Importance of providing basisOfRecord and dcterms:type

[pzermoglio]

(These questions were posed during the second chapter of the DwC Hour, 7 Mar 2017).

Why should one provide basisOfRecord? How does not having this field inhibit research? If we provide basisOfRecord, is it necessary to provide dcterms:type? Couldn't these fields be filled in by aggregators?

Some answers were:

• It may not inhibit research, but may inhibit reuse outside our field.
• Having FossilSpecimen in basisOfRecord makes it really easy to pull only fossil records when searching.
• They could be filled in by aggregators if we as a community did some pre-work.

[tucotuco]

Very generally speaking, every field we can provide, especially if it is one of the fields that recommends standardization of some kind, helps people to find relevant information. A simple example that would inhibit research without having basisOfRecord is the capacity to filter vouchered specimens from any of the various types of non-vouchered observations.

The dcterms:type field is the one in which various kinds of Darwin Core MachineObservations (StillImage, MovingImage, Sound) can be distinguished from each other. These values belong to the controlled vocabulary for dcterms:type, not to the controlled vocabulary for basisOfRecord. So yes, to distinguish records of distinct types of media from each other, dcterms:type is needed. The Darwin Core term basisOfRecord could be filled in by aggregators for records that had any of these media filled in for dcterms:type, but not vice versa.