CLASSIFICATION OF CONVERSATIONS ON MOBILE INSTANT MESSAGING

Abstract

Mobile Instant Messaging (MIM)(e.g. WhatsApp) has seen a sudden growth among peers for communication, having many benefits like reduced cost and time for response, however, it has been shown to increase anxiety levels and addiction among users. It follows that to use MIM to our benefit, the harms should be mitigated and controlled. In this work, various machine learning models like Naïve Bayes, SVM and AdaBoost are applied for classification problems. Through this paper, we aim to classify the conversations into work and leisure by selecting the model which gives the highest accuracy. This would ensure that conversations are separated into work and leisure, thereby enhancing productivity of MIM users.