Semantics and Validation of Shapes Schemas for RDF

TL;DR

This paper formalizes the semantics of a new expressive schema language for RDF graphs, enabling precise validation and description of graph structures with recursive and Boolean features.

Contribution

It introduces a formal semantics and proof of soundness for Shape Expressions Language 2.0, including validation algorithms and implementation guidelines.

Findings

Semantics proven to be well-defined under certain restrictions

Two algorithms for schema validation presented

Implementation guidelines provided

Abstract

We present a formal semantics and proof of soundness for shapes schemas, an expressive schema language for RDF graphs that is the foundation of Shape Expressions Language 2.0. It can be used to describe the vocabulary and the structure of an RDF graph, and to constrain the admissible properties and values for nodes in that graph. The language defines a typing mechanism called shapes against which nodes of the graph can be checked. It includes an algebraic grouping operator, a choice operator and cardinality constraints for the number of allowed occurrences of a property. Shapes can be combined using Boolean operators, and can use possibly recursive references to other shapes. We describe the syntax of the language and define its semantics. The semantics is proven to be well-defined for schemas that satisfy a reasonable syntactic restriction, namely stratified use of negation and recursion. We present two algorithms for the validation of an RDF graph against a shapes schema. The first algorithm is a direct implementation of the semantics, whereas the second is a non-trivial improvement. We also briefly give implementation guidelines.