End-to-End Process Extraction in Process Unaware Systems

Knowledge of current business processes is a critical requirement for organizational initiatives like compliance management, regulatory reporting, process optimization, reengineering the IT systems and outsourcing. Existing process discovery techniques expect process execution information or event logs while organization's business processes are often executed on heterogeneous systems across different departments, by integration and data hand-offs between systems. Traditional information systems, however, are designed for storing and processing transaction data which persists in databases and other data storage mechanisms. In this paper we identify the challenges and propose a solution for extracting end-to-end processes from persistent process execution data available in multiple heterogeneous applications. The approach consists of analyzing persistent system data to identify and obtain events in a non-intrusive manner. The approach to get the end-to-end process involves a combination of data and process mining.