Transaction Dataset

Catalog

Labeled Dataset

First-order Transaction Network of Phishing Nodes

Second-order Transaction Network of Phishing Nodes

Ethereum Phishing Transaction Network

Bitcoin Partial Transaction Dataset

Unlabeled Dataset

Ethereum On-chain Data

EOSIO On-chain Data

Ethereum Partial Transaction Dataset

Ethereum On-chain Data

Introduction

We run the Ethereum Full Node (up to 10,999,999 blocks) to get the on-chain data (block, trace, receipt) and process them into the following datasets:

Block ：11,000,000 blocks information.

NormalTransaction：There are 858,580,934 ​transactions generated from the block data.

InternalEtherTransaction：Ether is the native cryptocurrency of Ethereum. The transactions of Ether not only happen in the transactions recorded in the block, but also occur during the smart contract execution. 529,634,152 Ether transactions which occur among 87,570,650 addresses are collected.

ContractInfo: Ethereum can be considered as a platform for smart contracts. There are 31,949,110 smart contracts created by 182,142 addresses. It implies that there should be a number of users who create multiple contracts. 14,357,190 contracts are deleted while they refund the Ether balance to 19,139,210 addresses.

ContractCall: In EVM, a smart contract can call another one to invoke some codes or functions. It consists of 2,205,957,409 Contract Calls, among which 1,518,793,033 contain input codes.

ERC20Transaction：In order to collect the information of tokens, we process the receipt dataset to extract the standard events, which are defined in the standard ERC20 protocol of Ethereum community. Additionally, each ERC20 token contains basic information like name, symbol, total supply, etc. There are 467,603,485 ERC20 transactions among 66,719,139 holder addresses.

ERC721Transaction：ERC721 token is another contract protocol proposed by Ethereum community. We find that 6,255 ERC721 contracts contain 28,313,312 token transactions and 653,535 holder addresses.

You can get more details and analysis from the paper called “XBlock-ETH: Extracting and Exploring Blockchain Data from Ethereum“.

Data details

About this table
Ethereum block information.
Columns (14 columns)
blockNumber block number
timestamp timestamp
size block size
difficulty difficulty
transactionCount transaction count
minerAddress miner address
minerExtra miner extra
gasLimit gasLimit
gasUsed gasUsed
minGasPrice min gasPrice
maxGasPrice max gasPrice
avgGasPrice avg gasPrice
miner miner address
reward miner reward
About this table
Ethereum block normal transactions information.
Columns (10 columns)
blockNumber block number
timestamp timestamp
transactionHash transaction hash
from address
to address
creates contract created
value value
gasLimit gasLimit
gasPrice gasPrice
gasUsed gasUsed
About this table
Ether Transaction. The transactions of Ether not only happen in the transactions recorded in the block, but also occur during the smart contract execution.
Columns (8 columns)
blockNumber block number
timestamp timestamp
transactionHash transaction hash
from address
to address
fromIsContract from is contract or not
toIsContract to is contract or not
value value
About this table
Ethereum Contract Information
Colmns (11 columns)
address address
createdBlockNumber created block number
createdTransactionHash created transaction hash
creator creator
createValue createValue
creationCode creation code
contractCode contract code
decreatedBlockNumber decreated block number
decreatedTransactionHash decreated transaction hash
refunder refunder
refundValue refund value
About this table
Ethereum Contract Calling
Colmns (11 columns)
blockNumber block number
timestamp timestamp
transactionHash transaction hash
from address
to address
fromIsContract from is contract or not
toIsContract to is contract or not
callType call type
callingFunction calling function
value value
isError error or not
About this table
Ethereum ERC20 token transaction information.
Columns (7 columns)
blockNumber block number
timestamp timestamp
transactionHash transaction hash
tokenAddress token contract address
from address
to address
value value
About this table
Ethereum ERC721 token transaction information.
Columns (7 columns)
blockNumber block number
timestamp timestamp
transactionHash transaction hash
tokenAddress token contract address
from address
to address
tokenId token id

Citation

BibTeX

@article{zhen2020xblock,
   title={XBlock-ETH: Extracting and Exploring Blockchain Data From Ethereum},
   author={Zheng, Peilin and Zheng, Zibin and Wu, Jiajing and Dai, Hong-ning},
   journal={IEEE Open Journal of the Computer Society},
   year={2020},
   volume={1},
   number={},
   pages={95-106},
}

IEEE

P. Zheng, Z. Zheng, J. Wu and H. Dai, “XBlock-ETH: Extracting and Exploring Blockchain Data From Ethereum,” in IEEE Open Journal of the Computer Society, vol. 1, pp. 95-106, 2020, doi: 10.1109/OJCS.2020.2990458.

