECTOR MACHINE, NAIVE BAYES BERBASIS PARTICLE SWARM OPTIMIZATION

Abstract

Omnibus Law has been widely discussed by the Indonesian people. President Joko Widodo will simplify regulations. The government invites the DPR to issue two major laws, namely the Job Creation Law and the MSME Empowerment Law. The public shows the pros and cons of the omnibus law. The purpose of this study is to find out the best method for sentiment analysis using the support vector machine method, naive Bayes based on particle swarm optimization against omnibus law on MSMEs on social media. The method used to analyze the omnibus law sentiment on SMEs begins with data collection from twitter, preprocessing, evaluation. The results obtained from research using the SVM method are the results of the Confusion matrix, namely accuracy 97.23%, precision 97.22% sensitivity and recall 100%. The use of SVM and PSO with 97.23% accuracy, 97.22% sensitivity and 100% recall accuracy. The use of naive bayes produces 94.46% accuracy, 97.14% sensitivity, and 97.14% recall. The use of naive bayes and pso produces 94.46% accuracy, 97.14% precision, sensitivity and 97.14% recall. The final result shows the best performance using the support vector machine and support vector machine methods with particle swarm optimization in omnibus law sentiment analysis on SMEs