Design And Implementation Of A Subsystem For Indentification And Analysis Of Telecom Fraud

With the continuous development of Internet and communication technology,fraudsters use false telephone numbers to send a large number of fraud information in the communication network,to attract awareness of the weak,cheap people fall into their trap,and then cheat their money.Telecom operators manage huge domestic telephone and SMS communication networks,and generate a large amount of call data every day.By mining the hidden information in the data,and to quantify the known characteristics of telecommunications fraud,it can identify the fraudulent telephone numbers hidden in the data,and use the potential information to analyze the situation and trends of telecommunications fraud,and finally get real-time dynamic Telecom fraud public opinion.This thesis mainly studies the identification and analysis of fraud in the field of telecommunications fraud and the application of data visualization:based on the characteristics of telecommunication call list data and telecommunication fraud,a quantitative fraud feature model and threat migration analysis model are built by feature engineering to achieve multitier feature structure.Combining the feature rule classification method and the statistical weighted Bias classification method,we design a double-layer fraud classification model,aiming at improving the accuracy and efficiency of recognition based on the traditional classification method.Using information statistics and analysis rules to analyze and build data models through fraud identification,maintaining data model can make the data transmission of the system more stable.Using the Ajax technology and long polling mechanism in WEB technology to synchronize the front and back data,so that the data visualization has the ability to dynamically adjust the information display.Using SPA technology,view layer framework react and visualization chart library Echarts to detect and analyze data visualization of fraud,making the interface jump faster and display data more smoothly.Based on the above research content,this thesis mainly designs and implements a sub-system for identification and analysis of Telecom fraud.The specific content includes the following aspects:The database is used to store the data,and the feature model is built according to the characteristics of the fraud phone and the call list.Two classification methods are used to complete the two-level classification model of Telecom fraud recognition,and the possible fraud phone number in the call list is identified.Threat assessment and fraud analysis of identified telephone data based on analysis rules and threat migration analysis features.Realization of visual interface system by data visualization.After that,the accuracy and overall function of the system are tested through test data.The results show that a sub-system for identification and analysis of Telecom fraud can achieve better results.