Neuronal Network Credit Risk Identification via User Operation Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The rapid expansion of personal financial credit business into new scenarios and customer groups poses challenges in quickly and accurately identifying credit risks at the application stage, potentially leading to economic losses if not managed properly.
Innovation Solution
A method using a neuronal network to perform credit risk identification by generating operation vectors from user interactions on a credit business system, processing these vectors through attention models, gating functions, and GRU models to generate a final representation, which is then input into a multilayer perceptron for risk assessment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional credit risk assessment methods are used, then the process is simpler, but the speed and accuracy of risk identification deteriorates
Solution Approach 1:
The patent replaces traditional mechanical credit scoring methods with a neuronal network system that processes user operation data. The neuronal network automatically learns patterns from user behavior data on the credit business system, substituting manual or rule-based risk assessment with an intelligent system that achieves higher accuracy in risk identification.
Solution Approach 2:
The patent introduces operation data as an intermediary between user behavior and risk assessment. By capturing and analyzing user operation data (clicks, navigation, time spent) on the credit business system, the system creates a bridge that provides deeper insights into user intent and risk profile, enabling more accurate risk identification without direct complex intervention.
2Productivity
If rapid expansion of credit business is pursued, then business volume increases, but risk control effectiveness deteriorates
Solution Approach 1:
The patent performs risk identification at the application stage by analyzing user operation data collected before the loan request is fully processed. The neuronal network assesses risk indicators early in the process, allowing the system to maintain high processing speed while ensuring risk control effectiveness by identifying potential risks before they materialize into losses.
Solution Approach 2:
The system continuously monitors user operation data and feeds this information back to the neuronal network for real-time risk assessment. This feedback mechanism allows the system to adapt to changing user behaviors and risk patterns, maintaining effective risk control even as business volume expands rapidly.
Data Source
AI summary
A method and system for credit risk identification. The disclosed embodiments include receiving a loan request to a credit business system from a user; in response to the loan request, obtaining operation data by the user on the credit business system within a period of time before the loan request; performing risk identification on the user using a neuronal network according to the loan request and the operation data to generate a result, and determining a response to the loan request based on the result of the risk identification.