Mining Attribute Selection
Challenge
To run a successful process mining analysis, the minimum required input data consists of:
- CaseID – a unique identifier of the process instance
- Timestamp – a unique identifier of the time when the instance happened
- Activity – a unique identifier of action that happened to this specific case at this specific time
However, datasets usually contain also additional information such as cost centre, employee, country, region, plants, departments, and many others. During process analysis and evaluation, you might come to the point that you need to see the process map or dashboards visualizations from a different angle.
What if you need to know the process flow among employees? Or which departments create bottlenecks? Which countries are running the supply chain inefficiently?
Solution
To see your process map from a different perspective – e.g., how your process flows between departments, plants, regions, etc. instead of activities – simply choose the desired attribute in the process map.
Minit will recalculate and visualize the analysis results, including performance analysis, finance, or rework detection.
When changing the mining attribute to e.g. resource, you can investigate the cause of rework from a different angle, from the resource perspective.
You can look at the rework metric from the service center perspective or a service country when switching further. Then you can select the DeptID to see the process flow between departments.
Pro Tip: Use Machine Learning algorithms to uncover hidden connections in your process data and investigate where and why problems occur. More about AI-Powered Root Cause Analysis
Application
The Mining Attribute Selection is available to users in the customize panel, right under the map type selection. The default mining attribute is the Activity. Click the dropdown menu and choose a different attribute as the mining attribute.
The mining attribute defines the nodes and paths in the process map and also defines the variants. It doesn’t influence the statistics and the events that form the case.
Control
You can choose a different attribute in Variant analysis as well. At the Variants analysis, there’s a separate menu to choose the mining attribute from.
This way, you can visualize one mining attribute in the process map and a separate one in the variants analysis.
The clustering in the process map is also available. Minit will look at the data from the selected mining attribute perspective and find a suitable clustering attribute if available.
The selection of the mining attribute is available in Minit Dashboards through the settings in the Process Map widget.
