A Literature Survey on Transaction and Phishing URL Fraud Detection System

Abstract

The literature survey provides a comprehensive overview of the complexities surrounding cryptocurrencies, focusing on fraud detection and regulation within the global financial system. It traces the historical evolution of monetary systems, the emergence of cryptocurrencies, and regulatory approaches. Significant findings include the application of machine learning algorithms like LGBM and random forest in Ethereum fraud detection, and the effectiveness of unsupervised learning for anomaly mining in Bitcoin transactions. Novel frameworks for fraud detection through ensemble stacking models are also highlighted. The survey underscores the need for effective approaches to combat fraudulent activities within blockchain platforms, such as Ponzi schemes and phishing scams. Proposed solutions utilize methodologies like graph neural networks and ensemble learning, exhibiting high accuracy. Regulatory measures, classification technique refinement, and future research directions are emphasized to enhance fraud detection models. Overall, the survey illuminates the potential impact of cryptocurrencies on the global financial system, the challenges of unregulated circulation, and promising advancements in fraud detection and regulatory efforts within the digital economy