PM4Py is an open-source process mining library, primarily intended to be used as a data science / research scripting library. However, to showcase the possibilities of process mining with PM4Py, i.e., far beyond just a scripting language, we have build the Process Mining ToolKit (PMTK). PMTK is a web-based process mining solution that allows you to perform various different process mining analyses on top of your event data. Note that, the main intent of PMTK is to serve as a vehicle for our process mining research, i.e., we use the framework to implement advanced process mining technology in a user-friendly, cloud-based, envrionment. As such, PMTK is not a commercial product, i.e., usage of PMTK will be combined with a joint research project. At this moment, access to our demo server and/or the stand-alone version is only granted upon request. Interested? Feel free to contact us.

Process Statistics

After loading your event data, PMTKis able to show you various different statistics of your process, including:

  • Occurrence plots of the activities in the process
  • Occurrence plots of the number of events per case
  • Density plots of executed events over time
  • Density plots of newly started process instances over time
  • Overview of the number of active process instances over time
  • Various performance (throughput time) statistics
Process Statistics Page

Variant Inspector

PMTK has a dedicated variant explorer, showing all the different variants of the process

  • Inspect executions following a selected variant
  • Select and filter on a subset of process variants
Variants Explorer

Dotted Chart

PMTK contains the dotted chart analysis, i.e., a simple yet powerful event data visualization. In the dotted chart, the event data can be visualized in a variety of ways. Typically, the events (colors indicate the activity) are plotted over time, per process instance. However, virtually any event attribute can be plotted against any other attribute. Additionally, the dotted chart can be used to filter the data!

  • Visualize correlations between arbitrary data attributes
  • Select and filter on a subset of process variants
  • Interactive zoom-in/zoom-out
  • Event Filtering
Dotted Chart Analysis

Process Discovery

PMTK can discover process models, and, process maps! Furthermore, PMTK supports frequency and performance projection on top of the process models! The process models are highly interactive and allow us to filter the event data, e.g., selecting all traces that contain a specific selected activity.

  • Process Map Discovery
  • Process Model Discovery
  • Frequency / Performance Projection
  • Interactive Filtering
Process Discovery