Method and device for monitoring change of node and auxiliarily identifying blacklist as well as electronic device

Abstract

The disclosure relates to a method and device for monitoring changes of nodes and auxiliarily identifying blacklists as well as an electronic device. The method for monitoring the changes of the nodes comprises the following steps: acquiring the user data, wherein the user data includes user identities of a plurality of virtual users; taking the user identities as the nodes, connecting the nodes with the association relationship to form an edge, and establishing a foundation drawing according to the nodes and the edge; monitoring the changing degree of the association relationship between each node and other nodes; when the changing degree of the association relationship between one node and other nodes exceeds a first preset threshold, determining the node as an abnormal node; or selecting the preset number of nodes with the maximum changing degree as the abnormal nodes. According to the method and device disclosed by the invention, the abnormal nodes can be monitored according to the changing degree between the associated nodes.

Description

Technical field [0001] The present disclosure relates to the field of computer technology, and in particular, to a method, device, and electronic equipment for monitoring node changes and assisting in identifying blacklists. Background technique [0002] With the rapid development of Internet financial services, e-commerce, online transactions, etc., financial fraud has also become ganged and organized, and the requirements for the security of the risk management and control system are becoming higher and higher. In order to be able to effectively prevent fraud, the hidden blacklist needs to be identified in advance. [0003] The existing common fraud prevention methods are mainly divided into the following two types: [0004] 1. A model of anti-fraud rules engine is used to describe the characteristics of fraudulent behaviors, thereby distinguishing fraudulent behaviors from normal operations. In the anti-fraud rule engine, these rules for identifying fraudulent acts rely on "expe...

AI Technical Summary

Problems solved by technology

[0006] The above two methods, through the discovery of repeated individual behavior patterns in financial fraud from historical cases, are not suitable for ganged and organized behaviors

At the same time, these two methods are post-event processing methods. After the occurrence of a fraud case, after-event analysis and processing are carried out, but pre-event prediction and in-event alarm cannot be carried out.

Images