An Incremental Process Mining Approach to Extract Knowledge from Legacy Systems

Several approaches have already been proposed to extract both business processes and business rules from a legacy source code. These approaches consider static source code analysis for the extraction procedure. However, business processes have components that can not be directly extracted by static analysis (i.e., participants, responsibilities, and concurrent activities). Moreover, well-known static analysis algorithms do not support the incremental extraction of information from the legacy code. Large legacy systems can benefit from an incremental analysis strategy in order to provide iterative information extraction as well as to achieve partial results much earlier. This paper discusses a new approach for business knowledge extraction from legacy systems. The approach considers an incremental process mining technique to extract business process structures and the business rules associated to it. Discovery results can be used in various ways by business analysts and software architects, e.g. documentation of legacy systems or for re-engineering purposes.