Deteksi Toxic Comment TikTok dan Auto Filtering pada TikTok Menggunakan Support Vector Machine

Abstract

The rapid growth of social media, particularly TikTok, has increased user interaction through its comment feature. However, the large number of comments has also led to the emergence of toxic comments containing hate speech, insults, and cyberbullying. Manual moderation is considered ineffective due to the massive volume of comments and the complexity of informal language commonly used on social media. This study aims to implement a text mining approach using the Support Vector Machine (SVM) algorithm to detect toxic comments and develop an automatic filtering mechanism for TikTok comments. The dataset was collected by scraping comments directly from several TikTok videos. The preprocessing stage included case folding, cleaning, tokenization, stopword removal, and stemming. Feature extraction was performed using TF-IDF, followed by text classification using the SVM algorithm. Model performance was evaluated using accuracy, precision, recall, and F1-score. The experimental results show that the SVM model achieved an accuracy of 95.28%, with a precision of 1.00, recall of 0.67, and F1-score of 0.80 for the toxic class. Furthermore, the developed auto-filtering system successfully filters toxic comments automatically, making the content moderation process faster and more efficient. The proposed approach demonstrates that combining TF-IDF feature extraction with the SVM algorithm can effectively support automated content moderation and help reduce the spread of harmful comments on social media platforms.