How to model and express provenance?

Issue

The DCAT model treats the descriptions of datasets in a catalogue as entities that only exist in the context of the catalogue, and does not consider situations where these descriptions are imported from and exported to other catalogues.

In an environment where descriptions of datasets are exchanged among data portals, the situation that DCAT-AP is designed for, it may be important for users to understand where data comes from and how it may have been modified along the way. For example, it could support credibility of a dataset to know which organisation created the metadata for it in the first place and how the description was modified along a chain of exchanges.

DCAT-AP specifies an optional property dct:provenance for Dataset but does not provide any guidance on how to describe instances of the class dct:ProvenanceStatement.

Current situation

Only few national implementations of DCAT-AP use dct:provenance to provide provenance information and that is in free text.

Recommendation

Rationale

In the absence of commonly agreed approaches, ignoring provenance information does neither help nor hinder interoperability.

Example

The example is based on the Nobel Prize catalogue, which is available via http://www.nobelprize.org/datasets/dcat. Some modifications were made in order to clarify the guideline.