Penggunaan Naïve Bayes Classifier untuk Pengelompokan Pesan pada Ruang Percakapan Maya dalam Lingkungan Kemahasiswaan

Abstract

Group chat feature provided by instant messaging service like LINE or WhatsApp in student environment has high importance, but not all of the conversation discussed were relevant with the mission of the group itself. With certain reasons, when lot of messages are unread, unaware users will see the first chats displayed on the screen, and will ignore or clear the rest. Thus, it is common for important messages to be ignored this way if they were placed on the middle of the unread chats. That stack of messages could be classified automatically, so users could get the information about the categories of the messages. This research test Naïve Bayes Classifier usage to classify messages in student usage environment, and we show that with good classification model, Naïve Bayes classifier performs well with F-measure score reaching 90,57% for one of the categories.