ACM

Peilin Zheng, Zibin Zheng, Jiajing Wu, and Hongning Dai, “XBlock-ETH: Extracting and Exploring Blockchain Data From Ethereum,” IEEE Open Journal of the Computer Society, vol. 1, pp. 95-106, 2020, doi: 10.1109/OJCS.2020.2990458.
README

Ethereum Partial Transaction Dataset

Introduction

Different from the Ethereum On-chain Data, the Ethereum Partial Transaction Datasets are three relatively small Ethereum datasets (namely EthereumG1, EthereumG2, EthereumG3) for easier analysis. The transaction datasets are modeled as complex networks, which can be used in graph analysis such as link prediction.

In the constructed network, a node represents an Ethereum account and a link (i.e. edge) represents an Ethereum transfer transaction.

In our work, we conduct temporal link prediction with these three datasets. 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). You can learn more details in the Related Research.

The data details of EthereumG1 are described below. The file structure of EthereumG2 and EthereumG3 are similar to EthereumG1. You can know more information in the README file.

Data details

About this table
Each row represents a mapping from Ethereum address to unique node number (ID).
Columns (2 columns)
addr Ethereum address
idx Node number (ID)
About this table
Sort all the collected edges according to their timestamps.
Columns (4 columns)
From The sender of the transaction
To The recipient of the transaction
Value The amount of money transferred
Timestamp When the transaction happens
About this table
The edge list of the training data for temporal link prediction problem.
Columns (4 columns)
from_node_num Node number (From)
to_node_num Node number (To)
value Value
timestamp Timestamp
About this table
Dataset split. Load this pickle file as a dict.
Columns (4 elements)
train_edges_pos positive node pairs in training set
test_edges_pos positive node pairs in test set
train_edges_false negative node pairs in training set
test_edges_false negative node pairs in test set

Citation

BibTeX

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

IEEE

D. Lin, J. Wu, Q. Yuan, and Z. Zheng, “Modeling and understanding ethereum transaction records via a complex network approach,” IEEE Transactions on Circuits and Systems–II: Express Briefs, 2020, to be published, doi: 10.1109/TCSII.2020.2968376.

ACM

Dan Lin, Jiajing Wu, Qi Yuan, and Zibin Zheng, “Modeling and understanding ethereum transaction records via a complex network approach,” IEEE Transactions on Circuits and Systems–II: Express Briefs, 2020, to be published, doi: 10.1109/TCSII.2020.2968376.
README
Related Work
Run on Kaggle

Second-order Transaction Network of Phishing Nodes

Introduction

The second-order transaction network dataset contains 1660 target phishing nodes and 1700 non-phishing nodes crawled from Etherscan.

The second-order transaction network is divided into two parts. The csv files record the transaction information between the target node and its first-order transaction node, and the csv files also record the transactions between each first-order transaction node and its corresponding second-order transaction node information.

Data details

About this table
Transaction information between the target node and its first-order transaction node.
Columns (6 columns)
TxHash Transaction hash
BlockHeight Height of block
TimeStamp Transaction timestamp
From Address of transaction initiating node
To Address of transaction receiving node
Value Transaction amount

Citation

BibTeX

@InProceedings{yuan2020Phishing,
  author="Yuan, Zihao and Yuan, Qi and Wu, Jiajing",
  title="Phishing Detection on Ethereum via Learning Representation of Transaction Subgraphs",
  booktitle="Blockchain and Trustworthy Systems",
  year="2020",
  publisher="Springer Singapore",
  pages="178--191",
}

IEEE

Z. Yuan, Q. Yuan, and J. Wu, Phishing Detection on Ethereum via Learning Representation of Transaction Subgraphs, vol. 2. Springer Singapore, 2020.

