Process Discovery describes the data-based visualization of a process and is a basic method in Process Mining. The model is usually generated automatically from the available data (event logs). The discovered model is often displayed as a Direct Follower Graph. Primary goal is the creation of transparency and the attainment of process knowledge.
Are there specific requirements for Process Discovery?
The basic prerequisite for this is the availability of the data points. The data points should look as follows: Identification of the data point, name of the activity or event, time stamp for the start and end time of the activity, and any number of other attributes (these depend on the available data).
Without these data points the generation of a Direct Follower Graph is not possible, because the Process Mining system does not “know” what the process looks like. If data is available, but not in the mentioned form, a data transformation must be carried out. Whether this data is available at all must be discussed with the system user or data owner. The consent of the data owner is also required for data access. As a rule, the company is the data owner. In the case of personal data, the consent of the works committee is also required. As soon as the permission for data use is available, the desired data can be exported, transformed if necessary and imported into the Process Mining system. The Process Discovery is then carried out with the imported data.
How does Process Discovery work?
As explained earlier, Process Discovery is performed as follows:
1. Check availability of data points
2. Obtain permission from the data owner and, if applicable, the works committee
3. Extract data
4. Transform data
5. Import data into Process Mining system
6. Direct Follower Graph is generated
« Back to Glossary Index