Application And Research Of Network Community In Violation Valet Financial Behavior Monitoring Of Securities Companies

Recently,illegal phenomenon of irregular valet financial management is increasingly serious in security industry.As for there are a lot of wrong judgments,missing,subjective assumptions and other issues in traditional threshold-based monitoring mode,this paper chooses the Network Community Model to make a in-depth theoretical and empirical research about the illegal valet finance management,striving for enhancing monitoring accuracy and investigation efficiencyAfter a brief introduction of the background and theoretical basis,this paper selects the transaction data of a broker as object,and further studies the illegal behaviors through three levels.The first level are the Network Community Model establishment based on the relation among clients with same trade equipment,and the groupuscule division when accomplishing network integrality analysis and assessment;the second level ulteriorly discerns the secondary associations considering the similarity of transaction behaviors of members in each groupuscule,which locks the customer groups with valet financial management suspicion;the third level makes a ranking list for customers' importance,taking PageRank algorithm as model,which can provide strategic basis for follow-up telephone return visits,interviews and other qualitative work.The ultimate empirical result shows good effects while using Network Community Model to excavate suspicious behaviors of valet financing.Moreover,Python and Igraph can preliminarily achieve the systematic,visual and batched requirements.