ACM

Zihao Yuan, Qi Yuan, and Jiajing Wu, Phishing Detection on Ethereum via Learning Representation of Transaction Subgraphs, vol. 2. Springer Singapore, 2020.
README
Run on Kaggle

Ethereum Phishing Transaction Network

Introduction

Cryptocurrency, as blockchain’s most famous implementation, suffers a huge economic loss due to phishing scams. In our work, accounts and transactions in Ethereum are treated as nodes and edges, thus detection of phishing accounts can be modeled as a node classification problem. 

In this work, we collected phishing nodes from Ethereum that reported in Etherscan labeled cloud. Starting from phishing nodes we crawl a huge Ethereum transaction network via second-order BFS. Dataset contains 2,973,489 nodes, 13,551,303 edges and 1,165 labeled nodes.

MulDiGraph.pkl：This dataset is stored in pickle format, and it is the  networkx object. Each node is an address with an attribute called isp indicating whether it is a phishing node. Each edge has two attributes, including amount and timestamp, which represent the balance of the transaction and the timestamp of the transaction, respectively. In this data set, the total number of nodes is 2,973,489, the number of transactions is 13,551,303, and the average degree is 4.5574.

Data details

About this table
The networkx format data is not suitable for display in a table, so after reconstructing the three attributes of isp, amount, and timestamp, a form for display is generated as follows.
Columns (6 columns)
from Address
to Address
amount Balance of the transaction
timestamp When the transaction finished
fromIsPhi 1 means fishing mark node, otherwise 0
toIsPhi 1 means fishing mark node, otherwise 0

Citation

BibTeX

@article{chen2020phishing,
  author = {Chen, Liang and Peng, Jiaying and Liu, Yang and Li, Jintang and Xie, Fenfang and Zheng, Zibin},
  title = {Phishing Scams Detection in Ethereum Transaction Network},
  year = {2020},
  issue_date = {December 2020},
  publisher = {Association for Computing Machinery},
  address = {New York, NY, USA},
  volume = {21},
  number = {1},
  doi = {10.1145/3398071},
  journal = {ACM Trans. Internet Technol.},
}

IEEE

L. Chen, J. Peng, Y. Liu, J. Li, F. Xie, and Z. Zheng. 2020. Phishing Scams Detection in Ethereum Transaction Network. ACM Trans. Internet Technol. 21, 1, Article 10 (December 2020), 16 pages. DOI:https://doi.org/10.1145/3398071

ACM

Liang Chen, Jiaying Peng, Yang Liu, Jintang Li, Fenfang Xie, and Zibin Zheng. 2020. Phishing Scams Detection in Ethereum Transaction Network. ACM Trans. Internet Technol. 21, 1, Article 10 (December 2020), 16 pages. DOI:https://doi.org/10.1145/3398071
README
Run on Kaggle

EOSIO On-chain Data

Introduction

We run a EOSIO full node and replay all transactions (up to 89,829,999 blocks) to get the on-chain data (block, transaction receipts, action traces) and process them into the following datasets:

Block：89,829,999 blocks information.

EOSTransferAction(EOSTA)：EOS is the native cryptocurrency of EOSIO. There are 1,356,748,049 internal EOS transfers and 653,529,552 external EOS transfers that occur among 1,156,658 accounts.

ContractInvacationAction(CIA): Unlike Ethereum, all actions (transactions) in EOSIO are completed through calling contracts, including common EOS transfers. There are several system contract accounts in EOSIO, such as eosio, eosio.token, eosio.msig, and so on. In order to investigate the contract development ecology of EOSIO, we extracted the invocation data of all contracts except the system contracts. There are 775,082 authorization accounts initiated a total number of 2,189,162,705 contract invocations.

ContractCodeAction(CCA): EOSIO can be considered as a platform for smart contracts. There are 55,735 SetCode actions for creating or updating 5,594 contracts. It is worth noting that users can easily update contract code in EOSIO, which is not allowed in Ethereum.

ContractAbiAction(CAA): It is worth noting that users can also easily update contract abi, which is not allowed in Ethereum.

TokenInfoAction(TInfoA)：In EOSIO, a contract that contains three standard functions of create, issue, and transfer can be regarded as a standard token contract. 1,826 contracts are considered as standard token contracts, and a total of 4,811 tokens have been created and issued. It implies that in EOSIO, a contract can issue multiple tokens, which is different from that of Ethereum.

