- From: LJ.Garcia <lj.garcia.co@gmail.com>
- Date: Sun, 5 Feb 2023 11:12:03 +0100
- To: "Clark, Timothy (twc8q)" <twc8q@virginia.edu>
- Cc: Daniel Arend <arendd@ipk-gatersleben.de>, public-bioschemas <public-bioschemas@w3.org>, "Garcia-Castro, Leyla Jael" <ljgarcia@zbmed.de>, "sneumann@ipb-halle.de" <sneumann@ipb-halle.de>, "Beier, Sebastian" <s.beier@fz-juelich.de>
- Message-ID: <CAPZUG=AGyKJY9OJFycy=PC-qksa_oEJGuNrmOaSNy8nOdQKKVA@mail.gmail.com>
Dear Tim, Thanks for your input. We have created a GitHub Issue "Discussion: how to describe the nature of the content for Datasets?" <https://github.com/BioSchemas/specifications/issues/629> to gather input, I think some people have added some comments there. If you are interested in the topic, please have a look. wrt your comments. Bioschemas already provides a Dataset schema specification (https://bioschemas.org/profiles/Dataset/) which includes a property "description" for free text descriptions of a Dataset. The discussion at GitHub Issue "Discussion: how to describe the nature of the content for Datasets?" <https://github.com/BioSchemas/specifications/issues/629> in more about a quick way to describe (e.g., express the metadata corresponding to) the content of the dataset, e.g., a dataset of proteomics for butterflies. Kind regards, On Tue, Jan 24, 2023 at 9:38 PM Clark, Timothy (twc8q) <twc8q@virginia.edu> wrote: > My thought on this is that you have multiple kinds / levels of content > description. > > (1) *Full text description*, similar to an “abstract” of the dataset. > Explains to humans what the dataset is concretely, and can serve to aid in > searches. This is contained in “Bibliographic” style metadata - as in > Datacite - and obviously goes along with other relevant terminology such as > “version” etc. > (2) *Guide metadata*, i.e. keywords, using controlled vocabularies, which > are specifically designed to aid in tuning searches - use of controlled > vocabularies will enable synomyms etc. to ensure semantics included in > search. > (3) *Dataset schema specifications*. There are various approaches to > doing this, but essentially you serialize names, descriptions, datatypes, > and controlled vocabulary term alignments for each data element in the > dataset. These not only enable proper reuse of the datasets but also > strengthen description specificity, as in “this dataset contains data with > these attributes”. > > Cheers, > > Tim > > > On Jan 24, 2023, at 1:06 PM, LJ.Garcia <lj.garcia.co@gmail.com> wrote: > > Hi Bioschemas community, > > Some of you have expressed interest in adding information about the nature > of the content of your Dataset. For instance, if your Dataset is compiling > information about a particular Taxon, you would want to add that > information to the description (i.e., metadata) of your Dataset. @Daniel > Arend <arendd@ipk-gatersleben.de> this would be your use case. > > There is, so far, no consensus at Bioschemas on how to do so. We have > created a GitHub Issue "Discussion: how to describe the nature of the > content for Datasets?" > <https://github.com/BioSchemas/specifications/issues/629> to gather > options and views on the subject, please contribute to the discussion. > > Kind regards, > Leyla Jael Castro > Bioschemas Chair > > >
Received on Sunday, 5 February 2023 10:12:27 UTC