W3C home > Mailing lists > Public > semantic-web@w3.org > July 2007

Re: How to name what you get back? was: Terminology Question concerning Web Architecture and Linked Data

From: Pat Hayes <phayes@ihmc.us>
Date: Mon, 30 Jul 2007 12:56:37 -0500
Message-Id: <p0623090ac2d3d621f8a6@[]>
To: "John Black" <JohnBlack@kashori.com>
Cc: "Chris Bizer" <chris@bizer.de>, "Frank Manola" <fmanola@acm.org>, "Tim Berners-Lee" <timbl@w3.org>, <www-tag@w3.org>, <semantic-web@w3.org>, "Linking Open Data" <linking-open-data@simile.mit.edu>

>Chris Bizer wrote:
>>Hi Frank, Pat and Bernard,
>>thanks a lot for all your ideas and comments.
>>So what we are having on the table right now is a definition:
>>"The term XXX refers to the description of a 
>>non-information resource that a client obtains 
>>by dereferencing a specific URI that identifies 
>>this non-information resource."
>>and various proposals for the term:
>>Pat Hayes:
>>- represented description
>>- redirected description
>>- redescription
>>- transmit
>>- infon
>>Frank Manola
>>- associated representation
>>Bernard Vatant:
>>- description
>If you want to maintain the idea of 
>"identifying" a resource, then I would suggest 
>the term "Definite Description", 
>that means something like this: a set of logical 
>statements about a universe of discourse such 
>that there is one individual or one set of 
>individuals, and only one individual or set of 
>individuals, about which all those statements 
>are true.

I would strongly suggest not doing this, on the 
grounds that the thing we are talking about will 
almost never be a definite description. It is 
dangerous to co-opt precise terms from another 
discipline unless you are prepared to accept the 
likely consequences that they bring with them.

>I once began to compile a list of all the ideas 
>I came across for what you should name what you 
>get back, 
>Among the terms I thought were interesting were 
>"Discriminant Description", "Concise Bounded 
>Description", "Rdf Neighbor", "Ctx Meaning", 
>"Wordnet Synset", "Web Proper Name".

It will almost never be a Wordnet Synset or a 
proper name, either. Again, these are precise 
technical terms.


