-
cwd should be src/
-
rule output: first atom is head, rest is body
In ./Data/KG add a folder containing an .nt file with the same name as the folder.
In src/Data/Constraints add a folder containing a .ttl file containing SHACL-constraints both with the same name as used before. If there are SPARQL querys in the constraints please use "$this" when referencing the target node. Also requires writing "SELECT ($this AS ?this)". This is needed for error free validation with TravSHACL.
In ./Data/Constraints add a .ttl file containing the ontology. Here the name doesn't matter.
"KG": name of KG "prefix": namespace prefix "rules_file": rule file to write to (will be generated if not exists) "rdf_file": .nt file in KG folder, must be same name as KG, "constraints_folder": also same name as KG, "ontology_file": name of ontology file .ttl, "max_body_length": leave empty for default of 3, "example_set_size": num of pos/ neg examples per target predicate, leave empty for default 15 "type_predicate": leave empty for default of 'http://www.w3.org/1999/02/22-rdf-syntax-ns#type', "alpha": alpha used in weight formula, leave empty for default 0.5, "mine_negative_rules": leave empty for False --> will mine positive rules, put anything for True
use this as input.json for testing
{ "KG": "musicKG", "prefix": "http://example.org/", "rules_file": "musicKG.csv", "rdf_file": "musicKG.nt", "constraints_folder": "musicKG", "ontology_file": "musicKGOntology.ttl", "max_body_length": "", "example_set_size": "", "type_predicate": "", "alpha": "", "mine_negative_rules": "" }