TokenTransferAction(TTA)： A total of 1,128,111,142 token transfers occurred in 1,295,389 holding accounts.

TokenIssueAction(TIssueA): Token issuers can send tokens directly to any account without permission, being commonly known as Token Airdrop.

NewAccountAction(NAA): In most public blockchain systems, creating a new address (account) is easy and free. However, in EOSIO, creating a new account requires a creator to buy RAM for storing account information. In addition, the creator will generally stake some CPU and NET resources for the new account to initiate transactions. There are 1,636,043 different accounts (or NewAccounts), which were created by only 45,350 account creators.

CPUNETAction(CPUNETA): In EOSIO, users need to stake CPU and NET for transaction calculation and network transmission. There are 5,474,353 CPU-Related actions, including 3,805,742 stakecpu actions and 1,668,611 unstakecpu actions. Meanwhile, there are 3,100,820 NET-Related actions, including 2,324,444 stakenet actions and 776,376 unstakenet actions.

RAMAction(RAMA): Users need to buy RAM to store information in EOSIO. There are a total number of 2,983,276 RAM-related actions, including 2,546,849 buyram actions and 436,427 sellram actions.

REXAction(REXA): In order to solve the problem that users do not have enough EOS to stake CPU, EOSIO officially launched the CPU and NET leasing mechanism, i.e., the REX mechanism, on May 1, 2019 (around the block 56,000,000). Users can store some EOS tokens in REX pool through buyrex action to lease to others, and retrieve EOS and get the corresponding rent at any time through sellrex action. Meanwhile, users can rent CPU or NET from the REX pool by rentcpu or rentnet actions.

You can get more details and analysis from the paper called “XBlock-EOS: Extracting and Exploring Blockchain Data From EOSIO“.

Data details

About this table
EOSIO block information
Columns (12 columns)
id block id
block_num block number
timestamp block generation time
producer block producer
cpu_usage_ms total CPU usage of the block
net_usage_words total NET usage of the block
confirmed number of blocks that the producer confirms when signing a block. The default is 1, and only the previous block is confirmed.
transaction_count total transaction count of the block
deferred_transaction_count total deferred transaction count of the block
tx_deferred_transaction_count the count of deferred transaction triggered by contract(triggered by another transaction)
person_deferred_transaction_count the count of deferred transaction triggered by person
action_count total action acount of block(exclude inline action)
About this table
Internal and External EOS Transfer Action Information
Columns (8 columns)
txId transaction id
block_num block number
block_time the timestamp of the block contains this transaction (action)
from EOS sender (from)
to EOS receiver (to)
quantity EOS transfer amount
is_inline Whether this action is inline. true means that it is an internal EOS transfers, otherwise, it is an external EOS transfers.
memo transfer memo
About this table
Contract Invacation Action Information (exclude inline actions)
Columns (10 columns)
txId transaction id
action_ordinal the sequence number of the action in a transaction
block_num block number
block_time the timestamp of the block contains this transaction (action)
status execution result
authorization authorization accounts
contract the called contract name
contract_func the called contract function name
error_code error code
except error detail
About this table
Contract Code Deploy or Update Action Information
Columns (7 columns)
txId transaction id
block_num block number
block_time the timestamp of the block contains this transaction (action)
authorization authorization accounts
account contract name
code_size contact code size (bytes)
code contact hex code
About this table
Contract Abi Deploy or Update Action Information
Columns (7 columns)
txId transaction id
block_num block number
block_time the timestamp of the block contains this transaction (action)
authorization authorization accounts
account contract name
raw_abi contact hex abi
json_abi contact json abi, it will null if the ran_abi is unsupported version
About this table
Token Information
Columns (8 columns)
txId transaction id
block_num block number
block_time the timestamp of the block contains this transaction (action), that is the time when this token is created
authorization authorization accounts
token_contract the contract that created and issued the token
token_name token name (symbol)
issuer token issuer
maximum_supply maximum supply of the token
About this table
Token Transfer Action Information
Columns (10 columns)
txId transaction id
block_num block number
block_time the timestamp of the block contains this transaction (action)
token_contract the token contract name
token_name token name (symbol)
from token sender
to token receiver
quantity the amount of token that transferred
is_inline whether it is an inline action
memo transfer memo
About this table
Token Issue Action Information
Columns (8 columns)
txId transaction id
block_num block number
block_time the timestamp of the block contains this transaction (action)
token_contract the token contract name
token_name token name (symbol)
to token receiver
quantity the amount of token that issued
memo issue memo
About this table
Account Creation Action Information
Columns (5 columns)
txId transaction id
block_num block number
block_time the timestamp of the block contains this transaction (action)
creator the account name of the creator
name the new account name
About this table
CPU-related and NET-related Action Information
Columns (9 columns)
txId transaction id
block_num block number
block_time the timestamp of the block contains this transaction (action)
type the action type, delegatebw (stake) or undelegatebw (unstake)
from the account that pay for CPU and NET if the typs is delegatebw; the account that receive EOS if the typs is undelegatebw;
receiver the account that receive CPU and NET if the typs is delegatebw; the account that return CPU and NET to system if the typs is undelegatebw
net_quantity the amount of EOS for NET
cpu_quantity the amount of EOS for CPU
transfer 1 means from give EOS to receiver when it stake CPU or NET for receiver
About this table
RAM-releated Action Information
Columns (8 columns)
txId transaction id
block_num block number
block_time the timestamp of the block contains this transaction (action)
type the action type, buyrambytes, buyram, sellram
payer/account the account that pay for RAM if the type is buyram or buyrambytes; the account that gain EOS if the type is sellram
receiver the account that receive RAM if the typs is buyram or buyrambytes
bytes the amount of RAM, it will null if the type is buyram
quant the amount of EOS for RAM
About this table
REX-releated Action Information
Columns (9 columns)
txId transaction id
block_num block number
block_time the timestamp of the block contains this transaction (action)
type the action type, withdraw, deposit, buyrex, sellrex, rentcpu, rentnet
owner/from see more detail in the paper
receiver see more detail in the paper
amount/rex/loan_payment see more detail in the paper
loan_fund see more detail in the paper
rex_received/proceeds/rented_tokens see more detail in the paper

