Node based Label Propagation for Bitcoin Transaction Pattern Identification Over Similar Community

• Published in: International Conference on Inventive Computation Technologies (Online) pp. 1147 - 1153
• Main Authors: Manwal, Manika, Alvi, Ahmad Murtaza, Turaga, Nagendra Kumar, Mittal, Amit, Rivera, Richard, Degadwala, Sheshang
• Format: Conference Proceeding
• Language: English
• Published: IEEE 26.04.2023
• Subjects: Analytical models; Behavioral sciences; Bitcoin; Blockchain Transaction Logs; Community; Data mining; Data models; Label Propagation; Network topology; Peer-to-peer computing
• ISSN: 2767-7788
• DOI: 10.1109/ICICT57646.2023.10134101

People have conducted decentralized transfer transactions through Bitcoin addresses ever since the Bitcoin system launched online, drastically improving the convenience of transactions. Simultaneously, peer-to-peer transaction logs have become a subject of study. However, it requires a significant amount of time and computing capacity to examine the entire network directly, and it is not conducive to observing the transaction mode within the entity. Consequently, it is possible to construct and analyses the transaction network based on the entity service community and further investigate entity behavior and Bitcoin entities within the service community. A central node-based label propagation algorithm is proposed by enhancing the conventional label propagation algorithm, which divides the Bitcoin entity transaction network into districts and analyses the core communities, including exchanges and mining pools. The transaction method is realistic and straightforward to comprehend. The enhanced label propagation algorithm can converge more quickly and reduce the randomness of the results of community division. The experimental results demonstrate that the internal transaction modalities of various services vary. The simplified display makes the Bitcoin transaction network more readable.