Image-Based Scam Detection Method Using an Attention Capsule Network

摘要: In recent years, the rapid development of blockchain technology has attracted much attention from people around world. Scammers take advantage pseudo-anonymity to implement financial fraud. The Ponzi scheme, one main scam methods, defrauded investors large amounts money, thereby harming their interests and hindering application blockchain. Unfortunately, current detection typically largely relies on source code contract or uses a single feature which does not fully represent characteristics. such case, schemes with high efficiency becomes urgent. this paper, we propose an image-based method using capsule network (SE-CapsNet) focused Ethereum. sequence bytecode, opcode frequency, binary interface (ABI) call are extracted as features bytecode ABI, further converted into grayscale images, then mapped three color channels generate RGB used input model for detecting scheme contract. addition, employ fancy PCA data augmentation reduce impact imbalanced results. Experimental results show that deep learning models can effectively detect contracts before transactions occur. Among them, our proposed SE-CapsNet obtains great results, F1 score 98.38%.