An Improved Process Discovery Approach Based on the Markov Transition Matrix

Abstract

Process mining has been widely used to discover the predefined process model from an event log. The discovered model either shows the connection of reality and original model, or indicates the conformance of the process and the event log, which can be used to discover, monitor and improve the original process.The paperaims to enhance the flexibility and adaptability of the process discovery algorithm. The paperfirst analyses the process patterns using hierarchical structure, then proposes an improved multi-step process discovery approach based on the first-order Markov transition matrix. Finally, this paper verifies the feasibility and applicability of the improved approach through a simulation example.