Cryptocurrency Fraud Detection via Unsupervised Domain Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies lack effective tools to detect fraudulent blockchain transactions due to the absence of historical data on fraudulent cryptocurrency transactions, making it impossible to develop fraud detection systems similar to those used in traditional payment transactions.
Innovation Solution
The use of unsupervised domain adaptation with deep neural networks to combine labeled data from electronic payment transactions with unlabeled cryptocurrency transaction data, aligning features through domain adaptation algorithms like Deep CORAL, to identify fraudulent transactions without relying on labeled blockchain data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional fraud detection tools are used for blockchain transactions, then fraud detection capability is improved, but the tools cannot be developed due to lack of labeled fraudulent blockchain data
Solution Approach 1:
The patent uses an intermediary domain (electronic payment transactions with labeled fraud data) to bridge the gap between having no fraud detection capability and needing fraud detection for blockchain transactions. The labeled electronic payment data serves as a mediator that enables training of domain adaptation models without requiring labeled blockchain fraud data.
Solution Approach 2:
The patent changes the domain parameter from electronic payment transactions to blockchain transactions while maintaining the fraud detection function. By using domain adaptation techniques, the model transforms parameters (features) from the source domain to align with the target domain, enabling fraud detection in the new context without labeled target data.
2Measurement precision
If domain adaptation is used to leverage labeled electronic payment data, then fraud detection accuracy is improved, but the complexity of the system increases
Solution Approach 1:
The patent replaces the traditional mechanical approach of collecting and labeling fraudulent blockchain transactions with an automated domain adaptation system. The unsupervised domain adaptation algorithm automatically aligns feature distributions between domains, substituting manual data collection and labeling processes with automated computational methods.
3Loss of time
If features from electronic payment transactions are directly applied to blockchain transactions, then development time is reduced, but detection accuracy deteriorates due to domain differences
Solution Approach 1:
The patent performs preliminary domain adaptation processing to align feature distributions from electronic payment transactions with blockchain transaction features before applying the model. This preliminary alignment ensures that features transferred from the source domain are compatible with the target domain, maintaining detection accuracy while avoiding the time-consuming process of collecting labeled blockchain fraud data.
Data Source
AI summary
A method for identifying fraudulent cryptographic currency transactions using a deep neural network includes: receiving, by a receiver of a processing server, a source dataset, the source dataset including labeled source data associated with a plurality of source features and being associated with a source domain; receiving, by the receiver of the processing server, a target dataset, the target dataset including unlabeled target data associated with a plurality of target features and being associated with a target domain; combining, by a processor of the processing server, at least a subset of the plurality of source features and at least a subset of the plurality of target features into a combined data layer; training, by the processor of the processing server, a deep neural network using a domain adaptation algorithm and the combined data layer to identify a set of final features.