Method, system, storage medium and electronic device for identifying order fraud

Abstract

The invention provides an order fraud identification method, system, storage medium and electronic equipment, wherein the method includes: obtaining historical orders, including historical fraud orders; extracting the values ​​of characteristic attributes of historical fraud orders; establishing multiple characteristic attribute groups, each characteristic Attribute groups have different value groups; traverse historical orders according to each value group of each characteristic attribute group, calculate the proportion of historical fraudulent orders under each value group to historical orders, and use it as the fraud rate of each value group; The value group with a fraud rate greater than the fraud threshold is regarded as a high-risk value group, and the sum of the historical fraud orders of all value groups under the characteristic attribute group corresponding to the high-risk value group is calculated, and the characteristic attribute group whose sum is higher than the preset threshold is screened out. The high-risk value group of is used as the fraudulent value group; the fraudulent value group is used to identify the fraudulent attributes of the new order. The invention can detect fraud characteristic attributes in real time, intercept new fraudulent orders in time, and automatically update interception rules.

Description

technical field [0001] The invention relates to the technical field of the Internet, in particular to an order fraud identification method, system, storage medium and electronic equipment. Background technique [0002] Based on the research on historical fraud cases, it can be found that fraud cases always have a certain concentration in some aspects. When we further study the elements with high concentration of fraud, we can find that there are historical reasons behind the emergence of these elements. For example, through research on fraud, it can be found that the locations of users of many fraudulent orders are always concentrated in certain fixed areas. After drawing this conclusion, making targeted fraud rules for orders in these fixed areas can prevent this well. occurrence of fraud. [0003] Or, research has found that many fraudulent orders are created by users who log in from another location. Based on this, setting interception rules for users who log in from ot...

AI Technical Summary

Problems solved by technology

[0004] Then, the existing interception rules for fraudulent orders are basically based on a single fraudulent element. Although the interception is simple and easy to implement, it is difficult to cover other types of fraud

Moreover, the interception rules for fraudulent orders have a certain lag and cannot be automatically updated in real time

Images