Sorted hyper-predicate knowledge bases for ontologies and rules
Ken Kaneiwa, Philip H. P. Nguyen · 2012
In the Semantic Web context, the meta-modeling of concepts in ontologies and rules is required for defining the meanings of upper-level data and vocabularies. For this purpose, we propose a decidable fragment of order-sorted logic programming with hyper-predicate hierarchies. In the sorted language, we can employ simple- and hyper-predicates to express the concepts of sorts, predicates, and meta-predicates by unifying and enhancing them in a hyper-predicate hierarchy. The sorted Horn-clause calculus is extended to develop a query-answering system that can answer queries such as atoms and hyper-atoms generalized to contain predicate variables. We show that each query is computable in NEXPTIME if knowledge bases are function-free and safe. Furthermore, the computation can be reduced to EXPTIME if additional arguments in super-predicates and multiple predicate declarations are restricted.