Explorer les transactions Bitcoin avec ORBITAAL, le plus grand jeu de données sur Bitcoin

Exploring Bitcoin Transactions with ORBITAAL, the Largest Bitcoin Dataset Ever Built

As cryptocurrencies play an increasingly important role in both the economy and scientific research, access to reliable, comprehensive, and exploitable data has become essential. ORBITAAL ( cOmpRehensive BItcoin daTaset for temporAl grAph anaLysis) addresses this need by providing the largest dataset ever built from Bitcoin transactions, enabling in-depth analysis of network dynamics over nearly 13 years. How can such a dataset open new perspectives in network science, economics, or artificial intelligence?

Researchers from the DM2L team at LIRIS (Célestin Coquidé et Rémy Cazabet) have published a paper in Scientific Data (Nature Portfolio) presenting ORBITAAL, the largest and most comprehensive dataset ever constructed from Bitcoin transactions. This dataset offers a detailed model of the evolution of the Bitcoin transaction network, covering the period from the very first transaction in January 2009 to January 2021—almost 13 years of data.

ORBITAAL stands out for its richness and precision. It provides two complementary representations of the network: snapshot graphs and stream graphs. It also includes the USD value of each transaction, computed using the Bitcoin price at the time of the transaction. Additionally, the dataset accounts for transaction fees and mining rewards, which are modeled through a dedicated node.

ORBITAAL is a reference dataset for research across multiple domains, including crypto-economics, network science, and machine learning. Potential applications include anomaly detection, user behavior prediction, and Bitcoin price modeling.

Article: https://doi.org/10.1038/s41597-025-04595-8
Dataset: https://doi.org/10.5281/zenodo.12581515