Mining Product Data Models: A Case Study

This paper presents two case studies used to prove the validity of some data-flow mining algorithms. We proposed the data-flow mining algorithms because most part of mining algorithms focuses on the control-flow perspective. First case study uses event logs generated by an ERP system (Navision) after we set several trackers on the data elements needed in the process analyzed; while the second case study uses the event logs generated by YAWL system. We offered a general solution of data-flow model extraction from different data sources. In order to apply the data-flow mining algorithms the event logs must comply a certain format (using InputOutput extension). But to respect this format, a set of conversion tools is needed. We depicted the conversion tools used and how we got the data-flow models. Moreover, the data-flow model is compared to the control-flow model.