基于模型堆叠的以太坊钓鱼诈骗账户识别方法OA北大核心CSTPCD

In recent years,phishing scams have become a type of fraud that cannot be ignored in blockchain platforms,posing a major threat to users'financial security.To solve this problem,this paper proposes a framework for phishing account detection based on blockchain transactions,and verifies its effectiveness by taking ethereum as an example.Specif-ically,the framework alleviates data imbalances and reduces computational effort by introducing sample filtering rules,adopts a cascading feature extraction method to extract valid features,and builds an ensemble classification algorithm based on model stacking to identify phishing accounts.The experimental results show that the framework can effectively identify phishing fraud accounts on ethereum and has certain practical application value.