unibz-core/constraintlearning

ConstraintLearning Paper: Data and Processes

ConstraintLearning

ConstraintLearning-2021 Paper: Data and Processes.

Quickstart Guide

Install python 3.x.

Install the following libraries:

$ pip install problog
$ pip install probfoil
$ pip install regex
$ pip install pandas
$ pip install owlready2

Take any OntoUML model generated through Visual Paradigm, install the onto-vp-plugin to convert the model into .ttl (or .rdf, .owl) format (see related model examples).

Run the the gufo2alloy converter to generate Alloy .als specifications (see related model examples).

Customize the simulations scope. Run simulations. Save the output simulations into .txt files.

Use the input model (.rdf format) and the generated .txt as inputs of the probfoil converter.

Generate .data and .settings probfoil input files with the corresponding .py files:

python file.py > output.txt

Play with the generated probfoil data and elicit neg/pos. For the probfoil set-up see related guidelines.

As an example see the output rules described in the paper.