This document provides a summary of a research paper on conformance checking between designed and observed processes. It includes:
- An introduction to process mining quality measures like fitness, precision, generalization, and simplicity.
- A paper that proposes a new entropy-based precision and recall measure for process mining that satisfies important properties. This measure uses the topological entropy of regular languages to calculate precision and recall between event logs and models encoded as deterministic finite automata.
- An evaluation showing the new measure works for both synthetic and real-life event logs and outperforms existing approaches at fulfilling important precision properties. The implementation also scales to large real-life logs.