Object-Centric Process Mining for Blockchain Applications. Extracting and Representing Ethereum Execution Data in OCEL 2.0

Analyzing the execution behavior of Ethereum Decentralized Applications (DApps) with process mining presents significant challenges due to the multi-object nature of DApp data. Traditional event logs, like XES, struggle to capture the respective structures and interactions effectively. This paper proposes a method for extracting DApp execution data from Ethereum and representing it in the Object-Centric Event Log (OCEL) 2.0 format. We address central challenges in this process, including dynamic contract deployments, preserving the order of operations within transaction traces, and accurately representing object types and their evolving roles. Our findings demonstrate that, while OCEL 2.0 offers some advantages for capturing the rich interactions within DApps, certain limitations regarding hierarchical object types and event granularity require workarounds. We evaluate the practicality of our approach with a case study of the prediction market platform Augur, highlighting how object-centric process mining can provide insights into DApp behavior. This work contributes to a better understanding of object-centric process mining in the context of blockchain data.