CCDCD:Community Core Mining with Dynamic Constrains Based on Graph Density
Taiyong Li · Jisuanji kexue yu tansuo · 2009
Community core mining(CCM)is a current hot topic in the field of data mining.Existing methods reduce the community to the graphs with specific constraints and treat CCM as dense sub-graphs mining.The existing methods are inefficient on complex problem with multi-graph and dynamic constraints.To improve the efficiency,a novel method for community core mining with dynamic constrains based on graph density(CCDCD).The main contributions include:(1)Reveal the inner disciplines in the graph density with dynamic constraints;(2)Present algorithm dynamic calculation based on updated edges(DCUE)to fast evaluate graph density with variant constraints;(3)Give extensive experiments to show that newly proposed DCUE gets more satisfied results while saving more than 80% time over complex problem,and discover more interesting knowledge about community core with dynamic constraints.