Identifying and Intercepting Telecommunications Fraud Numbers on the Internet Through Big Data Technolog

With the advancement of network technology, methods of telecommunication fraud have become increasingly diverse. The identification and interception of fraudulent numbers are particularly crucial. This study utilized the network operator’s big data as the basis and conducted Synthetic Minority Over-sampling Technique (SMOTE) processing and feature selection on the original data. The eXtreme Gradient Boosting (XGBoost) algorithm was employed to identify fraudulent numbers and enable early interception. Additionally, an improved Sparrow Search Algorithm (ISSA) algorithm was designed to optimize the parameters of the XGBoost algorithm, resulting in the development of the ISSA-XGBoost algorithm. Experiments were conducted using a collected dataset. The results indicated that after SMOTE processing and feature selection, the recognition effectiveness of the ISSA-XGBoost algorithm for fraudulent numbers was improved. Furthermore, compared to parameter optimization methods such as Grid Search and Bayesian Optimization (BO), the ISSA algorithm demonstrated superior performance. It achieved a precision of 0.945, a recall rate of 0.816, and an F1 value of 0.876. The method also exhibited a low resumption rate of 21.12% in practical applications. These findings validate the effectiveness of the proposed method for identifying and intercepting network telecommunications fraud numbers, further supporting its potential practical applications. © (2024), (International Journal of Network Security). All rights reserved.