This post is by Kefang Ding, Scientific Assistant in the Process And Data Science group at RWTH Aachen University. Contact her via email for further inquiries.
ProM is a scientific open-source platform for process mining techniques, which is popular among researchers. Many algorithms in process mining are implemented as ProM plugins. However, users interact with those plugins separately. It makes it difficult and time-consuming to conduct analyses which require multiple plugins or tests which require repeated execution of the same sequence of plugins.
Workflow management systems are software tools designed to create, perform and monitor a defined sequence of tasks. Among the current workflow management systems, KNIME is a free and open-source data analytics platform, which is implemented in Java but also allows for wrappers to run Java, Python, Perl and other frameworks.
In order to overcome the drawbacks of ProM on workflow management, as well as to enable the use of process mining techniques in KNIME, we have developed a project called PM4KNIME. PM4KNIME integrates the ProcessMining tools from ProM into KNIME platform by wrapping ProM plugins as nodes in KNIME.
To conduct tasks in Process Mining with PM4KNIME, nodes are connected to compose a workflow. Then the workflow can be executed multiple times by one click. For example, to complete the task on checking the performance, fitness and precision of an event log and a Petri net, the workflow in KNIME is shown below.
With the same tasks, compared to ProM, PM4KNIME allows an easy configuration with less interaction, easy reuse and sharing with the workflow.
The current PM4KNIME extensions implement the nodes for frequent-used process mining techniques. The PM4KNIME taxonomy is listed below.
Click on the link (https://github.com/pm4knime/pm4knime-document/wiki) to learn more about PM4KNIME! If you are interested in PM4KNIME, please contact us with kefang.ding@pads.rwth-aachen.de