The main file to run is retrieve_facts which is run as ./retrieve_facts facts_data

You should prepare the following beforehand:

• A folder facts_data which contains (see example folder):

1. The positive examples for a learning problem stored as a list under a file named positives.txt, see the example folder
2. The negative examples for a learning problem stored as a list under a file named negatives.txt
3. The query corresponding to the learned class expression stored under a file query.txt
4. The SPARQL endpoint where the knowledge base is loaded, stored under a file sparql_endpoint.txt

Running and storing results:

./retrieve_facts facts_data which produces two files under the current directory: positive_results.csv and negative_results.csv which contain the triples we are looking for.

What is the Python file ce_str_to_sparql.py?

This file is optional. It is only used to convert a class expression as string into a SPARQL query. This is useful if you want to test your explainer on other class expressions. However, you would need to have Ontolearn and OWLAPy installed, and modify the sparql_endpoint to yours.

enexa-cel-explainer