An Analysis Of The Fund Flow Of The Case Of Illegal Absorption Of Public Deposits

As a typical stakeholder economic crime,the crime of illegally absorbing public deposits not only brings huge economic losses to the people,but also brings unstable factors to the country.For economic crimes,the account is the scene and the funds are traces.The flow of criminal funds has become the DNA of most economic crimes.The analysis of the capital flow of illegally absorbing public deposit cases has important application value for public security apartments to investigate and handle cases.Nowadays,as the crime of illegally absorbing public deposits becomes more complicated,the criminals ’criminal tactics become more and more novel,and the criminal role in the case is often no longer as fixed and single as in the past,so the specific crime model of the case has not been investigated.When personnel have mastered or emerged a new type of criminal model,investigators need to use unsupervised machine learning and other methods to explore issues such as the core members of the case,account classification,hierarchical composition,and capital flow.This article first analyzes the clustering results of such cases through traditional clustering methods,and then clarifies the distribution of huge data in illegally absorbing public deposit cases and the flow of funds through community discovery algorithms.The main work of the thesis includes:Firstly,a clustering algorithm is proposed for the structural characteristics of the fund trading network in cases of illegally absorbing public deposits,as well as the data characteristics of the common account types in such cases in terms of inflow and outflow frequency,amount,direction,time and other attributes.The necessity of thinking to deal with such case data.Secondly,preprocess the data of the fund trading network,and call the K-means clustering algorithm,hierarchical clustering algorithm,and DBSCAN clustering algorithm for the total amount of inflow,total frequency of inflow,total amount of outflow,and total frequency of outflow for each trading account through Python to perform cluster analysis on datas.Finally,in view of the inadequacy of traditional clustering algorithms on this kind of data processing,the application of community discovery algorithm is further proposed,and the concept of modularity is used to analyze the counterparty account,transaction direction,transaction frequency,transaction amount,and capital flow in the fund trading network.The community finds continuous optimization.Establish relevant models through community discovery algorithms,comprehensively analyze and judge the behavior characteristics of accounts and funds transactions,combined with the behavior characteristics of different levels of suspects in illegal absorption of public deposit cases,which can be used by the handling unit in investigating criminal suspects in illegal absorption of public deposit cases.Provide the basis and ideas for personnel lock-in,hierarchical positioning,classification,discovery of potential criminals and recovery of stolen losses.Because the data processing is more convenient and the visualization results are clearer,the work efficiency in capital flow analysis is improved,thetime for investigation and evidence collection is shortened,and a favorable opportunity is obtained for the investigation work.