Enterprise fraud identification method

Abstract

The invention relates to an enterprise fraud identification method. The method comprises the following steps: acquiring enterprise data conforming to normal distribution; based on the enterprise data,detecting primary features through an abnormal point analysis method to obtain an abnormal enterprise, and performing labeling; constructing a supervised learning model based on the labeled abnormalenterprise to obtain identification features; and acting a final supervised learning model composed of the identification features and the primary features on the enterprise data to obtain an enterprise with fraudulent behaviors. Compared with the prior art, starting from a certain significant feature, other significant features are mined, multiple dimensions are integrated for identification, thecheating difficulty of an enterprise is improved under the condition that the accuracy is improved, and the method has high applicability and robustness, is lower in cost, is simple and easy to implement, and does not affect transaction experience.

Description

technical field [0001] The invention relates to the field of enterprise fraud identification, in particular to an enterprise fraud identification method. Background technique [0002] The Internet has penetrated into all aspects of our lives at present, bringing convenience to our life and work, but also bringing certain risks to business transactions. Traditional offline business transactions are basically P2P (person-to-person), and human attributes are easy to identify and verify. But on the Internet, identifying whether someone is a real person poses challenges. As technology develops, online fraud becomes more invisible and harder to detect. [0003] The scalpers that emerged in the e-commerce era always challenge the normal operation of the market. They use their technology / experience to pretend to be normal users and seriously damage the interests of all parties. [0004] The essence of this type of fraud is to camouflage tens of thousands of users through programmi...

AI Technical Summary

Problems solved by technology

In the corporate welfare scenario, scalpers forge tens of thousands of employees to simulate welfare distribution, consumption and other behaviors. Although the use of biometric technology can also solve this problem, it will bring about an increase in costs and a decline in transaction experience.

Images