Data Quality
Process-mining results depend on the data behind them. This research studies how missing or incorrect information changes the conclusions we draw, and how to trace those problems back to their source.
LIQUID: Linking Data Quality Issues to Deviations in IoT-Enhanced Process Mining
EDOC 2026 · Accepted / in press · Best Paper Award.
LIQUID preserves information about data-quality problems when sensor readings are transformed into event logs. It then traces apparent process deviations back to possible problems in the original sensor data.
A Framework for Measuring Data Quality Sensitivity in Process Discovery
EdbA at ICPM 2025 · Process Mining Workshops (2027).
We introduce a framework for measuring how data-quality issues affect the process models discovered from event logs. Its configurable pipeline lets users test how sensitive discovery methods are to different kinds of data errors.
For work on preparing event data and discovering object-centric processes, see object-centric process mining.