Intelligent financial fraud detection practices in post-pandemic era

Area of Science:

• Financial Fraud Detection
• Artificial Intelligence
• Risk Management

Background:

• Financial fraud causes significant global losses, exacerbated by the COVID-19 pandemic.
• The pandemic accelerated digital financial services, introducing new fraud detection challenges.
• Existing fraud detection methods require adaptation to evolving digital landscapes.

Purpose of the Study:

• To provide a comprehensive overview of intelligent financial fraud detection practices.
• To analyze pandemic-induced changes in fraud risk and data types.
• To explore the evolution of fraud detection methodologies, focusing on emerging techniques.

Main Methods:

• Literature review of intelligent financial fraud detection practices.
• Analysis of fraud risk features influenced by the COVID-19 pandemic.
• Examination of data type evolution from quantitative to unstructured data.
• Summary of traditional and advanced fraud detection methods.
• Discussion of Graph Neural Network applications in post-pandemic fraud detection.

Main Results:

• The COVID-19 pandemic has reshaped financial fraud landscapes and accelerated digital service adoption.
• Fraud detection practices have evolved to incorporate diverse data types beyond traditional quantitative data.
• Graph Neural Networks show significant promise for enhancing financial fraud detection in the post-pandemic era.

Conclusions:

• Intelligent financial fraud detection requires continuous adaptation to new risks and technologies.
• The integration of diverse data and advanced methods like Graph Neural Networks is crucial.
• Future research should address key challenges and explore novel directions in AI-driven fraud detection.