TY - GEN
T1 - A Model for Detecting Cryptocurrency Transactions with Discernible Purpose
AU - Baek, Hyochang
AU - Oh, Junhyoung
AU - Kim, Chang Yeon
AU - Lee, Kyungho
N1 - Funding Information:
ACKNOWLEDGEMENT This research was supported by the Institute for Information c& ommunications Technology Planning &Evaluation(IITP) grant funded by the Korea government(MSIT) (No.2017-0-01853, Machine Learning based Intelligent Malware Analysis Platform)
Publisher Copyright:
© 2019 IEEE.
PY - 2019/7
Y1 - 2019/7
N2 - The perpetration of financial fraud progresses parallel with the innovation in the field of finance. Consequently, the emergence of the blockchain technology has also manifested financial transaction obfuscation through the use of de-anonymization of the blockchain technology. This study identifies the suspicious transaction from Binance, an open-source cryptocurrency, through the means of defining and detecting the cryptocurrency wallets. By drawing the metadata of 38,526 wallets from etherscan.io, this study investigates the transactions with discernible purpose. This study performed an unsupervised learning expectation maximization (EM) algorithm to cluster the data set. Based on the features engineered from the unsupervised learning, we performed an anomaly detection using Random Forest (RF). In this study, we offered an insight into labeling the cryptocurrency wallets by providing a model for detecting the cryptocurrency with anomalous transactions. We advocate that labeling the wallets with discernible transactions may help financial institutions, private sectors, financial intelligence, and government agencies identify and detect the transactions with illicit activities.
AB - The perpetration of financial fraud progresses parallel with the innovation in the field of finance. Consequently, the emergence of the blockchain technology has also manifested financial transaction obfuscation through the use of de-anonymization of the blockchain technology. This study identifies the suspicious transaction from Binance, an open-source cryptocurrency, through the means of defining and detecting the cryptocurrency wallets. By drawing the metadata of 38,526 wallets from etherscan.io, this study investigates the transactions with discernible purpose. This study performed an unsupervised learning expectation maximization (EM) algorithm to cluster the data set. Based on the features engineered from the unsupervised learning, we performed an anomaly detection using Random Forest (RF). In this study, we offered an insight into labeling the cryptocurrency wallets by providing a model for detecting the cryptocurrency with anomalous transactions. We advocate that labeling the wallets with discernible transactions may help financial institutions, private sectors, financial intelligence, and government agencies identify and detect the transactions with illicit activities.
KW - Anti-Money Laundering
KW - Blockchain
KW - Cryptocurrency
KW - Ethereum
KW - Machine Learning
KW - Smart Contract
UR - http://www.scopus.com/inward/record.url?scp=85071889450&partnerID=8YFLogxK
U2 - 10.1109/ICUFN.2019.8806126
DO - 10.1109/ICUFN.2019.8806126
M3 - Conference contribution
AN - SCOPUS:85071889450
T3 - International Conference on Ubiquitous and Future Networks, ICUFN
SP - 713
EP - 717
BT - ICUFN 2019 - 11th International Conference on Ubiquitous and Future Networks
PB - IEEE Computer Society
T2 - 11th International Conference on Ubiquitous and Future Networks, ICUFN 2019
Y2 - 2 July 2019 through 5 July 2019
ER -