Fraud number identification method based on graph embedding

Abstract

The invention discloses a fraud number identification method based on graph embedding. The method comprises the following steps: calculating a call characteristic index of a user based on a call signaling of the user in a period of time, forming a call characteristic index vector of each user, and extracting a contact person having a call relationship with the user in a period of time as a directneighbor of the user to form a direct neighbor group of each user; constructing and training a graph for identifying the fraud number, embedding the graph into a neural network, inputting conversationcharacteristic index vectors of a target user and a plurality of associated users, and outputting label information for identifying whether the number of the target user is a suspected fraud number or not; and inputting the call characteristic index vectors of the target user to be identified and the plurality of associated users into the trained graph, embedding the graph into the neural network, and then judging whether the target user number to be identified is a suspected fraud number according to the output. The invention belongs to the technical field of information, and can realize accurate identification of a fraud number based on a multi-point-point communication relationship structure mode of a fraud phone.

Description

technical field [0001] The invention relates to a method for identifying fraudulent numbers based on graph embedding, which belongs to the field of information technology. Background technique [0002] With the deepening of the country's prevention work on communication and information fraud technology, traditional telephone fraud has been curbed to a certain extent, but criminal gangs are using complex and changeable methods to carry out new types of fraud. The new type of fraud is generally based on gang crimes, usually using number tampering, camouflage, etc., to carry out routine and scripted precise fraud activities against specific individuals. During this process, the typical "point-to-multipoint" contact map of fraudulent calls has been transformed into a "multipoint-to-point" communication relationship. Multi-point communication fraud is even carried out across borders, and the monitoring of a single international gateway cannot restore the change in fraud patterns....

AI Technical Summary

Problems solved by technology

During this process, the typical "point-to-multipoint" contact map of fraudulent calls has been transformed into a "multipoint-to-point" communication relationship. Multi-point communication fraud is even carried out across borders, and the monitoring of a single international gateway cannot restore the change in fraud patterns. , such fraudulent activities usually cause heavy losses and are extremely harmful to society

[0003] In the face of such sparse and cross-border integration of telephone communication relationships, traditional template matching, signaling statistical analysis, and simple international gateway monitoring and other technical means for detecting single-point fraudulent call behaviors have little effect. Therefore, how to It has become a technical problem that technicians need to solve urgently

Images