Enhancing Spam Comment Detection on Social Media with Emoji Feature and Post-Comment Pairs Approach using Ensemble Methods of Machine Learning

Every time a well-known public figure posts something on social media, it encourages many users to comment. Unfortunately, not all comments are relevant the post. Some spam which can disrupt overall flow of information. This research employed two strategies address issues in text detection media. The first strategy was utilizing emojis that had been frequently discarded studies. In fact, media use convey their intentions. second stacked post-comment pairs, different from systems solely focused comment-only data. pairs were required detect whether comment (not spam) or based post context. used SpamID-Pair dataset derived for Indonesian detection. After comprehensive investigation, emoji-text feature, and ensemble voting could boost performance (in terms accuracy F1). Adding manual features also improved performance. Based experiment, best stand-alone methods SVM (RBF kernel) soft method average