Aligning Data-Aware Declarative Process Models and Event Logs

Alignments are a conformance checking strategy quantifying the amount of deviations of a trace with respect to a process model, as well as providing optimal repairs for making the trace conformant to the process model. Data-aware alignment strategies are also gaining momentum, as they provide richer descriptions for deviance detection. Nonetheless, no technique is currently able to provide trace repair solutions in the context of data-aware declarative process models: current approaches either focus on procedural models, or numerically quantify the deviance with no proposed repair strategy. After discussing our working hypotheses, we demonstrate how such a problem can be reduced to a data-agnostic trace alignment problem, while ensuring the correctness of its solution. Finally, we show how to find such a solution leveraging Automated Planning techniques in Artificial Intelligence. Specifically, we discuss how to align traces with data-aware declarative models by adding/deleting events in the trace or by changing the attribute values attached to them.