define use of basisOfRecord

[fdschneider]

Darwin Core uses basisOfRecord as a very fundamental marker for the referenced object of a single data row. The field takes only Darwin Core 'classes'.

i.e. in context of a traitdata table, the entry would always be MeasurementOrFact. The resolution we are intending is concerning the single occurence at which the trait was measured, i.e. a specimen that is alive, or preserved or stored in a museum, or a more virtual term if there hasn't been a individual occurence, but rather a meta-information extracted from literature or expert knowledge.

[caterinap]

That sounds quite reasonable and also fits well with some of the botany data where they have e.g. a leave from a branch from a tree

[fdschneider]

If basisOfRecord refers not to the main data table of measurements, but rather to the occurenceID level of information, the definition of DWC and what we had in mind matches again. Thus, I would define basisOfRecord as a data column in the Occurence Extension, i.e. it describes an individual specimen, even if many measurements have been taken on it. (This would also argue for renaming specimenID into occurenceID, see #13 )

[nadjasimons]

Both your suggestions sound reasonable to me, I agree to the new names.

[aostrow]

DwC differs here, e.g. LivingSpecimenis possible as entry. Does it make sense to distinguish further instead of using just MeasurementOrFact.

[fdschneider]

I think we use it in exactly the same way as described for MeasurementOrFact Extension. However, we have a couple of options that are not offered by the controlled vocabulary suggested there. For instance we should specify "ExpertKnowledge" or "Database" as possible entries.

We said once that it should be a mandatory field, but I don't remember why exactly. Its not essential for primary trait datsets, rather for the metadata.

However, as I said, I don't see it in the core data section, but rather in the occurence Extension. It does not really fit into our measurementOrFact category, since multiple traits measured on the same specimen all receive the same entry in BasisOfRecord. I guess this 'problem' will be further clarfied when you see the R script.

[fdschneider]

I now put originOfRecord into extension MeasurementOrFact. http://fdschneider.de/bexis_traits/traitdatastandard.html#basisofrecord ; I think it makes most sense on the side of information resolution. Authors might also chose to link it to the occurenceID to describe the kind of specimen observed. In practice this will not make much difference, though. I think that will do it for the beginning.