Consumer finance anti-fraud system and method based on dynamic rule library

Abstract

The invention discloses a consumer finance anti-fraud system and method based on a dynamic rule library. The system comprises a data preparation module, an anti-fraud engine module based on the dynamic rule library, and a modeling human-machine interface module. The method comprises the following steps that: (1) on the basis of data on stocks, establishing a tag system; (2) carrying out EDA (Exploratory Data Analysis) search on tag sample data; (3) on the basis of the preliminary screening of a casting concrete variable, utilizing a decision tree to obtain variable characteristic significancesorting, and determining a final casting concrete variable; (4) on the basis of the established decision tree model, generating a dynamic rule library; and (5) on the basis of the anti-fraud engine module of the dynamic rule library, making a decision. The system has the advantages that the safety and the availability of the rule library are improved, the rule library is unlikely to be cracked; arequired data sample amount is small so as to be suitable for a company or mechanism which just starts consumer finance business, and the cold boot difficulty of the consumer finance anti-fraud systemis lowered; and the degree of participation of the business personnel is improved so as to be favorable for the quick iteration update of the rule library.

Description

technical field [0001] The invention relates to a consumer finance anti-fraud system and method based on a dynamic rule base, which relates to the technology in the field of financial risk control, and in particular to a consumer finance anti-fraud system and method. Background technique [0002] As an important part of inclusive finance, consumer finance has broad market prospects under the background of the country's policy of developing inclusive finance. According to the data released by the international authoritative statistics agency IDC, the scale of my country's consumer finance market will reach more than 12 trillion in 2020. However, the customer groups of consumer finance are natural persons and small and micro enterprises, showing typical sinking characteristics and complex characteristics. Therefore, in the anti-fraud link, the traditional anti-fraud theoretical system and technology relying on human experience can no longer be applied to consumer finance. Fo...

AI Technical Summary

Problems solved by technology

Under the severe reality that group fraudulent behaviors such as group fraud and breach of contract are gradually becoming high-tech and scripted, the traditional static rule base responds slowly to sudden and group fraudulent behaviors, and once the static rule base is cracked, it will take a short time Huge losses can be caused in a short period of time

Although the latter purely data-driven method can satisfy the rapid response to group fraud and is not easy to be cracked, for most licensed consumer financial institutions or emerging Internet consumer finance companies, the accumulated stock The quantity and dimensional diversity of data are difficul

Images