In social data mining classification is considered as the most effective decision making techniques among all the human activities. In this paper, the authors propose a new link based classifier using neuro fuzzy logic for analyzing the social behavior based on weblog dataset. In this system, data are processed using a multistage structure. This system provides a diagnosis using a neuro fuzzy link based classifier that analyses the user's behavior to specific diagnostic categories based on their cluster category in social networks. It uses random walks method to organize the labels.