Indonesia Covid-19 Pandemic Social Media Analysis With Text Mining
Hilda Adelia Ahmad, Angelina Novianti Winarlie, Eka Miranda · 2022 International Conference on Information Management and Technology (ICIMTech) · 2022
This research aimed to analyze public opinion regarding Indonesia's situation in facing community restrictions (PPKM) during the Covid-19 pandemic. This research used the text mining approach to classify public opinion into two classes of the Pro and Cons classes regarding the policies, along with comparing the accuracy, precision, and recall values using two text classification methods of Naive Bayes and Support Vector Machine (SVM). The data collected were 217 tweets from Indonesia in November 2020. The Naive Bayes method showed 64% accuracy, 72% precision, and 53% recall, while the SVM method showed 63% accuracy, 70% precision, and 53 % recall. Based on these classification text methods results, researchers concluded that SVM's accuracy, precision, and recall values were not higher than Naive Bayes.