Transaction anti-fraud method, device and system

Abstract

The embodiment of the invention discloses a transaction anti-fraud method, a transaction anti-fraud device and a transaction anti-fraud system. The method comprises the steps: obtaining to-be-recognized transaction data, and extracting the transaction feature data of the to-be-recognized transaction data; judging the to-be-identified transaction data by utilizing a preset judgment rule to obtain ajudgment result of the to-be-identified transaction data; inputting the transaction feature data into a pre-constructed transaction evaluation model to obtain an evaluation result of the to-be-identified transaction data, the transaction evaluation model being constructed and obtained according to sample data confirmed as fraudulent transaction, sample data confirmed as non-fraudulent transactionand unconfirmed sample data; and performing decision processing on the evaluation result and the judgment result of the to-be-identified transaction data by utilizing a preset decision rule to obtainan anti-fraud decision result of the to-be-identified transaction. By means of the embodiments of the invention, efficient and stable risk prevention and control of real-time transactions can be realized.

Description

technical field [0001] The invention relates to the technical field of computer data processing, in particular to a transaction anti-fraud method, device and system. Background technique [0002] Traditional transaction anti-fraud systems usually use expert rules to identify fraud risks. Expert rules are often formulated only for one risk event, and require expert experience and post-event data analysis, which has a lag. At the same time, expert rules are not easy to update and iterate. After a period of deployment, the prevention and control effect gradually deteriorates and needs to be re-formulated. Once the expert rules are cracked by fraudsters, they will be completely invalid. Therefore, there is an urgent need in the technical field for a more efficient and stable transaction anti-fraud method. Contents of the invention [0003] The purpose of the embodiments of this specification is to provide a transaction anti-fraud method, device and system, which can realize ...

AI Technical Summary

Problems solved by technology

Expert rules are often formulated only for one risk event, which requires expert experience and post-event data analysis, which has a lag

At the same time, expert rules are not easy to update and iterate. After a period of deployment, the prevention and control effect gradually deteriorates and needs to be re-formulated. Once the expert rules are cracked by fraudsters, they will be completely invalid

Images