[Application-profiles-ig] DataCite example

Wed Jun 3 14:37:27 BST 2020

OK, I got a start on modelling the example. There are some issues I've run into already, which might be interesting to discuss.https://docs.google.com/spreadsheets/d/1F0zGQcHKTM9rq0P7Tz1E9qSMRsyG97l-xAV0czyKxAA/edit?usp=sharing首先,由于哒taCite Schema (and this is an older version (3.1) vs. current 4.3) started as an XML schema, there are subelements that are not always modelled in the "RDF" way and some elements that take either a Literal or a URI. I've added comments on some cells. You can see the XML influence also in the use of xsi:type to record the URIs of controlled vocabulary sources (i.e., these would be XML attributes in an element tag). I tried to take the modelling in the AP "as is" and translate it into this new format. Later it might be interesting to make an attempt from stratch. My doc is more of a worksheet at this point, as I've added a couple extra columns to preserve columns from the source while I'm working on it. (Clause Number and "Dublin Core or Other schema Property" which is the "mapping" information I am translating into the URI column (this is so you can see how I translated it). I combined two columns into the Comments column: "Definition" and "Values, examples, constraints" There is also reference to using this document with the "DataCite Ontology." That could be this SPAR authored ontology, but I'm not sure.https://sparontologies.github.io/datacite/current/datacite.htmlAnyway, I can see the challenges involved here of starting with an established XML schema for a custom metadata schema, and then trying to move it to a mapping/application profile format. Not easy! John -- *John Huck*, MLIS Metadata Librarian, University of Alberta Library +1 (780) 248-1337 (office) 5-25E Cameron Library, University of Alberta, T6G 2J8 ᐊᒥᐢᑲᐧᒋᐋᐧᐢᑲᐦᐃᑲᐣ / Amiskwaciwâskahikan / Edmonton *The University of Alberta respectfully acknowledges that we are situated on Treaty 6 territory, traditional lands of First Nations and Métis people.* On Tue, 2 Jun 2020 at 21:36, Karen Coyle <kcoylenet at gmail.com> wrote: >Great, John, thanks. - kc>>On Tue, Jun 2, 2020 at 8:35 PM John Huck <john.huck at ualberta.ca> wrote:>>>Hello all,>>>>I finally have the link to the DataCite DC Application profile (highest>>draft number, I presume). So I'll try to work up the example.>>>>>>https://github.com/datacite/schema/blob/master/source/DataCite%20Dublin%20Core%20AP%20-%20Draft%201_8.pdf>>>>On Sat, 30 May 2020 at 12:13, Karen Coyle <kcoyle at kcoyle.net> wrote:>>>>>Actually, mapping between metadata sets came up in our requirements and>>>use cases. See:>>>>>>https://github.com/dcmi/dcap/issues/36>>>>>>>>Ah, yes, I see. Thanks!>>>>Cheers,>>>>John>>>>On Sat, 30 May 2020 at 12:13, Karen Coyle <kcoyle at kcoyle.net> wrote:>>>>>>>>>>>On 5/28/20 11:48 AM, John Huck wrote:>>>> Thanks, Karen.>>>>>>>> This also raises the question of when to draw the line between>>>> application profiles and straight-forward mappings/crosswalks. There's>>>a>>>> connection/continuum, of course, and while I tend to think there won't>>>> be many hopelessly ambiguous cases (with a little care in>>>discernment),>>>> I wonder if some people who create APs might do so as a first step or>>>> part of a larger process to develop mappings, and so there might be a>>>> connection between those activities.>>>>>>John, good comments.