RDFS(c) Schema Inconsistency Checking Based on a Key Instance and Its Query Rewriting

Abstract

The introduction of a constraint mechanism to the resource description framework schema (RDFS) poses challenges for keeping RDFS ontologies consistent, particularly in the presence of reasoning rules (RRs). This study presents a method to detect inconsistencies in a given constraint enhanced RDFS, termed RDFS <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$_{\mathrm {(c)}}$ </tex-math></inline-formula> , under a given set of RRs. This method can be used to efficiently infer the type qualifications and constraints that may be violated when the RRs are applied to the graph instances of the schema. First, the concept of an internal schema is developed, and the conversion of an RDFS <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$_{\mathrm {(c)}}$ </tex-math></inline-formula> to an internal schema is examined. An internal schema is more versatile than the corresponding original schema language. Second, the concept of key instances of an internal schema is established as a basis for detecting schema inconsistencies. The number of key instances is considerably smaller than that of original instances, resulting in a significantly higher detection efficiency. Moreover, a rigorous formal proof is given to ensure that the schema generated based on the key instances is semantically equivalent to the original schema. Finally, an algorithm is provided to detect schema inconsistencies in the key instances of the corresponding internal schema. This algorithm extends the key instances by query rewriting and constructs potential instances that violate the constraints. The minimality of rewriting ensures that a minimum number of instances are constructed, thereby further improving the efficiency for inferring constraint violations. Test results show that key instances and query rewriting can be used to significantly enhance the efficiency of detecting inconsistencies and that the proposed method is superior to relevant methods in terms of time performance.