Towards a vocabulary for data quality management in semantic web architectures

Abstract

Reliable decision-making and reliable information based on Semantic Web data requires methodologies and techniques for managing the quality of the published data. To make things more complicated, the judgment of what is good data will often depend on the task at hand or the subjective requirements of data owners or data consumers. Some data quality requirements can be modeled using data quality rules, i.e. executable definitions that allow the identification and measurement of data quality problems. In this paper, we provide a conceptual model that allows the representation of such rules and other quality-related knowledge using the Resource Description Framework (RDF) and the Web Ontology Language (OWL). Based on our model, it is possible to monitor and assess the quality of data sources and to automate data cleansing tasks. The use of a generic conceptual model based on Semantic Web formalisms supports the definition of reusable, broadly applicable SPARQL queries and portable applications for data quality management (DQM). Furthermore, the explicit representation of rules in RDF/OWL facilitates rule management tasks, e.g. for analyzing consistency among the rules, and allows to collaborate and create a shared understanding.