Modelling of not "ready to publish" Datasets - Do we need something else than the dcat classes?
A Dataset in DCat is more of a bibliographic entry in a catalog than a representation of some file with data inside. The question arises, if the Dataset class of DCat has the right meaning for a dataset that is an input or output of a processing step. Would Distribution be better? Or do we need some other class to model the role of datasets that are generated or used within research processes?