Sentiment classification using Twitter data regarding the Covid-19 pandemic in Indonesia

The first COVID-19 pandemic case in Indonesia occurred on March 2, 2020, officially confirmed by the Republic of Indonesia. This has led to various reactions and views from the public on social media. Until now, Twitter is still a very popular communication tool with many comments, and in it, there is valuable information for text mining. Sentiment analysis is necessary to see whether the public’s views on the corona pandemic are positive or somewhat. This classification process uses the K-Nearest Neighbors method. The classification process starts with crawling Twitter data, preprocessing text, and classifying text using KNN, and testing with Confusion Matrix. The test results using the Confusion Matrix obtained 0.673 at an Accuracy value of 0.96. The response to comments from Twitter users in Indonesia regarding the Korona pandemic is more positive when viewed from the classification results.