>>>>>>>>>>>>>>In that use case, as I understand it, the AP would be used to select>>>metadata from external sources such that it would be the most compatible>>>with whatever other metadata sources one is processing. I also think>>>that Nishad may have examples similar to this. So this is worth looking>>>at when we get to the step of extensions to the simple profile. What>>>makes this more complex is that it seems to often require decisions: "if>>>no x:id then use a y:name if available". One-to-one mapping could simply>>>be another column, but as we know it's rarely that neat.>>>>>>>>>>> More concretely, though, this does remind me of a question I had about>>>> the Label property in our AP model. Do we need to clarify whether the>>>> label is the (or one of the) original label(s) of the predicate, or>>>the>>>> label for element within the context or implementation of the>>>> application? Do we need two label properties? Or do we not especially>>>> care about the original label? I tend to think people will want to>>>> record their implementation label. You could end up with a wide>>>semantic>>>> gap between predicate and label if people perceive it as a way to>>>> "customize their metadata", but that could be addressed in best>>>> practices. What do others think?>>>>>>Oh, yes, this is one of the cans of worms waiting for us. Your example>>>using the label is a good one, and one that we could probably tackle in>>>our "best practices" document (which we will have to write), but it>>>brings up the bigger question of how to prevent contradictions between>>>the original vocabulary terms and their use in an AP. This reuse>>>question applies not only to APs but all reuse and it's something that I>>>don't think has been addressed in any metadata universe. So let's see>>>what we can conclude about labels, and then how we will step neatly away>>>from the bigger question, which is not something we can resolve.>>>>>>kc>>>>>>>>>>> As an example, here's the Dataverse metadata implementation>>>spreadsheet>>>> <>>>https://docs.google.com/spreadsheets/d/10Luzti7svVTVKTA-px27oq3RxCUM-QbiTkm8iMd5C54/edit#gid=0>.>>>>>>> Dataverse implements DDI in a database structure as a core, and then>>>> adds in domain blocks from other standards (so this could be another>>>> example of a hybrid profile/crosswalk). Column A is the field label>>>and>>>> D/E are the database field name. The metadata elements Dataverse>>>> implements are not all RDF properties, so it sits outside our current>>>> scope, but it's just for illustration.>>>>>>>> Cheers,>>>>>>>> John>>>>>>>> -->>>约翰·哈克> * *,信息学硕士学历>>>> Metadata Librarian, University of Alberta Library>>>> +1 (780) 248-1337 (office)>>>> 5-25E Cameron Library, University of Alberta, T6G 2J8>>>> ᐊᒥᐢᑲᐧᒋᐋᐧᐢᑲᐦᐃᑲᐣ / Amiskwaciwâskahikan / Edmonton>>>> /联合国iversity of Alberta respectfully acknowledges that we are>>>> situated on Treaty 6 territory, traditional lands of First Nations and>>>> Métis people./>>>>>>>>>>>> On Thu, 28 May 2020 at 11:31, Karen Coyle <kcoyle at kcoyle.net>>>> kcoyle at kcoyle.net>> wrote:>>>>>>>> Thanks, John. I've run into similar "oops!" with both the Wikidata>>>> example and ContentDM. I'm thinking that we might want to use some>>>of>>>> these less compatible examples to show a counter example using an>>>> application profile, and hopefully that will show the advantages>>>of the>>>> AP approach.>>>>>>>> Thanks for volunteering to use your own project as an example. I'm>>>> looking forward to seeing that!>>>>>>>> kc>>>>>>>> On 5/27/20 12:58 PM, John Huck wrote:>>>> > Hello all,>>>> >>>>> > When I suggested DataCite as an example, I forgot that it's not>>>> strictly>>>> > speaking a profile of DC terms, it's just closely based on them>>>> (not the>>>> > first time I have forgotten this fact either!).