mjy
commented
4 years ago Hi @mabecabrera, thanks for your question! The short answer- provenance and improved semantics that will lead to interoperability betwen your dataset and others.
Most of the user customization inherits from TaxonWorks "Controlled Vocabulary Term".
There we are encouraging curators to spend a couple minutes thinking about adding something that has long term consequences to curatorial workflows, i.e. to think about what it means to tag something. It's often far more subtle than we think at first glance. For example, which of the many different versions of 'female' do you mean? http://bioportal.bioontology.org/search?utf8=✓&query=female. I suspect it's likely a tag based on examination of terminalia (or not- perhaps exual dimorphism in antennae), as opposed to a genetic assay. This might seen pedantic, but it can make a difference. If you are creating predicates it could be this bad: https://github.com/tdwg/dwc-qa/blob/master/data/GBIFDistinctValues/GBIF_distinct_sex_2017-02-27.csv. Ouch!
A good definition describes not only what you mean by the term or label, but also how a user should consistently apply it. I.e. a naive curator could read it and act in a consistent manner without talking to the person who created it. This is important for downstream computing/statistics/data merging etc. etc.
Finally, we strongly encourage curators to add not only a definition, but also find a URI that points to the definition you'd like others to find. See the bioportal query above. If you do this it means: things with this value are the same as someone data somewhere else that also points at the URI, you can combine them.
Hope this helps.
mabecabrera
I would like to know why is it necessary to have a minimum of 20 characters for descriptions while existing some categories that can be very shortly described. For example: female or egg