Deep clustering fraud detection method and device

Abstract

The invention discloses a deep clustering fraud call detection method, which comprises the following steps: carrying out deep clustering on all call bill data to form a plurality of clusters, comparing the plurality of clusters with index values of a fraud cluster, and taking the cluster with the highest matching degree with the index values as a fraud cluster; obtaining each called number calledby the calling number in the fraud cluster, determining all the calling numbers called by the called numbers according to the call bill data, and performing complex network modeling by utilizing the called numbers and all the calling numbers; performing community discovery in a complex network of modeling, and determining a fraud high-risk community according to the proportion of the calling number in the fraud cluster contained in each community; and carrying out voice recognition on each call in the fraud high-risk community, and judging and classifying fraud calls according to a voice recognition result. By applying the method and the device, the fraud call can be found more accurately on the basis of ensuring the real-time performance.

Description

technical field [0001] The present application relates to fraud detection technology, in particular to a deep clustering fraud detection method and device. Background technique [0002] With the continuous development of the communication industry, while bringing more convenience, the rampant telecommunications and network fraud activities follow, and there are more and more means of telephone fraud, which makes people hard to guard against. [0003] Currently, fraudulent call detection methods mainly include call source detection, blacklist interception, etc. These methods generally have problems of poor real-time and flexibility. The update of fraudulent techniques and methods can easily lead to the failure of the original interception means. At present, the high incidence of fraudulent calls is mainly concentrated on fixed-line telephone users. Most of the calling numbers are from overseas. Fraudsters use number changing software or VOIP technology to bypass existing int...

AI Technical Summary

Problems solved by technology

The main disadvantage of this solution is that, from the perspective of behavioral characteristics, it is difficult to deal with changing fraudulent methods only by comparing whether the behavioral characteristics match; from the perspective of voice information, it is often difficult to judge only by matching whether the voice contains fraud keywords. not that precise

Images