Network Analysis

Network science is the study of complex networks. It provides theories, techniques and tools that help us understand the structure and evolution of blockchain networks. 
To analyze blockchain transaction networks, we use methods of network construction, sampling, and embedding.

Network Construction

In the network construction, we modelsthe original transaction records into the research object of this paper which is called complex information network. Network analysis based on Ethereum data is significantly different from traditional network analysis. The transaction time and amount in the financial transaction network are two important attributes in the transaction network. Simply constructing transaction data into a complex network will lose a lot of useful information. Therefore, in the future, more information needs to be considered in the modeling choice of Ethereum transaction data analysis, such as directed network, weighted network, time / dynamic network, etc.

Construction Method of Temporal Weighted Multiple Directed Network

Introduction

In order to investigate transactions conveniently, we will abstract the transaction record about Ether transfer as a four-tuple (src, dst, w,t) which means the sender src transfer w ether to the receiver dst at time t. In order to retain transaction information as much as possible, we build Ethereum transaction data as a self-defined network model-Temporal Weighted Multidigraph (TWMDG), referred to as Temporal Weighted Network.
TWDMG is not a simple graph but a multigraph, due to the presence of parallel sides. Therefore, TWDMG no longer defines edge as node pair (u, v). It encompasses the edge time and weight information (u, v, w, t). We can uniquely identify multiple network an edge through the four-tuple. We collect the four-tuples to construct a sequence representing a TWMDG where each node represents a unique Ethereum account and each edge represents a unique Ether transfer transactions.

Citation

BibTeX

@article{lin2020modeling,
title={Modeling and Understanding Ethereum Transaction Records via A Complex Network Approach},
author={Dan Lin, Jiajing Wu, Qi Yuan, Zibin Zheng},
journal={IEEE Transactions on Circuits and Systems--II: Express Briefs },
year={2020},
month={to be published},
publisher={IEEE},
doi={10.1109/TCSII.2020.2968376}
}

IEEE

D. Lin, J. Wu, Q. Yuan and Z. Zheng, "Modeling and Understanding Ethereum Transaction Records via A Complex Network Approach," in IEEE Transactions on Circuits and Systems II: Express Briefs, doi: 10.1109/TCSII.2020.2968376

ACM

Dan Lin, Jiajing Wu, Qi Yuan, and Zibin Zheng. 2020. Modeling and Understanding Ethereum Transaction Records via A Complex Network Approach. IEEE Transactions on Circuits and Systems II: Express Briefs. DOI: 10.1109/TCSII.2020.2968376

Network Embedding

The network embedding vector is an effective representation of the network, and it is the bridge and medium between the original network and the network application tasks. Intuitively, similar nodes in the original network should have similar node vectors in the embedded low-dimensional space. The role of network embedding is to mine the implicit characteristics of nodes in the transaction networks, reduce the dimension of transaction data, and apply the large-scale transaction networks into a wide range of scenarios based on embedding vectors.

Random walks netwok embedding method

Introduction

The random walk algorithm can get the node-to-node association in the network. Two nodes with similar topologies often show similar functions. In the defined new network (TWMDG), we consider the timing and multiplicity of edges in the network, redefine the temporal walk, and combine the temporal and weight information of edges to propose a Temporal WEighted MultiDiGraph Embedding algorithm. The walking sequence should consider time and weight. The proposed method is expected to learn more meaningful and accurate time-dependent node embeddings that capture more comprehensive properties from dynamic transaction networks。

Citation

BibTeX

@article{lin2020modeling,
title={Modeling and Understanding Ethereum Transaction Records via A Complex Network Approach},
author={Dan Lin, Jiajing Wu, Qi Yuan, Zibin Zheng},
journal={IEEE Transactions on Circuits and Systems--II: Express Briefs },
year={2020},
month={to be published},
publisher={IEEE},
doi={10.1109/TCSII.2020.2968376}
}

IEEE

D. Lin, J. Wu, Q. Yuan and Z. Zheng, "Modeling and Understanding Ethereum Transaction Records via A Complex Network Approach," in IEEE Transactions on Circuits and Systems II: Express Briefs, doi: 10.1109/TCSII.2020.2968376

ACM

Dan Lin, Jiajing Wu, Qi Yuan, and Zibin Zheng. 2020. Modeling and Understanding Ethereum Transaction Records via A Complex Network Approach. IEEE Transactions on Circuits and Systems II: Express Briefs. DOI: 10.1109/TCSII.2020.2968376

Temporal Link Prediction​

Temporal Link Prediction on Ethereum Transaction Network

Introduction

The task of link prediction aims to predict the occurrence of links in a given graph on the basis of observed information. In the temporal link prediction problem, unlike the static link prediction where links have no timestamp, we use the existing links in the past (with smaller timestamps) as the training data to predict the occurrences of links in the future (with larger timestamps).

Citation

BibTeX

@article{lin2020modeling,
title={Modeling and Understanding Ethereum Transaction Records via A Complex Network Approach},
author={Dan Lin, Jiajing Wu, Qi Yuan, Zibin Zheng},
journal={IEEE Transactions on Circuits and Systems--II: Express Briefs },
year={2020},
month={to be published},
publisher={IEEE},
doi={10.1109/TCSII.2020.2968376}
}

IEEE

D. Lin, J. Wu, Q. Yuan and Z. Zheng, "Modeling and Understanding Ethereum Transaction Records via A Complex Network Approach," in IEEE Transactions on Circuits and Systems II: Express Briefs, doi: 10.1109/TCSII.2020.2968376

ACM

Dan Lin, Jiajing Wu, Qi Yuan, and Zibin Zheng. 2020. Modeling and Understanding Ethereum Transaction Records via A Complex Network Approach. IEEE Transactions on Circuits and Systems II: Express Briefs. DOI: 10.1109/TCSII.2020.2968376

Transaction Network Analysis

Ethereum Transaction Network Analysis via A Complex Network Approach

Introduction

We model the Ethereum transaction network as a Temporal Weighted Multidigraph (TWMDG)  introduced in the Network Construction. In a TWMDG, we define the problem of Temporal Weighted Multidigraph Embedding (T-EDGE) by incorporating both temporal and weighted information of the edges, the purpose being to capture more comprehensive properties of dynamic transaction networks.

