PLG - Processes and Logs Generator
Multiperspective Processes Randomization and Simulation for Online and Offline Settings
Introduction
Evaluating process mining algorithms could require the availability of a suite of real-world business processes and their execution logs, which hardly are available.
Process Log Generator is a application capable to generate random business processes, starting from some general "complexity paramenters". PLG is also able to "execute" a given process model in order to generate a process log.
This software is designed to help researchers in the construction of a large set of processes and corresponding execution logs. This software is released with a small library which could help in the programmatical creation of processes.
Main features
- Random process generation, with different complexity parameters
- Random process evolution (to generate slight variations of existing processes)
- Configuration of time for activities duration and time between activities (via Python scripts)
- Generation of static/dynamic data objects for multi-perspective event log generation (via Python scripts)
- Import of process from
- PLG file format
- BPMN files (generated from SAP Signavio)
- Export of generated processes as
- PLG file format
- BPMN 2.0 XML file
- BPMN as Graphviz Dot file
- Petri net as Graphviz Dot file
- Petri net as LoLA file
- Petri net as PNML file
- Petri net as TPN file
- Generation of an event log with any number of traces
- Fine-tuned configuration of noise parameters for event log generation
- Export of the generated event log as
- XES file (both compressed as
.xes.gz
and not compressed as.xes
) - MXML file (both compressed as
.mxml.gz
and not compressed as.mxml
)
- XES file (both compressed as
- Generation of an infinite stream of events
- Event streams generated as MQTT-XES format (cf. https://www.beamline.cloud/mqtt-xes/)
- Ability to dynamically switch the process generting the events (to simulate concept drift in streams)
- Generation of noise into the stream
Screencast
Download
You can read or download the updated Java source code on the project GitHub (https://github.com/delas/plg). Sources of the old releases of PLG are still available on the old GitHub project (https://github.com/delas/plg-old).
PLG makes use of the following libraries:
libPlg
: library for processes and event generationlibPlgStream
: library for stream generationlibPlgVisualizer
: library for process visualization
Older versions of PLG:
- PLGLib-1.4beta
- Release date: 2011-07-22
- Download this release (sources and binary)
- Release notes:
PLG-1.4-beta.zip
the application with graphic interfacePLG-CLI-ProcessCreator-0.1.jar
the command line version of the random process generatorPLG-CLI-LogCreator-0.1.zip
the command line version of the process simulator (and log generator)
- PLGLib-1.2
- Download this release (sources and binary)
- PLGLib-1.1
- Download this release (sources and binary)
- PLGLib-1.0
- Download this release (sources and binary)
Help and Contact
You can find more information on how to use the tool and how to configure it, on the GihHub Wiki pages. You can find all the contact information at the author's webpage.
Citation
Please, cite this work as:
- Andrea Burattin. "PLG2: Multiperspective Process Randomization with Online and Offline Simulations". In Online Proceedings of the BPM Demo Track 2016; Rio de Janeiro, Brasil; September, 18 2016; CEUR-WS.org 2016.
Other relevant publications:
- Andrea Burattin. "PLG2: Multiperspective Processes Randomization and Simulation for Online and Offline Settings". In CoRR abs/1506.08415, Jun. 2015.
- Andrea Burattin and Alessandro Sperduti. "PLG: a Framework for the Generation of Business Process Models and their Execution Logs". In Proceedings of the 6th International Workshop on Business Process Intelligence (BPI 2010); Stevens Institute of Technology; Hoboken, New Jersey, USA; September 13, 2010.10.1007/978-3-642-20511-8_20.