Towards Object-Centric Process Mining for Blockchain Applications.

Process mining event logs are traditionally formatted to reflect the execution of a collection of individual process instances, with a fixed case notion. In practice, process instances are often intertwined, and the scope of a particular process is less static. When flattening complex application data to traditional event log formats, like XES, problems such as divergence and convergence occur in the resulting event logs. Object-centric process mining with its object-centric event log (OCEL) format were introduced to tackle these issues, supporting multiple case notions with a single event log. While the adoption of object-centric logging is starting to gain momentum in several domains, the use of OCEL for event logs of blockchain applications has seen little research. In this paper, we investigate blockchain data structures and map them to OCEL logging capabilities. We present an approach to extracting data from blockchain applications that requires little domain knowledge. We discuss how to map data items to fit object-centric event logs and provide and analyze a corresponding OCEL event log for a blockchain application. The approach is evaluated based on a use case and contrasted to a previous case study.