>>>> >>>>> > There is mention of a DC application profile/mapping in the>>>> > documentation>>>> >>>>> <>>>https://schema.datacite.org/meta/kernel-4.3/doc/DataCite-MetadataKernel_v4.3.pdf>>>> (p.4)>>>>>>>> > and collaboration with DCMI SAM>>>> > <//www.voudr.com/groups/sam/>, but this may have>>>been a>>>> > project that was left incomplete for one reason or another,>>>> judging from>>>> > what I can find. It seems they were close to having one ready in>>>> 2016>>>> > for DataCite 3.x, but I can't see it was released and an online>>>> comment>>>> >>>>> <>>>https://blog.datacite.org/datacite-to-dublin-core-application-profile/>>>>>>>>> > last year said there was no work ongoing on it. I've emailed>>>their>>>> > support for more info, so we'll see.>>>> >>>>> > So, I could try to create a mapping to DC from the schema and>>>model>>>> > that, but for our present purposes, I think that could be more>>>time>>>> > consuming than it's worth, so I will focus on modelling the>>>other>>>> >我提到的项目(我自己),开发APof geospatial>>>> predicates.>>>> >>>>> > I'll just observe that the DataCite/DC situation seems like a>>>> case of a>>>> > project that could have benefited from a simpler way to create>>>an>>>> > application profile, so it's an interesting example in that>>>sense.>>>> >>>>> >欢呼,>>>> >>>>> > John>>>> >>>>> > -->>>> > *John Huck*, MLIS>>>> > Metadata Librarian, University of Alberta Library>>>> > + 1(780) 248 - 1337(办公室)>>>> > 5-25E Cameron Library, University of Alberta, T6G 2J8>>>> > ᐊᒥᐢᑲᐧᒋᐋᐧᐢᑲᐦᐃᑲᐣ / Amiskwaciwâskahikan / Edmonton>>>> > /The University of Alberta respectfully acknowledges that we are>>>> > situated on Treaty 6 territory, traditional lands of First>>>> Nations and>>>> > Métis people./>>>> >>>>> >>>>> > On Tue, 26 May 2020 at 11:42, Karen Coyle <kcoyle at kcoyle.net>>>> kcoyle at kcoyle.net>>>>> > kcoyle at kcoyle.net< mailto:kcoyle at kcoyle.net>>> wrote:>>>> >>>>> > Here are decisions and issues from the May 20 meeting, which>>>> have been>>>> > added to the meeting agenda and minutes [1].>>>> >>>>> >>>>> > 1. Group decided to focus first on APs for RDF, and to later>>>> determine>>>> > if the result can be generalized to non-RDF metadata.>>>> > 2. Examples will be developed for:>>>> > schema.org <http://schema.org> <http://schema.org> - Phil>>>> > Wikidata - ?>>>> > contentDM? - KC>>>> > BIBFRAME - KC>>>> > DataCite - John H>>>> >>>>> > Some areas that need more discussion are:>>>> >>>>> > 1. Where to put namespace columns and what to call those>>>columns>>>> > 2. Should entity descriptions be on their own row?>>>> (Advantages: can>>>> > annotate, can give cardinality)>>>> > 3. Can one designate properties as usable anywhere (aka:>>>> > free-floaters)?>>>> > (Idea: place them first before first entity)>>>> > 4. There is still the question of IDs for properties (KC>>>will>>>> > attempt an>>>> > example)>>>> >>>>> > kc>>>> > [1]>>>> >>>>>>>>https://github.com/dcmi/dcap/blob/master/meetings/2020/2020-05-20.dcap_zoom_call.md>>>> > -->>>> > Karen Coyle>>>> >kcoyle at kcoyle.net< mailto:kcoyle at kcoyle.net>>>>> kcoyle at kcoyle.net< mailto:kcoyle at kcoyle.net>>>>>http://kcoyle.net>>>> > skype: kcoylenet>>>> >>>>>>>>> -->>>> Karen Coyle>>>>kcoyle at kcoyle.net< mailto:kcoyle at kcoyle.net>http://kcoyle.net>>>> skype: kcoylenet>>>>>>>>>>>>-->>>Karen Coyle>>>kcoyle at kcoyle.nethttp://kcoyle.net>>>skype: kcoylenet>>>>>-->-- --->Karen Coyle / Digital Library Consultant>kcoyle at kcoyle.nethttp://www.kcoyle.net>skype: kcoylenet>mo.: 510-435-8234>------------------------------------>-------------- next part -------------- An HTML attachment was scrubbed... URL: <https://lists.www.voudr.com/pipermail/application-profiles-ig/attachments/20200603/ce7fe7fb/attachment-0001.htm>