Account Link Analysis for Early Fraud Risk Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to effectively detect fraudulent user accounts before they engage in fraudulent activities, as malicious users can create new accounts to continue their illicit activities, leading to ongoing losses for service providers.
Innovation Solution
An account classification system analyzes shared attributes between a new user account and multiple known fraudulent accounts to determine a risk level, using a graph-based approach to identify links and assign weights to these attributes, leveraging machine learning to detect potential fraud proactively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the service provider deactivates fraudulent user accounts to prevent further losses, then the immediate fraudulent activities are stopped, but the malicious user can create new accounts to continue fraudulent activities
Solution Approach 1:
The system performs preliminary actions by analyzing account attributes before fraudulent activities occur. It compares attributes of new accounts against known fraudulent accounts to identify patterns and classify accounts as potentially fraudulent before they can execute malicious operations, thereby preventing fraud rather than just responding to it after detection
2Difficulty of detecting and measuring
If the service provider monitors account activities to detect fraudulent accounts, then fraudulent activities can be identified, but the anonymous nature of the Internet allows malicious users to create new accounts continuously
Solution Approach 1:
The system creates a profile or fingerprint of fraudulent account attributes and uses this copy to identify new fraudulent accounts. By storing and comparing attribute patterns (such as device identifiers, IP addresses, or behavioral patterns) from known fraudulent accounts, the system can recognize new accounts that replicate these patterns, effectively detecting fraudsters who create new accounts under different names
3Measurement precision
If the service provider uses traditional fraud detection methods, then existing fraudulent accounts can be identified, but new fraudulent accounts created by the same malicious user are not detected
Solution Approach 1:
The system merges multiple detection signals and attribute comparisons into a unified classification decision. It combines information from various account attributes (device identifiers, IP addresses, transaction patterns, etc.) and compares them collectively against known fraudulent patterns, rather than relying on single-point detection, thereby improving the reliability of identifying associated fraudulent accounts
Data Source
AI summary
Methods and systems are presented for classifying a particular user account as a fraudulent user account by analyzing links between the user account and two or more known fraudulent user accounts collectively. Attributes of the particular user account are compared against attributes of a plurality of known fraudulent accounts to determine that the particular user account has shared attributes with a first known fraudulent account and a second known fraudulent account. The shared attributes with the first known fraudulent account and the second known fraudulent account are analyzed collectively to determine a risk level for the particular user account. The risk level may indicate a likelihood that the particular user account corresponds to a fraudulent account.