It is not just about describing the dataset, some interface level things also depend on domain specific knowledge (related filters, reasonable ordering of columns, descriptions of properties, everything about the dataset in other languages). These things will change as the community uses a dataset and we need to be able to improve on this.
In particular:
we do not have the knowledge to do them,
the author do not have the energy for them,
they will be regularly changed and are not part of the research product (and thus should not be part of the same provenance).
It is not just about describing the dataset, some interface level things also depend on domain specific knowledge (related filters, reasonable ordering of columns, descriptions of properties, everything about the dataset in other languages). These things will change as the community uses a dataset and we need to be able to improve on this.
In particular: