Assessing and Refining Mappingsto RDF to Improve Dataset Quality

Abstract

Abstract. rdf dataset quality assessment is currently performed primarily after data is published. However, there is neither a systematic way to incorporate its results into the dataset nor the assessment into the publishing workflow. Adjustments are manually -but rarely-applied. Nevertheless, the root of the violations which often derive from the mappings that specify how the rdf dataset will be generated, is not identified. We suggest an incremental, iterative and uniform validation workflow for rdf datasets stemming originally from (semi-)structured data (e.g., csv, xml, json). In this work, we focus on assessing and improving their mappings. We incorporate (i) a test-driven approach for assessing the mappings instead of the rdf dataset itself, as mappings reflect how the dataset will be formed when generated; and (ii) perform semi-automatic mapping refinements based on the results of the quality assessment. The proposed workflow is applied to diverse cases, e.g., large, crowdsourced datasets such as dbpedia, or newly generated, such as iLastic. Our evaluation indicates the efficiency of our workflow, as it significantly improves the overall quality of an rdf dataset in the observed cases.

Sustainable Linked Data Generation: The Case of DBpedia

“…domain validation If the dbpedia community uses ontology terms which cause violations, there is no support for schema validation. Domain validation of the mapping templates defined in the custom dbpedia mapping language can not be supported [6]. Nevertheless, the dbpedia dataset quality would significantly improve if schema violations is applied to the mapping templates [6,13,19].…”

“…Domain validation of the mapping templates defined in the custom dbpedia mapping language can not be supported [6]. Nevertheless, the dbpedia dataset quality would significantly improve if schema violations is applied to the mapping templates [6,13,19]. Currently custom dbpedia mapping templates are only validated for syntax violations (using the Mapping Syntax Validator 15 [12]).…”

“…A corresponding underlying implementation should generate all rdf triples relying on declaratively defined mapping rules, either they refer to schema or data transformations [5]. If the dbpedia mapping rules are formalized in a generic and complete approach, the dbpedia ef mapping process and maintenance will be improved, and, thus, the dbpedia dataset quality will improve too, as it is already indicated by e.g., [6,13]. 2. modular and decoupled implementation The dbpedia community should be able to add new mapping rules for schema annotations, and alternate or add new transformation rules, as well as data transformation libraries, without requiring to adjust the underlying implementation.…”

“…Mapping Templates, i.e., rules generating most of the dbpedia dataset from Wikipedia are executed by the dbpedia ef, defined by a world-wide crowd-sourcing effort, and maintained via the DBpedia mappings wiki 6 .…”

“…We distinguish two types of issues with dbpedia: schema-level rules that define how to apply vocabularies to raw data [6,10,14], e.g., the dbo:militaryBranch property is used for entities of dbo:MilitaryUnit type, but it should only be used with entities of dbo:Person type [13]. data-level extracted, or processed and transformed data values.…”