TY - GEN
T1 - Identification of High Yielding Investment Programs in Bitcoin via Transactions Pattern Analysis
AU - Toyoda, Kentaroh
AU - Ohtsuki, Tomoaki
AU - Mathiopoulos, P. Takis
N1 - Funding Information:
This work is partly supported by the Grant KAKENHI (No.16H07168) from Ministry of Education, Sport, Science and Technology, Japan.
Publisher Copyright:
© 2017 IEEE.
PY - 2017/7/1
Y1 - 2017/7/1
N2 - Although Bitcoin is one of the most successful decentralized cryptocurrency, recent research has revealed that it can be used as fraudulent activities such as HYIP (High Yield Investment Program). To identify such undesired activities, it is important to obtain Bitcoin addresses related with fraud. So far, the identification of such activities is based upon relating Bitcoin addresses with graph mining procedures. In this paper, we follow a different approach for identifying Bitcoin addresses related with HYIP by analyzing transactions patterns. In particular, based on the individual inspection of HYIP activity in Bitcoin, we propose a number of features that can be extracted from transactions. In particular, a signed integer called pattern is assigned to each transaction and the frequency of each pattern is calculated as key features. By evaluating the classification performance with more than 1,500 labeled Bitcoin addresses, it is shown that about 83% of HYIP addresses are correctly classified while maintaining false positive rate less than 4.4%.
AB - Although Bitcoin is one of the most successful decentralized cryptocurrency, recent research has revealed that it can be used as fraudulent activities such as HYIP (High Yield Investment Program). To identify such undesired activities, it is important to obtain Bitcoin addresses related with fraud. So far, the identification of such activities is based upon relating Bitcoin addresses with graph mining procedures. In this paper, we follow a different approach for identifying Bitcoin addresses related with HYIP by analyzing transactions patterns. In particular, based on the individual inspection of HYIP activity in Bitcoin, we propose a number of features that can be extracted from transactions. In particular, a signed integer called pattern is assigned to each transaction and the frequency of each pattern is calculated as key features. By evaluating the classification performance with more than 1,500 labeled Bitcoin addresses, it is shown that about 83% of HYIP addresses are correctly classified while maintaining false positive rate less than 4.4%.
UR - http://www.scopus.com/inward/record.url?scp=85046340448&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85046340448&partnerID=8YFLogxK
U2 - 10.1109/GLOCOM.2017.8254420
DO - 10.1109/GLOCOM.2017.8254420
M3 - Conference contribution
AN - SCOPUS:85046340448
T3 - 2017 IEEE Global Communications Conference, GLOBECOM 2017 - Proceedings
SP - 1
EP - 6
BT - 2017 IEEE Global Communications Conference, GLOBECOM 2017 - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2017 IEEE Global Communications Conference, GLOBECOM 2017
Y2 - 4 December 2017 through 8 December 2017
ER -