To evaluate the effectiveness of the proposed embedding method, we conduct experiments of predictive tasks, including temporal link prediction and node classification, on real-world transaction data collected from Ethereum. Experimental results demonstrate that T-EDGE outperforms baseline embedding methods, indicating that time-dependent walks and multiplicity characteristic of edges are informative and essential for time-sensitive transaction networks.

Citation

BibTeX

@article{lin2020modeling,
title={Modeling and Understanding Ethereum Transaction Records via A Complex Network Approach},
author={Dan Lin, Jiajing Wu, Qi Yuan, Zibin Zheng},
journal={IEEE Transactions on Circuits and Systems--II: Express Briefs },
year={2020},
month={to be published},
publisher={IEEE},
doi={10.1109/TCSII.2020.2968376}
}

IEEE

D. Lin, J. Wu, Q. Yuan and Z. Zheng, "Modeling and Understanding Ethereum Transaction Records via A Complex Network Approach," in IEEE Transactions on Circuits and Systems II: Express Briefs, doi: 10.1109/TCSII.2020.2968376

ACM

Dan Lin, Jiajing Wu, Qi Yuan, and Zibin Zheng. 2020. Modeling and Understanding Ethereum Transaction Records via A Complex Network Approach. IEEE Transactions on Circuits and Systems II: Express Briefs. DOI: 10.1109/TCSII.2020.2968376

Mixing Service Detection

Mixing services in Bitcoin, originally designed to enhance transaction anonymity, have been widely employed for money laundry to complicate trailing illicit fund. The goal of this sub-project is to build identification for addresses belonging to mixing services via mining transaction network.

Detecting Bitcoin mixng services via transaction network mixing with motifs

Introduction

In this paper, we focus on the detection of the addresses belonging to Bitcoin mixing services, which is an important task for anti-money laundering in Bitcoin. Specifically, we provide a feature-based network analysis framework to identify statistical properties of mixing services from three levels, namely, network level, account level and transaction level. To better characterize the transaction patterns of different types of addresses, we propose the concept of Attributed Temporal Heterogeneous motifs (ATH motifs). Experiments on real Bitcoin datasets demonstrate the effectiveness of our detection model and the importance of hybrid motifs including ATH motifs in mixing detection.

Citation

BibTeX

@misc{wu2020detecting,
title={Detecting Mixing Services via Mining Bitcoin Transaction Network with Hybrid Motifs},
author={Jiajing Wu and Jieli Liu and Weili Chen and Huawei Huang and Zibin Zheng and Yan Zhang},
year={2020},
eprint={2001.05233},
archivePrefix={arXiv},
primaryClass={cs.SI}
}

IEEE

J. Wu, J. Liu, W. Chen, H. Huang, Z. Zheng, and Y. Zhang, “Detecting Mixing Services via Mining Bitcoin Transaction Network with Hybrid Motifs,” vol. 14, no. 8, pp. 1–10, 2020.

ACM

Jiajing Wu, Jieli Liu, Weili Chen, Huawei Huang, Zibin Zheng, and Yan Zhang. 2020. Detecting Mixing Services via Mining Bitcoin Transaction Network with Hybrid Motifs. 14, 8: 1–10. Retrieved from http://arxiv.org/abs/2001.05233

Phishing Scams Detection

Blockchain has attracted an increasing amount of researches, with lots of refreshing implementations in different fields. Cryptocurrency, as its most famous implementation, suffers economic loss due to phishing scams. In our work, accounts and transactions are treated as nodes and edges, thus detection of phishing accounts can be modeled as a node classification problem.

Phishing Scams Detection in Ethereum Transaction Network

Introduction

Blockchain has attracted an increasing amount of researches, with lots of refreshing implementations in different fields. Cryptocurrency, as its most famous implementation, suffers economic loss due to phishing scams. In our work, accounts and transactions are treated as nodes and edges, thus detection of phishing accounts can be modeled as a node classification problem. To tackle the problem, we propose a detecting method based on Graph Convolutional Network (GCN) and autoencoder to precisely distinguish phishing accounts. Experiments on different large-scale real-world datasets from Ethereum show that our proposed model consistently performs promising results compared with related methods.

Citation

BibTeX

@misc{xblockEthereum,
author = {Chen, Liang and Peng, Jiaying and Liu, Yang and Li, Jintang and Xie, Fenfang and Zheng, Zibin},
title = {{XBLOCK Blockchain Datasets}: {InPlusLab} Ethereum Phishing Detection Datasets},
howpublished = {\url{http://xblock.pro/ethereum/}},
year = 2019
}

IEEE

L. Chen, J. Peng, Y. Liu, J. Li, F. Xie, and Z. Zheng “{XBLOCK Blockchain Datasets}: {InPlusLab} Ethereum Phishing Detection Datasets,” \url{http://xblock.pro/ethereum/}, Accessed: Nov 2019.

ACM

Liang Chen, Jiaying Peng, Yang Liu, Jiatang Li, Fenfang Xie, Zibin Zheng “{XBLOCK Blockchain Datasets}: {InPlusLab} Ethereum Phishing Detection Datasets,” \url{http://xblock.pro/ethereum/}, Accessed: Nov 2019.