Cyberbullying Analysis on Instagram Using K-Means Clustering

Abstract

Social Media, in addition to having a positive impact on society, also has a negative effect. Based on statistics, 95 percent of internet users in Indonesia use the internet to access social networks. Especially for young people, Instagram is more widely used than other social media such as Twitter and Facebook. In terms of cyberbullying cases, cases often occur through social media, Twitter, and Instagram. Several methods are commonly used to analyze cyberbullying cases, such as SVM (Support Vector Machine), NBC (Naïve Bayes Classifier), C45, and K-Nearest Neighbors. Application of a number of these methods is generally implemented on Twitter social media. Meanwhile, young users currently use Instagram more social media than Twitter. For this reason, the research focuses on analyzing cyberbullying on Instagram by applying the K-Mean Clustering algorithm. This algorithm is used to classify cyberbullying actions contained in comments. The dataset used in this study was taken from 2019 to 2021 with 650 records; there were 1827 words and already had labels. This study has successfully classified the tested data with a threshold value of 0.5. The results for grouping words containing bullying on Instagram resulted in the highest accuracy, which is 67.38%, a precision value of 76.70%, and a recall value of 67.48%. These results indicate that the k-means algorithm can make a grouping of comments into two clusters: bullying and non-bullying.