A novel online social network (Twitter)message (Tweet)classifier based on message diffusion in the network

Abstract

Online social message classification is an important task for E-Commerce companies to mine and classify the customer opinions. In this paper, we have proposed a first of its kind of an efficient message classification algorithm which is independent of tweet content and considers the set of followers who will retweet during the retweet peaks. By including the followers who will retweet during retweet peaks will get a better sampling of the followers set and reduces the computation and storage complexities drastically. Also, we have eliminated the heavy weight operations like DTW to perform the comparison task between the test vector and training vector. The preliminary experiment results authorize that the proposed system attains an accuracy of 95.96% in classification of tweet messages.