PENERAPAN ALGORITMA OPPOSITION-BASED WHALE DALAM KLASIFIKASI SVM UNTUK ANALISIS SENTIMEN TERHADAP KEBIJAKAN PPKM

Corona Virus Disease 2019 (COVID-19) has had a serious impact, forcing the Indonesian government to establish various policies to deal with the spread of COVID-19. One of these policies is the Implementation of Restricting Community Activities (PPKM). During its implementation, this policy raised pros and cons in the community, especially on Twitter social media. The existence of this public opinion, can be used as an effective source of information to assist the government in taking and evaluating policies. The purpose of this study was to determine public sentiment towards PPKM policies based on the classification of tweets or public opinion on Twitter. This process is carried out by applying the OBWOA method for selecting appropriate and optimal SVM parameters and feature selection to select the best features thereby reducing computation time and producing good classification performance. The results of the optimization of the SVM parameters obtained C = 4.99522643 and gamma = 1.4236435 with an average accuracy of 75.20%, precision of 79.73%, recall of 71.65%, and f measure of 70.88%. The feature selection results obtained an average accuracy of 82.40%, precision of 84.23%, recall of 79.63%, and f-measure of 78.96%. In addition, the sentiment classification of 1,389,481 public tweet data in the period January to December 2021 obtained 53% of tweets with Negative sentiment, 25% of Tweets with Neutral sentiment, and 22% of tweets with Positive sentiment.