A Vocabulary for Describing Mapping Quality Assessment, Refinement and Validation

Abstract

As the number of RDF datasets being published on the Web increases, so does the need to establish the quality of their contents. The generation of RDF datasets may be performed by transforming non-RDF resources to RDF. This process is termed “uplift”, and often involves the definition of mappings where the rules on how data is transformed to RDF are described. Another use for mappings is found when defining relationships, or links, between concepts within those datasets. The definition of these mappings involves various quality related activities in which these are iteratively refined until they satisfy the user s expressed requirements. Information related to such requirements and quality related mapping activities, however, is rarely captured or provided. In this paper, we present the Mapping Quality Vocabulary (MQV), which enables quality metadata and provenance information relating to the assessment, refinement and validation of mappings to be captured and published. We furthermore present a demonstration of the proposed model applied to a use case. It is expected that the generation and publication of such metadata and provenance information will facilitate mapping maintenance and reuse, whilst also providing the means for data consumers to evaluate mapping and dataset quality.