IMPLEMENTATION OF CLUSTERING ON TWEET UPLOADING SIDE EFFECTS OF COVID-19 POST VACCINATION USING K-MEANS ALGORITHM

Abstract

The Covid-19 Vaccination Program has become pros and cons among Indonesian people including Twitter social media users. When the program was running, Twitter users started uploading tweets regarding the side effects that occurred, ranging from mild to severe, both scientifically proven. or not. Of all uploaded tweets, by applying text mining, only tweets containing the queries "vaccine effect" and "post vaccine" in the period from January to June 2022 and Indonesian language tweets will be used and based on these parameters a total of 4800 tweets have been collected, of the total These tweets will be further processed using the clustering method, the k-means algorithm and the silhouette coefficient. The results of implementing the silhouette coefficient show that the best cluster is in cluster 2 with a score of 0.6228720387313319 and the results of the clustering algorithm k-means for 4800 tweets obtained 3917 members in cluster 0, and 883 members in cluster 1 placement. The feature is that cluster 0 contains tweets that state program effects explicitly or explicitly, while cluster 1 states program effects that arise implicitly.