Citation

BibTeX

@article{zheng2020xblock,
   title={XBlock-EOS: Extracting and Exploring Blockchain Data From EOSIO},
   author={Zheng, Weilin and Zheng, Zibin and Dai, Hong-Ning and Chen, Xu and Zheng, Peilin},
   journal={arXiv preprint arXiv:2003.11967},
   year={2020}
}

IEEE

Zheng, W., Zheng, Z., Dai, H. N., Chen, X., & Zheng, P. (2020). XBlock-EOS: Extracting and Exploring Blockchain Data From EOSIO. arXiv preprint arXiv:2003.11967.

ACM

Zheng, Weilin, Zibin Zheng, Hong-Ning Dai, Xu Chen, and Peilin Zheng. "XBlock-EOS: Extracting and Exploring Blockchain Data From EOSIO." arXiv preprint arXiv:2003.11967 (2020).
README

Bitcoin Partial Transaction Dataset

Introduction

The  Bitcoin  Partial  Transaction  Datasets  contain  three  snapshots  of  Bitcoin  transaction  data  for  easier  analysis,  namly    dataset1_2014_11_1500000,  dataset2_2015_6_1500000  and  dataset3_2016_1_1500000.  We  sample  the  snapshots  from  November  2014  to  January  2016  with  six  months  as  the  sampling  interval.  Each  snapshot  contains  the  first  1,500,000  transaction  records  in  its  corresponding  month,  namly  Nov.  2014,  Jun.  2015  and  Jan.  2016.

We  also  provide  a  file  including  the  labeled  addresses  belonging  to  mixing  services,  and  these  addresses  were  active  during  the  observing  time  of  our  snapshots.

Due  to  the  pseudonymous  requirements  of  Bitcoin,  it  is  unlikely  to  enforce  Know-Your-Customer  (KYC)  processes,  which  are  guidelines  in  anti-money  laundering.  However,  mixing  services  in  Bitcoin,  originally  designed  to  enhance  transaction  anonymity,  have  been  widely  employed  for  money  laundry  to  complicate  trailing  illicit  fund. 

In  our  work,  we  study  mixing  service  detection  with  this  dataset.  For  further  study,  we  can  chase  up  users  involved  in  criminal  activities  by  analyzing  users  who  take  part  in  Bitcoin  mixing.

