Group fraud identification method and device based on spatial search algorithm

Abstract

The embodiment of the invention discloses a group fraud identification method and device based on a spatial search algorithm, and the method comprises the steps: extracting the position information ofa fraud number and a non-fraud number during the participation of a marketing activity, building a two-dimensional plane according to the position information, and dividing the two-dimensional planeinto a plurality of grids; determining a central grid in the plurality of grids according to a DPC algorithm, searching by taking the central grid as a search starting point, calculating feature similarity of adjacent grids of the central grid, and clustering according to a spatial search algorithm to obtain a clustering result; and determining whether the to-be-processed number is a group fraudulent number according to the clustering result. The user is projected into a two-dimensional plane through the geographic position; grids are divided, the grids with high similarity with the grids withthe high fraud probability are quickly searched, group fraud numbers are found, fraud risks are directly calculated from group behavior characteristics, fraud user groups can be efficiently and accurately recognized, major fraud group risks are found, and enterprises are helped to effectively recognize fraud behaviors.

Description

technical field [0001] The embodiments of the present invention relate to the field of communication technology, and in particular to a group fraud identification method and device based on a space search algorithm. Background technique [0002] With the development of the Internet and the continuous investment of capital, companies in related fields such as e-commerce, Internet finance, new media, and sharing economy increase user value in order to attract new users and activate active users, and launch various marketing activities for new and old users, including: free experience, Discounts and even cash back. General marketing activities only allow a single user (number) to obtain a limited number of activity participation qualifications. In order to obtain more activity participation qualifications, there are special gangs to participate in batches in an organized and planned manner. [0003] A professional fraud team registers accounts by mastering a large number of nu...

AI Technical Summary

Problems solved by technology

[0004] The fraudulent user identification method in the prior art has the following disadvantages: on the one hand, professional fraud is gang behavior, and it is difficult to detect group fraud for the identification of a single user, and users are independent of each other, and the concept of group fraud has not been established; on the other hand , the model and rules based on business characteristics ignore the direct connection of fraudulent users in geographical location, which affects the accuracy and coverage of predictions, and cannot detect major risks of fraudulent groups, helping enterprises to effectively identify fraudulent behavior

Images