Yes. In TaxonWorks we have a multi-step process where we transcribe first into an equivalent of “verbatimLabel” and then records are parsed. So the records you are seeing are “in-progress” and have not yet been parsed.
1 Like
@datafixer this makes me smile (and perhaps wryly). It’s exactly this kind of thing that collection managers need help with. They need ways to “see” these issues in their data. Yes, they could export data and use the command line. I’d suggest they need better tools in their local CMS to help them. (see for example Secret Sauce to Visible Sauce! Controlled Vocabularies - #3 by Debbie). For at least some of what you found, it seems like there was a column shift b/c of a separator issue (that is, localities or other values like dates in the scientificName column). Perhaps folks could make more use of the dwc gbif validation tool that would be startled to find such data types in the sciName column.
@ymgan at some point there will be the MIDS standard implemented at GBIF which will look at some select fields and allow you to find records that have “something” in them (or not). It does not assess fidelity of these data, only presence of or absence of characters in a give (select) term. Using it, you could filter out records that are NULL for a given field, helping you narrow your dataset more effectively.
1 Like
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.