ExplainableMine

With ExplainableMine, we contribute to seminal work on event log data quality. We will devise approaches on explainable distributed outlierness for process mining by interweaving outlier detection techniques along the complete analytic pipeline from sensor data to processes.
Within the field of process mining, data quality is one of the most urgent problems hampering the direct application of process mining techniques on event data. The potential of process mining for significant organisational impact is constrained by contemporary consideration and treatment of (poor) quality of event data. Event logs in practice tend to suffer from significant data quality problems and these need to be recognized and resolved effectively for analysis results to be meaningful. Particularly, the presence of outliers strongly impacts the quality of the discovered process model. In data engineering, outliers, commonly referred to as ’anomalies’, refer to something that is out of range. ExplainableMine will develop an outlier quantification framework making the analysis results and sourced process mining explainable.

Verantwortlich für die Redaktion: Christian Imenkamp