Multi-party data source obtaining method and risk control system

Abstract

The invention discloses a multi-party data source obtaining method and a risk control system. The multi-party data source obtaining method comprises the steps of distributing external request data, and forming multiple queues comprising the corresponding request data; then selecting the request data in the queues to perform processing, and executing tasks corresponding to the request data; and introducing a Future mechanism to ensure that subsequent operation is performed after all requests are finished. The method effectively avoids system crash and processing timeout, and improves the data obtaining and processing efficiency. The risk control system is based on a non-blocking multi-thread data obtaining and processing mode subjected to optimization design, comprises a risk control request control system, a risk control task processing system, a risk control task management system and a risk control decision system introduced as needed, ensures the timeliness of request obtaining andtask processing, and introduces an expert assessment decision as needed, thereby enabling the risk assessment to be more targeted and flexible, and greatly improving the credit service efficiency.

Description

technical field [0001] The invention relates to a wind control data processing method, in particular to a method for acquiring multi-party data sources and a wind control system. Background technique [0002] With the development of Internet finance and e-commerce, people use the Internet to conduct transactions, payments and loans. Usually, in the process of borrowing and payment, risk control assessment is very critical. However, risk control assessment often needs to be based on risk control data, such as user historical payment records, user historical loan records and other multi-party data. [0003] Risk management has been a topic for researchers for many years and an important issue for business professionals. Risk management helps business professionals identify possible future operating risks to an entity (such as a factory) and can help establish procedures for measuring, mitigating, and managing risk. If a bank or credit institution is to extend credit to some ...

AI Technical Summary

Problems solved by technology

[0006] Especially with the continuous increase of online information data on the Internet, massive data analysis and processing directly affect the timeliness and quality of credit ratings. However, as the difficulty of data acquisition and processing increases, the existing credit data processing model can no longer meet actual needs

[0007] In order to improve the efficiency of credit processing, those skilled in the art are making various attempts and efforts, such as a corporate credit risk early warning and monitoring method disclosed in Chinese patent ZL201710459415.9, which includes: collecting multiple Data sources, these extracted associated data are cleaned by deleting invalid data and data merging and deduplication, generating data association relationship diagrams, and repeatedly performing association mining; according to the obtained multiple data sources, combined with relevant business personnel to provide According to the scoring card model, the factors that meet the risk trigger conditions are judged, and the risk score of the main enterprise itself is calculated; the associated influence factors of the affiliated enterprises are calculated, and different risk transfer weights are set to measure the transfer coefficients of different affiliated relationships. Risk score; the calculated risk score of the affiliated enterprise is weighted to the risk score of the main enterprise to obtain the final risk score; finally, according to the calculated risk score, the mechanical graph model is used to draw the relationship diagram, but the The method is only aimed at the early warning of corporate credit, and does not give a corresponding technical solution for how to realize the extraction of multi-party data sources, which has great application limitations for the timeliness of corporate credit ratings, especially personal credit ratings

[0008] In order to obtain a large number of data sources required for credit risk control, the current field generally uses a serial call method or an asynchronous result generation scheme. However, this method will have the problem of relatively long data processing time or poor timeliness; even if the technical solution of the thread pool is adopted It will also cause serious local resource locks

Images