Bitcoin address-transaction bipartite graph

Original Paper: 

1: Gaihre, Anil, Yan Luo, and Hang Liu. "Do bitcoin users really care about anonymity? an analysis of the bitcoin transaction graph." 2018 IEEE International Conference on Big Data (Big Data). IEEE, 2018.

2: Gaihre, Anil, Santosh Pandey, and Hang Liu. "Deanonymizing cryptocurrency with graph learning: The promises and challenges." 2019 IEEE Conference on Communications and Network Security (CNS). IEEE, 2019.

Please cite the above papers in case the dataset is used.

A graph reader code to read the datasets in binary format is also included in this package. Note, the weight in the graph is the satoshi (BTC).

Following steps are taken to generate the datasets: We download the publicly available Blockchain raw data using Bitcoin Core v0.16.0 [1]. This file consists of the transactions for 9.5 years, ranging from January 3, 2009, to June 7, 2018.  We exploit the popular rusty-block-parser [2] script to decrypt and parse the encrypted blocks and transactions. The output of this process contains several comma-separated value (CSV) files that consist of the information of each block, including the timestamps, hash of blocks, transaction IDs, output addresses in a transaction, and inputs as a reference to the previous transaction output. 

With the available transaction and address information from the tool, we create mappings from transactions and addresses to numerical vertex IDs. Since each edge in the graph always connects to one address and one transaction, this graph is thus a bipartite graph.

In total, the amounts of transactions and addresses are 321,043,952 (321 million) and 399,344,697 (399 million), respectively, as well as edges are 1,692,308,191 (1.6 billion).

Please read the README.pdf file in the dataset collection to understand the dataset and how it is generated.

References:

[1] Bitcoin Core Software. Available at https://bitcoin.org/en/bitcoin-core/.

[2] Rusty block parser github repository. Available at https://github.com/gcarq/rusty-blockparser

[3] Graph Project Start. Available at https://github.com/asherliu/graph_project_start