Improved Alignment Repair
MSc Thesis Project
Title: Improved Alignment Repair
Author: Zheqi Lyu
1st examiner: Prof.dr.ir. Wil M.P. van der Aalst
2nd examiner: Prof. Dr. rer. nat. Martin Grohe
Daily supervisor: Dr.ir. Sebastiaan J. van Zelst
Summary
Most organizations, across different domains, e.g., banking and healthcare, execute various business processes. It is in the organization’s interest to assess what execution ordering the process has. Hence, it is vital to analyze the degree of consistency of the process execution with the corresponding process specifications. Conformance Checking Technique, a part of Process Mining, analyzes the agreement’s extent of a process model with observed behaviors of the business process. Currently, Alignments are the de-facto standard for conformance checking. Unfortunately, computing alignments is a combinatorial process. As a consequence, the researchers consider obtaining a repaired alignment of a model based on an existing optimal alignment of the other model. In this thesis, we present two approaches through expanding the parts that no longer fit the given process model to compute an enhanced repaired alignment by given two process models and an existing alignment. Our experiments show that this repaired alignment approximates to the optimal and the trade-off between loss of the optimality and time is reasonable.