Process Mining with the Partially Ordered Workflow Language
The POWL Miner allows you to perform process discovery from event logs, leveraging the Partially Ordered Workflow Language (POWL) 2.0. The generated POWL 2.0 models can be viewed and exported as BPMNs or Petri nets (PNML). For more details on POWL 2.0, please refer to the paper: Unlocking Non-Block-Structured Decisions: Inductive Mining with Choice Graphs.
You have two options for running the POWL Miner as a Streamlit App:
Access the hosted version directly: https://powl-miner.streamlit.app/
To run the Streamlit application on your own machine:
- Clone this repository.
- Install the required dependencies ('requirements.txt') and packages ('packages.txt').
- Run:
streamlit run app.py
Alternatively, you can install it using the provided Docker image:
- Pull the Docker image:
docker pull ghcr.io/humam-kourani/powl:latest
- Run the app:
docker run -p 8501:8501 powl
You can also install the POWL Miner as a Python library to integrate its functionalities into your own scripts.
- Install the required packages ('packages.txt').
- Install the library via pip:
pip install powl
👉 Usage Example:
Check the examples/ directory of this repository.
This project bundles bpmn-auto-layout,
which is licensed under the MIT License. See THIRD_PARTY_LICENSES.txt for details.