Provenance-aware query execution in open linked data systems

Provenance is suggested as a concept which directs the trustworthiness and quality assurance process of data on the Web. The applications publishing and consuming linked data represent definitions about trust with the help of provenance metadata. In the thesis, filtered data obtained from queries concerning metadata of datasets is elaborated. To enhance trustworthiness, it is worked on filtering most suitable datasets with provenance metadata to be queried and transferring these datasets to federated query engine. The first contribution of this thesis is to introduce a provenance-aware linked data querying infrastructure. In the infrastructure, the representation of provenance metadata and the process during querying linked data by the organization is described. The second contribution of this thesis is to filter most appropriate datasets between available datasets towards provenance metadata. In the work, metadata about datasets are used for filtering. An architecture benefited from metadata and required vocabularies is represented byconsidering organizational preferences and dataset-based requirements. The third contribution of this thesis is to append Interior Data and Exterior Data concepts to the linked data literature. When observing linked data from organizational perspective, whole accessible raw or semantic data on the web and the metadata of those data is called as Exterior Data. Interior Data is the aggregation of organization data, organizational preferences and metadata of accessible datasets on the web which is referred to organizational data source metamodel warehouse. The fourth contribution of this thesis is to express provenance metadata for extending dataset metadata which belongs to Exterior Data. Provenance definitions related to raw dataset, semantic dataset and dataset metadata are clarified. The fifth contribution of this thesis is to build an ontology that considers organizational and provenance concepts constructing Interior Data. The ontology includes terms for querying provenance independent of the domain. Additionally, it helps organizations to have ideas about each other. The sixth contribution is to build another ontology that represents provenance preferences of the organizations. While there are different provenance dimensions and related different provenance preferences, sample preference dimensions involving different conditions are handled. Accordingly, method of creating organizational preferences are exemplified. The last contribution of this thesis is to generate rules by the help of Internal Data, External Data and built ontologies. Final list of datasets used by query engine are obtained by enforcing rules on available datasets.