What is process mining?
process mining is a technique in which business processes are extracted from the information system events and analyzed. This is a practice for managing business processes employed to discover new processes, compare the existing process with a workflow model and improve the process. Mining of event log data can provide valuable information that may not be obtained by other methods.
There are three categories of process mining. The first is the model of discovery, so it is called because it involves discovering previously unknown or undocumented processes. This type of data mining is done if there is no model for workflow or when it is known that the existing documentation is defective. The event protocols are then mined for information that is analyzed to recreate this process. The documentation is then created for this process based on data extracted from event protocols.
The second type of process mining is the conformation model.V has been derived from its purpose of checking whether the ongoing workflow corresponds to the planned process. Event logs are mined data to find the differences between the existing process and the model.
Once such differences are placed, they are analyzed to see if they have improved this process. If such changes prove to be beneficial for this process, the model is revised to include these deviations. The decisions made on procedural checkpoints are reviewed regarding information available at any point and limits affecting such decisions. If such changes are disadvantageous, then changes in the existing process can be made to make it easier to adapt to the model.
The third class of process mining is the extension model. This type of data mining is trying to expand the existing model with improvement. Data from event logs are analyzed from possible areas of structuresUry model. For example, narrow places can be checked for possible alternative routes in the workflow.
process mining is not without problems. Some tasks are always hidden from the event logs and can not be mined by data. These can be reconstructed using a careful analysis of visible tasks, but not always. Conclusions based solely on information pulled out of events may therefore be doubtful.
Duplicate tasks in the event protocol also create problems because there may be different activities in the same category or name. Therefore, it may be difficult to distinguish the tasks of the same name from each other, even though they have different functions. Other problems include corresponding decision -making data, incorporating time into the model, various perspectives, incorrectly recorded data and simply insufficient information. Process mining must be alleviated by experience and good judgment to overcome drought problems when applying this technique.