Leveraging semantic labels for multi-level abstraction in medical process mining and trace comparison.

Area of Science:

• Health Informatics
• Process Mining

Background:

• Medical information systems generate event logs detailing process instances.
• Analyzing these logs is crucial for understanding and improving healthcare workflows.

Purpose of the Study:

• To present a framework for abstracting event log data into higher-level concepts.
• To improve semantic process discovery and trace comparison in healthcare.

Main Methods:

• Developed an abstraction mechanism to convert event log actions into domain-knowledge-based concepts.
• Integrated abstraction with trace comparison using a similarity metric.
• Utilized semantic process discovery within the ProM framework.

Main Results:

• The abstraction mechanism handles complex situations like interleaved actions and delays.
• Trace comparison accounts for abstraction penalties and temporal constraints.
• Process discovery yields more interpretable models, demonstrated in stroke care.

Conclusions:

• The proposed framework enhances the analysis of medical event logs.
• Abstraction facilitates clearer understanding of healthcare processes and identification of deviations.
• The approach improves the interpretability of process models for medical professionals.