Transaction subgraph network-based Ethereum phishing node detection method and system

Abstract

An Ethereum phishing node detection method based on a transaction subgraph network comprises the following steps: 1) acquiring data; 2) constructing a network, and constructing a target node self-network according to transaction information of an Ethereum external account; 3) performing graph mapping: processing the self-network according to a novel directed graph mapping method to obtain a mapping graph containing intrinsic implicit information, namely a transaction sub-graph network; 4) performing graph feature extraction and fusion: performing feature extraction on the original self-network and the transaction sub-graph network by using a Graph2vec model, and splicing the features of the original self-network and the transaction sub-graph network; 5) classifying a detection model, taking the network fusion features as input, and training a random forest model to detect the phishing attribute of the target node. The invention further discloses an accurate fishing fraud detection system based on the method. According to the method, the self-network is constructed, the corresponding graph mapping technology is combined, the phishing node detection task is lightly and accurately completed, and the method is different from most researches based on extraction of a large amount of or higher-order neighbor information.

Description

technical field [0001] The invention relates to the fields of block chain technology and network science, in particular to a method and system for detecting phishing nodes in an Ethereum platform based on a transaction subgraph network. Background technique [0002] Blockchain is a point-to-point distributed ledger database technology, which realizes the purpose of decentralization through distributed storage and consensus mechanism and other technologies. In recent years, with the increasing price of virtual currency and the continuous advancement of blockchain financial projects, the blockchain field has attracted the attention of a wide range of users and researchers. [0003] As the largest blockchain-based application today, Ethereum is also the largest blockchain platform capable of supporting smart contracts. Compared with the blockchain 1.0 represented by Bitcoin, Ethereum is a Turing-complete scripting language that supports developers to create and publish any dec...

AI Technical Summary

Problems solved by technology

However, since the target of this method is an undirected and unauthorized network, the transaction flow and transaction amount in the actual transaction network are ignored, resulting in the loss of some information

Furthermore, this method makes more use of the first-order and second-order neighbor information and ignores the inherent implicit information between the target node and its neighbors.

Images