Topic adaptive sentiment classification based community detection for social influential gauging in online social networks

Online Social Networks (OSNs) such as Twitter, Facebook, Instagram, and WhatsApp are turned as a place for many of people in recent years to spend much of their time, due to their huge network structure and massive amounts of user-generated data in it. Those data’s are widely used in various real-world applications such as online marketing, epidemiology, digital marketing, online product or service promotion, and online recommendation systems. Presently, the twitter has grown to become a mainstream medium for the dissemination of messages, which creates necessitated intensive research challenges in the field of social influential gauging, Influence Maximization Problems, alongside an information diffusion. First, to address the social influential gauging a novel Topic Adaptive Sentiment Classification based Community Detection (TASCbCD) algorithm is proposed to detect communities in twitter network based on the results of topic based sentiment classification using robust topic features. In the topic modelling, the initial topics of each extracted data and the robust topic features were used to classify using a multi-class support vector machine. The WordNet and SentiWordNet are benchmark data sets that are used for supporting those classification to achieve the desired results. The resultant communities give a better visualization of identifying the overlapping communities that helps to gauge the topic based social influential user in OSNs. However, from the experimental result, it is observed that the proposed algorithm achieves better results in RandIndex and Scaled Density metrics than state-of-the-art methods for communities detection.