Towards a Hybrid Approach of Primitive Cognitive Network Process and K-Means Clustering for Social Network Analysis

Social network sites (SNSs) have been influencing the social activities of many people. Consequently, analysis of social network data may produce meaningful information for decision making. This paper represents the basic hybrid approach of Primitive Cognitive Network Process (PCNP) and classical K-Means Clustering for grouping users in social network sites (SNSs) into appropriate clusters by the similarities among users. This new method has combined the PCNP approach, which is a revised approach of the Analytic Hierarchy Process (AHP), and the K-means method for evaluating the weighted attributes influencing the similarity between users. The proposed approach can act as a friends referring function in various kinds of SNSs.