Research Resources

This page provides access to tools, data, and resources developed as part of the SNSF Leading House Asia research project.


Software & Tools

R Shiny Dashboard

Status: Completed Release: December 2024

An interactive dashboard for consumer lending credit risk analysis featuring:

  • Real-time data analysis capabilities
  • Graph-based model visualization
  • Prediction interface for lending decisions
  • Comparison with baseline models

Code Repository

GitHub: Digital-AI-Finance/SNFS-Leading-House-Asia Organization: Digital-AI-Finance

Project documentation, website source, and research methodology available on GitHub. The repository includes conference materials, publications, and the complete project website.


Data Sources

Consumer Lending Data

The project utilizes data from various Chinese consumer finance platforms:

Source Type Description
Internet-based consumer finance Ant Group, Tencent, JD.com platforms
Commercial banks Traditional banking consumer credit data
Licensed consumer finance companies Regulated lending institution data

Note: All data is anonymized and processed in compliance with data protection regulations.


Research Methodology

Graph-Theoretic Analysis Framework

Our methodology consists of four key components:

  1. Data Collection and Processing
    • Comprehensive consumer lending data collection
    • Privacy-compliant anonymization
    • Feature extraction and preprocessing
  2. Graph Construction
    • Network modeling of consumer interactions
    • Transactional behavior mapping
    • Similarity-based edge creation
  3. GNN Model Development
    • Graph Neural Network architecture design
    • Homophily-guided graph construction
    • Default prediction optimization
  4. Validation and Economic Analysis
    • Historical data backtesting
    • False positive/negative rate analysis
    • Financial benefit quantification

Publications & Presentations

Key Research Papers

Conference Materials

  • EcoSta 2025 Invited Session slides
  • ICSA 2024 China presentation
  • AUS ICMS 2025 materials

External Resources

Graph Neural Networks

Resource Description
PyTorch Geometric GNN library
DGL Deep Graph Library
NetworkX Network analysis in Python

Credit Risk Literature

Topic Reference
Traditional scoring FICO, Altman Z-Score
Machine learning XGBoost, Random Forest
Deep learning Neural network approaches
Graph-based GNN, GAT, GCN methods

China Consumer Finance

Resource Description
Ant Group Largest consumer finance platform
Tencent WeBank Digital banking services
JD Finance E-commerce integrated lending

Collaboration Resources

Partner Institutions

Research Networks

COST Action CA19130 - Fintech and AI in Finance

European research network with 300+ researchers from 51 countries. Prof. Osterrieder serves as Action Chair. Key 2024 activities:

  • Brussels Meeting at European Commission (May 14, 2024)
  • PhD School on Generative AI in Finance (2024)
  • Book Call: "Transparency in FinTech" (World Scientific)

MSCA Digital Finance Network

Marie Sklodowska-Curie Industrial Doctoral Network (2024-2027). 17 fully funded PhD positions across 15 institutions in 12 countries. Builds on COST Action CA19130 collaborations. Prof. Chu joined as associated partner in 2024.

SNSF Network-based Credit Risk Project

Swiss National Science Foundation grant for "Network-based credit risk models in P2P lending markets" (Aug 2022 - Aug 2025). CHF 347,836 funding.


Contact

For questions about research resources or collaboration opportunities:

  • Prof. Joerg Osterrieder: joerg.osterrieder@utwente.nl
  • Prof. Jeffrey Chu: jeffrey.jchu@ruc.edu.cn
  • GitHub: SNFS-Leading-House-Asia

(c) Joerg Osterrieder 2025