Telecommunication fraud identification method and device based on ticket data

Abstract

The invention provides a telephone bill data-based telecommunication fraud identification method and apparatus. The method comprises the following steps of obtaining telecommunication fraud case information and telephone bill data and performing preprocessing to form a basic database; constructing a knowledge graph taking a case, a suspect, a victim and a telephone as entities and taking a communication relation and a telephone affiliation relation as edges; calculating a shortest path between the two calls through a shortest path algorithm, and mining important intermediate contacts; constructing a clustering model capable of distinguishing different social relationships of callers to obtain the social relationship of fraud group members; and screening out the important intermediate contacts, the call rule of which with the known telecommunication fraud sub-accords with the fraud group member social relationship, from the important intermediate contacts, taking the screened important intermediate contacts and the known telecommunication fraud sub-as suspicious fraud group members, and further determining the fraud group members. According to the invention, the electricity fraud identification work can be improved, and the electricity fraud is avoided.

Description

technical field [0001] The invention relates to the field of bill data, in particular to a method and device for identifying telecommunication fraud based on bill data. Background technique [0002] Non-contact telecommunications network fraud cases have many characteristics, such as the creation of multiple application software by a suspect and the diversification of the same gang's crime methods. Therefore, it is difficult to use the exploration methods used in traditional identification to complete the work of concatenating cases and collecting evidence, which makes it generally difficult to crack down on such cases. Especially in the process of concatenation of cases, using the common features of different cases to find the concatenation points of cases has become an indispensable part of identifying telecom fraud. [0003] Therefore, it is urgent to study a method for identifying telecommunication fraud based on CDR data to improve the identification of telecommunicati...

AI Technical Summary

Problems solved by technology

Therefore, it is difficult to use the investigation methods used in the traditional identification to complete the work of combining cases and obtaining evidence, which makes it generally more difficult to crack down on such cases

Images