The essence of functional programming on semantic data Martin Leinberger 0 Ralf Lämmel 1 Steffen Staab 0 2 Institute for Web Science and Technologies, University of Koblenz-Landau , Germany The Software Languages Team, University of Koblenz-Landau , Germany Web and Internet Science Research Group, University of Southampton , England -

Programming with knowledge represented in description logics (DL) is errorprone. Untyped access, e.g., provided by the OWL API [ 1 ], does not leverage static typing which allows for proving the absence of runtime-errors. Mapping approaches, e.g., described by [ 2 ] cannot fully capture the conceptualization of semantic data. In [ 3 ], we present DL, a typed -calculus with constructs for operating on semantic data. This is achieved by the integration of description logics into the -calculus for both typing and data access.

We rely on ALCOI as a basic description logic. We assume the possibility of checking whether axioms follow logically from a knowledge base as well as class expression queries. Furthermore, we only allow named individuals as query results to avoid infinitely large result sets.

Key design principles of DL are to treat (1) concept expressions as types in the programming language (2) subtype inference by forwarding typing judgments to the knowledge system as axioms during type checking, (3) typing of queries to ensure satisfiability and proper result processing, (4) class expression queries as well as (5) open-world querying in which we treat axioms only as true if they are true in all models of the knowledge system.

Runtime semantics are modeled as a small-step operational semantics. As usual, runtime errors are modeled through stuck states during evaluation. Using the key design principles, we show that a straightforward extension of a basic -calculus, is sufficient to achieve a type-safe integration. The only exception comprises of accessing empty query results.

While the presented approach shows how a basic integration supporting subtyping and queries can be achieved, future work focuses on an extended query language, in particular, a subset of SPARQL, as well as type inference for all types of DL and polymorphism.

 1. M. Horridge and S. Bechhofer . The OWL API: A Java API for OWL ontologies . Semantic Web , 2 ( 1 ): 11 - 21 , 2011 . 2. A. Kalyanpur , D. J. Pastor , S. Battle , and J. A. Padget . Automatic Mapping of OWL Ontologies into Java . In Proc. SEKE 2004 , pages 98 - 103 , 2004 . 3. M. Leinberger , R. Lämmel , and S. Staab . The essence of functional programming on semantic data . In Proc. ESOP 2017 , LNCS. Springer.