PraeclarusPDQ: A Reference Architecture for Process Data Quality Management

After its emergence over two decades ago, process mining flourished as a discipline. Numerous theoretical contributions have enriched the field, leading to its widespread practical applications and support from mature commercial tools. However, its potential for significant organisational impact is constrained by contemporary consideration and treatment of (poor) quality of event data. In practice, event logs tend to suffer from significant data quality problems that need to be recognised and resolved effectively for analysis results to be meaningful. Despite its significance, data quality in process mining remains underexplored. This paper proposes PraeclarusPDQ, a reference architecture for process data quality management. The reference architecture is designed to be sufficiently comprehensive and flexible to incorporate current and future contributions to the multi-faceted problem of process data quality. The reference architecture has also been instantiated as an open-source software environment aiming to galvanise the process mining community and lead to significant future breakthroughs in the theory and practice of process data quality management. Further, the reference architecture and the software prototype have been evaluated regarding their applicability, usefulness, and ease of use.