Efficient Validation of SHACL Shapes with Reasoning

As the usage of knowledge graphs (KGs) becomes more pervasive in practical applications, there is a burgeoning need for high-quality data. The SHApes Constraint Language(SHACL) allows for expressing certain types of quality constraints that define sub-structures and correct values in KGs modelled with RDF. Nevertheless, per forming SHACL validation without entailment often yields one sided outcomes, as it falls short of validating crucial implicit data encoded in the KG ontology. Current solutions that incorporate entailment into SHACL validation are inefficient, due to the time intensive process of applying inference rules to the entire dataset. Moreover, applying entailment for SHACL validation can generate large amounts of redundant triples, exacerbating the validation work load and resulting in erroneous or redundant validation results. In light of these challenges, we propose Re-SHACL, an approach that combines targeted reasoning and entity merging techniques to generate a concise, consolidated RDF graph devoid of redundancy. Re-SHACL significantly reduces execution time and improves the accuracy of the validation reports. Our experiments demonstrate that Re-SHACL can be combined with state-of-the-art validators to deliver accurate validation reports efficiently.