A Hybrid Probabilistic Privacy Preserving Based Community Detection Model on Online Social Networking Data

Abstract

Privacy preserving plays a vital role on the online social networking sites due to high dimensionality and data size. Community detection is used to find the social relationships among the node edges and links. However, most of the conventional models are difficult to process the community structure detection due to high computational time and memory. Also, these models require contextual weighted nodes information for privacy preserving process. In order to overcome these issues, an advanced probabilistic weighted based community detection and privacy preserving framework is developed on the large social networking data. In this model, a filter based probabilistic model is developed to remove the sparse values and to find the weighted community detection nodes and its profiles for privacy preserving process. Experimental results show that the filter based probabilistic community detection framework has better efficiency in terms of normalized mutual information, Q, rand index and runtime (ms).