Influence Spread Analysis With Combined Approach of Centrality Measures and Similarity Based Link Prediction Over Diffusion Model

Social networking sites are online platforms offer interaction platform for its users worldwide. Social Network Analysis (SNA) is a scientific approach to extract data and ana- lyze the structural characteristics of networks. Influence spread means reach of particular post or information to users across the network. Structure based influence analysis have specified evaluation measures to quantify the effectiveness of algorithm in SNA. The analysis is useful to identify network structure, inter- connections, the evolution of the network, the information flow and dissemination and the pattern discovery that can be extracted from influencers in analysis. Here the network is scrutinizedwith structural based influence approach. Structure represents the connections, interactions linkage and the topology of the network. One of the most used indicators based on network data is centrality to identify influence spread. Here, Degree, PageRank, Eigenvector, Neighborhood Coreness, and Hypertext Induced Topic Selection centrality measurement techniques wereapplied to find influenced nodes. Network diffusion model is used for spread analysis using different centrality measures. The nodes are selected and grouped for link prediction basedon centrality measure result. Then link prediction is performed using similarity-based metrics including local, global and quasi local metrics. The interactive network visualization is used for evaluation of proposed methodology.