W3C home > Mailing lists > Public > public-dwbp-wg@w3.org > February 2015

Re: Dataset description

From: Michel Dumontier <michel.dumontier@stanford.edu>
Date: Tue, 24 Feb 2015 08:19:51 -0800
Message-ID: <CALcEXf5apkyqmtFkhs1WQBFCXp=Yqvy+kJnOFHVEHN=xtLwzVQ@mail.gmail.com>
To: Phil Archer <phila@w3.org>
Cc: Public DWBP WG <public-dwbp-wg@w3.org>
Hi Phil,
  great comments, let me answer inline


On Tue, Feb 24, 2015 at 3:42 AM, Phil Archer <phila@w3.org> wrote:

> Hi again Michel,
>
> I was looking at the dataset description work. That's obviously
> interesting in this WG. I was looking in particular in some of the
> semantics which bring me to a few questions.
>
> 1. What is the reason for choosing
>
> :chembl a dctypes:Dataset .
>
> rather than
>
> :chembl a dcat:Dataset .
>
> or
>
> :chembl a dctypes:Dataset, dcat:Dataset .   ??
>
>
> we voted on it :)


> 2. This:
>
> :chembl17
>     dct:creator <https://www.ebi.ac.uk/chembl/> .
>
> Really?
>
> Not
>
> :chembl17
>     dct:creator [foaf:page <https://www.ebi.ac.uk/chembl/>] .
>
> or some such??
>
>
i agree that there should be a proper unique identifier for the creator  /
publisher. i think it was meant as a placeholder until we had a real,
authoritative URI. the solution you provide is ok in the absence of one,
but I would probably suggest creating a . i created an issue:
https://github.com/joejimbo/HCLSDatasetDescriptions/issues/107



> 3. This one is very interesting and timely:
>
> :chembl_target_targetcmpt_linkset
>     pav:authoredBy <http://orcid.org/0000-0002-8011-0300> .
>
> So you see an ORCID as a person, not a page about a person? High on my to
> do list for the coming week or so is to make good on a promise to ORCID to
> offer some advice on improving the semantics of the data they offer. One of
> the things I am *very* keen to avoid of course is the HttpRange-14 rat hole
> and my hope is that we can say that an ORCID URI identifies a landing page
> about a person. The person themselves would http://orcid.org/0000-0002-
> 8011-0300#person (or similar). i.e. a frag ID to distinguish between the
> different things that currently the same identifier is used for (person,
> profile page, account etc.)
>
> So I'm keen to know how you see this.
>
>
same as above... like you, i worry about this, but am open to suggestions
for universal deployment in the absence of failure of the authority to
support it.


> 4. Similar to 2
>
> :chembl
>     dct:publisher <http://www.ebi.ac.uk> .
>
> Hmm...
>
> :chembl
>     dct:publisher [foaf:homepage <http://www.ebi.ac.uk> ].
>
> no ??
>
>
> yup.


> 5. cf. question 2
>
> :chembl
>     foaf:page <http://www.ebi.ac.uk/chembl/> ;
>     schemaorg:logo <http://www.ebi.ac.uk/rdf/sites/ebi.ac.uk.rdf/files/
> resize/images/rdf/chembl_service_logo-146x48.gif> .
>
> So you currently have <http://www.ebi.ac.uk/chembl/> as the creator of
> the dataset and a page about it.
>
> 6. I note with interest the inclusion of citation guidance in the rights
> statement. Have you looked at the Force11 Data Citation principles? Do you
> think we should have a specific term(s) in DCAT for citation text? This is
> going to be very relevant in DWBP as we work on the data usage vocabulary
> (which is v likely to be a DCAT extension).
>
>
can you tell me more about what you were thinking?


> I know I'm being a pedant - call it a professional hazard - but I'm
> interested to see how semantic pedantry plays out in the real world. If the
> community at large is happy that the same URI identifies a creator and a
> page about something, or that every person with an ORCID has a MIME type
> then, that's the way the world is and I'll be quiet, but if not, let's see
> what we can do to help!
>

 i'm totally with you. let's get it to where we want it to be :)

m.


> Cheers
>
> Phil.
>
>
>
>
> On 22/02/2015 19:31, Michel Dumontier wrote:
>
>> Thanks Phil!
>>    I wish I could attend the call, but 6am PST is too early for me :)
>>    Looks like you've all been making steady progress, but I must say that
>> I
>> haven't been able to keep up due to other commitments. If you can point me
>> to specific documents i can have a read through and provide some feedback.
>> otherwise, if there are specific things you want  me to contribute to, let
>> me know
>>   Thanks!
>>
>> m.
>>
>>
>> Michel Dumontier, PhD
>> Associate Professor of Medicine (Biomedical Informatics)
>> Stanford University
>> http://dumontierlab.com
>>
>> On Sun, Feb 22, 2015 at 11:25 AM, Phil Archer <phila@w3.org> wrote:
>>
>>  This is likely to be of interest to this group.
>>>
>>> http://htmlpreview.github.io/?https://github.com/joejimbo/
>>> HCLSDatasetDescriptions/blob/master/Overview.html
>>>
>>> I admit it's news to me - but nonetheless welcome for that.
>>>
>>> One of the editors, Michel Dumontier, is a (so far) silent member of this
>>> WG.
>>>
>>> Something for us to take into account!
>>>
>>> Phil.
>>>
>>>
>>> --
>>>
>>>
>>> Phil Archer
>>> W3C Data Activity Lead
>>> http://www.w3.org/2013/data/
>>>
>>> http://philarcher.org
>>> +44 (0)7887 767755
>>> @philarcher1
>>>
>>>
>>>
>>
> --
>
>
> Phil Archer
> W3C Data Activity Lead
> http://www.w3.org/2013/data/
>
> http://philarcher.org
> +44 (0)7887 767755
> @philarcher1
>
Received on Tuesday, 24 February 2015 16:20:43 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 24 February 2015 16:20:44 UTC