A Framework for Assessing and Refining the Quality of R2RML mappings

Abstract

"Uplift" mapping execution applies a set of mapping definitions to transform non-RDF data sources to RDF. During the uplift mapping process, mapping definitions are iteratively refined until they conform with the user's expressed requirements. The W3C standard R2RML is one language which allows for specifying the mappings needed to generate RDF datasets from relational databases. Many approaches have been proposed to assess the quality of the generated RDF datasets, even though the root cause of several of those quality violations are found in mappings. In this paper, we present a framework for assessing and refining the quality of the definitions used to transform non-RDF data to RDF. This paper also provides an overview of an implementation of the proposed quality assessment framework for the R2RML mapping language, which uses the W3C standard Shapes Constraint Language (SHACL). We also provide a demonstration of the proposed framework and its implementation through a walkthrough use case.