Abstract: In social data mining, classification is considered as the most effective decision making techniques among all the human activities. However, the existing algorithms for classification were based on decision making by a central decision manager. Therefore, the aim of this research, a new intelligent temporal model with an inference engine and a new algorithm called Temporal Neuro Fuzzy Decision Tree Classifier (TNFDTC) has been proposed for social network analysis. In this proposed model, inductive methods are proposed and used to classify values of attributes of unknown objects based on temporal features by providing appropriate classification using decision tree rules. Rules provided by TNFDTC are also useful for understanding the combinations of contents driving popularity over certain social networks. This proposed temporal neuro fuzzy decision tree has a fuzzy decision tree and a fuzzy decision model to handle uncertainty. It also uses temporal constraints to improve the classification accuracy by enhancing the existing neuro fuzzy decision tree classification algorithm. The parameters of the existing fuzzy decision trees have been adapted in this research which are based on stochastic gradient de-scent algorithm and hence it traverses back from leaf to root nodes. This research is useful to provide connection between consolidated features of users based on network data and also using the traditional metrics used in the analysis of social network users. From the experiments conducted in this research, it is observed that the proposed research provides better classification accuracy due to the application of neuro fuzzy classification method in decision model analysis.
Indira Priya Ponnuvel, Uma Rani Saravanan and Kannan Arputharaj, 2016. Intelligent Temporal Model Using Neuro Fuzzy Decision Tree Classification Algorithm for Online Social Network Analysis. Asian Journal of Information Technology, 15: 3247-3255.