Enriched Conformance Checking

The project implements an enriched conformance checking for an hybrid process model capable of representing human behavior.

Features

The project is composed by four main python scripts:

• Mining: executes discovery and conformance of the control flow dimension.
• Statsdata: executes discovery and conformance of the data dimension.
• Discovery: to run the discovery of the multi dimension model.
• Conformance: to run the conformance on the multi dimension model.

And by two jar files:

• dcr-discovery: discovery of DCR graphs (see https://github.com/tslaats/DisCoveR).
• dcr-conformance: conformance of DCR graphs.

Installation

Use the package manager pip to install the following packages.

  pip install numpy
  pip install pandas
  pip install scipy
  pip install matplotlib
  pip install pm4py

Install Declare4py, the version used in this project is the following https://github.com/francxx96/declare4py.

Usage/Examples

The approach is now divided into discovery and conformance. To run the discovery execute:

python Discovery.py path_xes path_discovered_model

• path_xes is the path of the event log used to discover the reference model (without extension)
• path_discovered_model is the path of the folder where to save the discovered models

For example:

python Discovery.py Example/Scenario2/logNormal Example/Scenario2

All the models derived by the application of the process discovery algorithms are saved in the folder /Models.

To run the conformance execute:

python Conformance.py path_xes_test path_xes_train path_discovered_model, case, exp_name

• path_xes_test is the path of the event log used to for the conformance (without extension)
• path_xes_train is the path of the event log previously used for the discovery
• path_discovered_model is the path of the folder where the discovered models are saved, i.e. the Models folder
• case and exp_name are attributes used to distinguish the different runs in the Results.csv file (case is a int, exp_name is the experiment name and it's a string)

For example:

python Conformance.py Example/Scenario2/logAsence Example/Scenario2/logNormal Example/Scenario2/Models 1 test1

The fitness values obtained by the application of the confromance checking are saved in the Results.csv file.

Badges