Process Mining is a process management technique that enables business processes to be reconstructed and evaluated on the basis of digital traces in IT systems. Process Mining is a discipline of business data analytics designed to improve processes in a company’s operations. The individual steps of the process stored in the systems are combined and the process is visualized in its entirety. Process Mining makes it possible to model the implicit and otherwise hidden process knowledge contained in data, making it tangible and transportable. The technique is often used when other approaches fail to formalize the processes or when the quality of existing process records is questionable. Contemporary management trends such as Business Activity Monitoring (BAM), Business Operations Management (BOM), and Business Process Intelligence (BPI), for example, show that there is a great deal of interest in further developing the opportunities for analysis in this area.
Fields of application
In principle, process mining can be used wherever individual steps of a process are stored in an IT system so that the coherence and chronology of the steps can be traced. This traceability is ensured by a process or process log. This is especially true for workflows that are stored and managed in workflow management systems. A workflow is a formally described business process that can be coordinated and controlled by a workflow management system. User interfaces allow users to interact with the system and store and edit individual steps in a workflow. The totality of the stored steps ultimately results in a process that can be lifted and reconstructed using process mining. So z. For example, the transactions from ERP systems, the history of tickets in a ticketing system, or clinical pathways of hospital patients are presented. Key application areas of process mining are process harmonization across various organizational units and companies, process optimization in terms of throughput times, process costs, process stability and ensuring compliance requirements. Further application possibilities for process mining can be found, for example, in knowledge management or in assistance systems. An application case of process mining would also be, for example, too long ordering processes in purchasing due to too long release times of the departments.
Process mining can be seen as a link between data mining and business process management. However, unlike data mining, process mining focuses on leveraging implicit process knowledge already contained in the data.
The starting point for process mining is a collection of data in which individual process steps are stored. The quality of this data is very important for process mining. A set of statistical models is now applied to this data, with the aid of which the standard course of the process (core process) is determined. This core process then serves as the basis for the rest of the process and enables deviations from the standard process to be determined.
Classes of Process Mining
The Task Force on Process Mining of the Institute of Electrical and Electronic Engineers IEEE, headquartered in New York, defines three different classes of process mining:
- Discovery : Existing process flow logs are used to reconstruct the processes contained therein without first having information or models of existing processes. Process Mining is used here for the pure enhancement of existing processes. This type of process mining application is currently the best known.
- Conformance : In this type of process mining, a model already exists via a process flow. Existing data will now be tested for conformance to the existing model based on the model and existing process-log process logs.
- Enhancement : Here, too, the process logs and a model of the existing process are already available. In contrast to the conformance type, however, not only theory and practice are to be checked for their conformity, but the existing model must be adapted and expanded if necessary. Ideally, this approach leads to a new, better model of the desired process.