Client-Side Risk Identification via Service Indicator Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing risk control systems in the Internet financial industry face high server resource consumption and costs due to centralized processing methods, where raw service data is typically stored and analyzed on server devices for risk identification and control, leading to increased costs and resource utilization.
Innovation Solution
A risk identification method and system where a client device processes service data to determine service indicator data, which is then used for risk assessment, reducing the need for raw data storage on server devices and lowering overall storage and calculation costs by performing calculations and analyses locally or in a cloud environment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If centralized server cluster processing method is used for risk identification and control, then risk identification accuracy is improved, but server resource consumption and hardware costs increase significantly
Solution Approach 1:
The patent segments the risk identification system into two parts: a centralized server cluster that provides risk identification models and rules, and distributed client devices that perform local risk assessment. This segmentation allows accurate risk identification to be distributed across multiple nodes, reducing the computational burden on the centralized server while maintaining identification accuracy through local processing.
Solution Approach 2:
The patent introduces service indicator data as an intermediary between raw service data and risk identification results. Client devices process raw service data into service indicator data locally, which then serves as input for risk identification models. This intermediary layer reduces the amount of data that needs to be transmitted and processed by the centralized server, thereby reducing server resource consumption while maintaining identification accuracy.
2Reliability
If centralized server cluster processing method is used for risk identification and control, then comprehensive risk analysis is achieved, but hardware costs and storage costs increase significantly
Solution Approach 1:
The patent extracts and stores only service indicator data on client devices rather than storing all raw service data. By extracting only the essential indicators needed for risk identification, the system achieves comprehensive risk analysis capability while significantly reducing storage requirements on both client and server sides.
Solution Approach 2:
The patent uses service indicator data as a disposable intermediate representation that is generated locally from raw data and then used for risk identification. This approach eliminates the need for long-term storage of expensive raw data on servers, as only the compact indicator data needs to be retained, reducing overall storage costs while maintaining analysis comprehensiveness.
Data Source
AI summary
This specification discloses techniques for risk identification. One example method includes receiving, by a client device, a risk identification request identifying a requested service operation and service data associated with the requested service operation; retrieving, by the client device, service data corresponding to the risk identification request; determining, by the client device, service indicator data associated with the service data; analyzing, by the client device, one or more of the service data and the service indicator based on a risk identification rule or a risk identification model to produce a risk result; and determining, by the client device, whether the requested service operation is a high risk operation based at least in part on the risk result.