Conformance checking helps us understand how well the process behavior recorded in event logs matches the process behavior prescribed by a process model. The goal of this project is to advance conformance checking methods by integrating timed and stochastic aspects of process behavior and optimizing alignments across entire event logs. We will:
- Analyze Effects of Timed and Probabilistic Features of a Process on the Quality of Alignments: Review and validate existing algorithms that generate trace-specific alignments and explore their limitations in handling time-sensitive and stochastic behaviors.
- Develop Enhanced Global Alignment Algorithms: Design and implement new algorithms that optimize the alignment for the entire event log, considering all traces, while also taking into account time-dependent variations and stochastic behavior.
- Create Realistic Test Data for Evaluation: Develop methods for generating synthetic event logs that reflect typical behavioral deviations, timing variations, and stochastic elements to thoroughly test the effectiveness of the global alignment algorithms.
- Contribute to Process Mining Tools: Enhance existing process mining tools by integrating these improved conformance checking techniques. These tools will enable companies and researchers to assess process models more effectively under realistic, time-sensitive, and uncertain conditions, considering the event log as a whole.