Bitcoin dust transactions

General information

This repository contains data regarding Bitcoin dust transactions. In the Bitcoin protocol, dust refers to the small amounts of currency that are lower than the fee required to spend them in a transaction. 

The repository comprises all transactions with at least one dust output or input. According to our definition, a dust output (or input) is considered dust if the associated amount is between 1 and 545 satoshis (where 1 satoshi = 10-8 bitcoin). For more details about the definition of dust, see [1].

All dust transactions have been extracted from the first 479,970 blocks of the Bitcoin blockchain, thus covering the time period between January 3rd, 2009 18:15 GMT and August 10th, 2017 18:03 GMT.

Data set description

File
Description

txs
A text file containing a representation of all Bitcoin transactions that create and consume dust. See the description below for more information about the structure of this file.

txs_addr_map.csv
A CSV file that maps numeric address identifiers to real Bitcoin addresses. This file comprises all addresses appearing in the txs data set.

labels.csv
A CSV file containing categorical entity labels for Bitcoin addresses appeared in transactions between 2010 and 2018. This file has been derived from the Entity-Address data set [2, 3] (see also: https://github.com/Maru92/EntityAddressBitcoin).

outputs_spent_stats.csv 
A CSV file containing statistics about all spent outputs in the first 479970 blocks of the Bitcoin blockchain. The file describes the durations of dust and non-dust outputs. The duration is defined as the difference between the height of the block where the output is spent and the height of the block where it was created. 

cluster_sizes_*.csv
These CSV files contain information about clusters of addresses induced by Bitcoin transactions. They have been used for the clustering analysis presented in [4]. See this GitHub repository for more information.

Transaction representation

The txs file contains a textual representation of dust transactions in the Bitcoin blockchain. Each row of the file corresponds to a transaction and is represented as a sequence of fields

info:inputs:outputs

with the following meaning.

The info section contains general information about the transaction. It is represented as a list of comma-separated fields, namely: timestamp,blockId,txId,isCoinbase,fee,approxSize.

The meaning of the fields is the following:

timestamp represents the Unix timestamp of the block containing the transaction.

blockId represents the height of the block containing the transaction.

txId is a numeric value that univocally identifies the transaction.

isCoinbase is equal to 1 if the transaction is a coinbase transaction, 0 otherwise.

fee denotes the transaction fee, expressed in satoshis (i.e., the smallest bitcoin denomination).

approximateSize denotes the approximate size of the transaction (expressed in bytes).

The inputs section contains a sequence of (0 or more) transaction inputs separated by a semicolon. Each input, in turn, is represented as a comma-separated string addrId,amount,prevTxId,offset where:

addrId represents the numeric identifier of the spending address;

amount is the amount of value associated with the input (expressed in satoshis);

prevTxId represents the numeric identifier of the transaction that created the output that is currently being spent;

offset represents the position, among all outputs of prevTxId, of the output that is currently being spent.

The outputs section contains a sequence of (1 or more) transaction outputs separated by a semicolon. Each output, in turn, is represented as a comma-separated string addrId,amount,scriptType where:

addrId represents the numeric identifier of the receiving address;

amount is the amount of value associated with the output (expressed in satoshis);

scriptType is a numeric identifier representing the type of the script associated with the output (i.e., 0=UNKNOWN; 1=P2PK; 2=P2PKH; 3=P2SH; 4=RETURN; 5=EMPTY).

Data analysis

Data included in this repository have been employed for the analyses presented in [4, 5]. This GitHub repository contains several tools, written in Java and Python, for analyzing the data.

Cite this work

If the data included in this repository have been useful, please cite the following article in your work.

@article{loporchio2023bitcoin,
  title={Is Bitcoin gathering dust? An analysis of low-amount Bitcoin transactions},
  author={Loporchio, Matteo and Bernasconi, Anna and Di Francesco Maesa, Damiano and Ricci, Laura},
  journal={Applied Network Science},
  volume={8},
  number={1},
  pages={1--28},
  year={2023},
  publisher={SpringerOpen}
}

References

Pérez-Solà, Cristina, et al. "Another coin bites the dust: an analysis of dust in UTXO-based cryptocurrencies." Royal Society open science 6.1 (2019): 180817.

Jourdan, Marc, et al. "Characterizing entities in the bitcoin blockchain." 2018 IEEE international conference on data mining workshops (ICDMW). IEEE, 2018.

Jourdan, Marc, et al. "A probabilistic model of the bitcoin blockchain." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. 2019.

Loporchio, Matteo, et al. "Is Bitcoin gathering dust? An analysis of low-amount Bitcoin transactions." Applied Network Science 8.1 (2023): 1-28.

Loporchio, Matteo, et al. "An Analysis of Bitcoin Dust Through Authenticated Queries." Complex Networks and Their Applications XI: Proceedings of The Eleventh International Conference on Complex Networks and their Applications: COMPLEX NETWORKS 2022—Volume 2. Cham: Springer International Publishing, 2023.