The  details  of  dataset1_2014_11_1500000  are  described  below.  The  file  structure  of  dataset2_2015_6_1500000  and  dataset3_2016_1_1500000  are  similar  to  EthereumG1.  You  can  know  more  information  from  the  README  file.

Data details

About this table
Information of block.
Columns (4 columns)
blockID Block ID
bhash Block hash (identifier in the blockchain)
btime Creation time of block
txs Number of transactions
About this table
Transaction ID and hash pairs.
Columns (2 columns)
txID Transaction ID
txhash Transaction hash
About this table
Bitcoin address ID and address pairs.
Columns (2 columns)
addrID Address ID
addr String representation of the address
About this table
Information of transaction.
Columns (5 columns)
txID Transaction ID
blockID Block ID
n_inputs Number of inputs
n_outputs Number of outputs
btime Creation time
About this table
List of all transaction inputs.
Columns (3 columns)
txID Transaction ID
addrID Sending address
value Integer sum in Satoshis (1e-8 BTC)
About this table
List of all transaction outputs.
Columns (3 columns)
txID Transaction ID
addrID Receiving address
value Integer sum in Satoshis (1e-8 BTC)
About this table
label.rar contains 38 files, the file structure of each file can refer to BitMixer.io.csv. These files contain addresses of BitMixer.io, BitLaunder.com, BitcoinFog, HelixMixer crawled from walletexplorer.
Columns (1 column)
Address Address belonging to this service （separated by ‘\n’）

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,” ArXiv Preprint ArXiv:2001.05233, 2020.

ACM

Jiajing Wu, Jieli Liu, Weili Chen, Huawei Huang, Zibin Zheng, and Yan Zhang, “Detecting Mixing Services via Mining Bitcoin Transaction Network with Hybrid Motifs,” ArXiv Preprint ArXiv:2001.05233, 2020.
README
Related Work
Run on Kaggle

First-order Transaction Network of Phishing Nodes

Introduction

Recently, blockchain technology has become a topic in the spotlight but also a hotbed of various cybercrimes. Ethereum is currently the largest blockchain platform that supports smart contracts and the corresponding cryptocurrency ether is the second-largest cyptocurrency. Besides, among various security issues of blockchain digital cryptocurrency, the number of phishing scams accounts for more than 50% of all cybercrimes in Ethereum since 2017 and this kind of scam has become as a main threat to trading security of Ethereum, thus emerging as a serious threat to the trading security of the blockchain ecosystem.

Our work shares phishing account information from Etherscan and the code for how to crawl it. In addition, the trans2vec algorithm for detection was also shared.

Data details

About this table
The node list is for phishing detection problem.
Columns (7 columns)
TxHash Transaction hash
BlockHeight Height of block
TimeStamp Transaction timestamp
From Address of transaction initiating node
To Address of transaction receiving node
Value Transaction amount
ContractAddress It’s a contract address or not
About this table
Transaction subgraph
Columns (4 columns)
From Address of transaction initiating node
To Address of transaction receiving node
Value Transaction amount
TimeStamp Transaction timestamp

Citation

BibTeX

@article{phishingdetection,
  
    title={ Who Are the Phishers? Phishing Scam Detection on Ethereum via Network Embedding},
  
    author={ Jiajing Wu , Qi Yuan, Dan Lin , Wei You, Weili Chen , Chuan Chen , and Zibin Zheng},
  
    journal={ IEEE Transactions on Systems, Man and Cybernetics: Systems},
  
    year={2020},
  
    note = {to be published, doi: \url{doi: 10.1109/TSMC.2020.3016821. }},
  
    publisher={IEEE}，
}

IEEE

J. Wu, Q. Yuan, D. Lin, W. You, W. Chen, C. Chen and Z. Zheng, "Who are the phishers? Phishing scam detection on ethereum via network embedding", IEEE Transactions on Systems, Man, and Cybernetics: Systems, to be published, doi: 10.1109/TSMC.2020.3016821.

ACM

Jiajing Wu, Qi Yuan, Dan Lin, Wei You, Weili Chen, Chuan Chen and Zibin Zheng, "Who are the phishers? Phishing scam detection on ethereum via network embedding", IEEE Transactions on Systems, Man, and Cybernetics: Systems, to be published, doi: 10.1109/TSMC.2020.3016821.
README