>>Some (of course subjective) comments: Like 
>>Bernard already stated, "redescription" sounds 
>>temporal. You describe something, then you 
>>redescribe it afterwards. Transmit and infon 
>>sound like Sci-Fi to me. Redirected description 
>>is a good explanation of what is happening, but 
>>maybe a bit to technical, process-oriented.
>>I think Frank's term "representation" does not 
>>work from the technical side, as you get 
>>redirected to an information resource, which 
>>has a representation and this representation 
>>contains the data we are talking about. But the 
>>representation of this information resource 
>>might also contain lots of data about other 
>>resources. For instance, when you think about a 
>>vocabulary definition as in the "Best Practices 
>>for Publishing RDF Vocabularies" guide 
>>http://www.w3.org/TR/swbp-vocab-pub/. I like 
>>Frank's idea of using the word "associated" as 
>>this term in also used throughout the TAG 
>>"Dereferencing HTTP URIs" document 
>>So, when I try to merge all these ideas and 
>>thoughts, I end up with "associated 
>>What do you think about this term?
>>Anybody strongly disagreeing?
>>A problem that remains with this term is that 
>>it is rather technical and therefore does not 
>>work well as a replacement for data item in the 
>>introduction of our tutorial.
>>For instance: The sentence "A basic tenet of 
>>Linked Data is to use RDF links to interlink 
>>data items from different data sources." would 
>>become "A basic tenet of Linked Data is to use 
>>RDF links to interlink associated descriptions 
>>from different data sources." Cough, cough :-)
>>But anyhow, the term "associated description" 
>>will work in the remaining technical chapters 
>>of the tutorial and we can save ourselves in 
>>the introduction by saying "A basic tenet of 
>>Linked Data is to use RDF links to interlink 
>>data from different data sources".
>>Chris Bizer
>>Freie Universitšt Berlin
>>Phone: +49 30 838 54057
>>Mail: chris@bizer.de
>>Web: www.bizer.de
>>----- Original Message ----- From: "Frank Manola" <fmanola@acm.org>
>>To: "Chris Bizer" <chris@bizer.de>
>>Cc: "Tim Berners-Lee" <timbl@w3.org>; 
>><www-tag@w3.org>; <semantic-web@w3.org>; 
>>"Linking Open Data" 
>>Sent: Wednesday, July 25, 2007 10:44 PM
>>Subject: Re: Terminology Question concerning Web Architecture and Linked Data
>>This probably indicates that I haven't gone back far enough in the
>>prior discussion (or still don't understand various details of the
>>Web architecture), but what's wrong with "representation"?  That is,
>>you're asking for "a term for "the information which you get about
>>the thing identified by it when you look up a URI".  I thought what
>>you got back when you dereferenced a URI was a "representation".  And
>>I thought that the difference between dereferencing the URI of an
>>information resource vs. that of a non-information resource is that:
>>(a) for a non-information resource there's supposed to be some
>>indirection messaging that goes on between the original dereferencing
>>and you getting a representation back, and
>>(b) the representation you get back doesn't contain all the
>>"essential characteristics" of the identified resource.
>>But you still get a representation back.  That is in fact what your
>>tutorial says. Under the heading "Dereferencing HTTP URIs", the last
>>sentence of the bullet that describes what happens for non-
>>information resources says "In a second step, the client dereferences
>>this new URI and *gets a representation* describing the original non-
>>information resource" [my emphasis].
>>I think I understand the sort of distinction you're getting at,
>>something like, as Pat suggests, the difference between getting a
>>copy (of sorts) of the thing itself vs. getting a "description" (in
>>some sense) of it, but if that's it, I'm not sure an entirely new
>>piece of terminology is what's needed.  Operationally the only way we
>>have of knowing whether a URI names an information resource or a non-
>>information resource is (at least in the tutorial) whether
>>redirection happens when we dereference it.  In the scenarios we're
>>talking about, the redirection is (as I understand it) ultimately to
>>the URI of an *information resource* (with its own URI) that
>>describes (in some sense) the original non-information resource.  The
>>draft TAG finding a http://www.w3.org/2001/tag/doc/
>>httpRange-14/2007-05-31/HttpRange-14 refers to this redirected-to
>>information resource as an "associated information resource" (without
>>necessarily intending to introduce that as a defined term).
>>Following those lines though, how about calling what you're after an
>>"associated representation"?  That is, when you dereference the URI
>>of a non-information resource, what you get back (after redirection)
>>is an "associated representation" (the details of the association
>>being determined by the details of the redirection).
>>On Jul 25, 2007, at 12:44 PM, Chris Bizer wrote:
>>>Hi Frank,
>>>>I'd seriously suggest you look for some 
>>>>alternative  to "data item" for the concept 
>>>>in question.
>>>OK, but this leads to a question which I 
>>>accutally wanted to try to avoid asking on 
>>>this list.
>>>Hmm, I will do it anyway and see what happens ;-)
>>>Question 4: What term should we use instead?
>>>Chris Bizer
>>>Freie Universitšt Berlin
>>>Phone: +49 30 838 54057
>>>Mail: chris@bizer.de
>>>Web: www.bizer.de
>>>----- Original Message ----- From: "Frank Manola" <fmanola@acm.org>
>>>To: "Chris Bizer" <chris@bizer.de>
>>>Cc: "Tim Berners-Lee" <timbl@w3.org>; 
>>><www-tag@w3.org>; <semantic- web@w3.org>; 
>>>"Linking Open Data" 
>>>Sent: Wednesday, July 25, 2007 6:07 PM
>>>Subject: Re: Terminology Question concerning 
>>>Web Architecture and Linked Data
>>>>I appreciate that we run into terminology 
>>>>conflicts all the time around here, but I'd 
>>>>seriously suggest you look for some 
>>>>alternative to "data item" for the concept in 
>>>>question.  An awful lot of people 
>>>>(particularly those involved with databases) 
>>>>are used to seeing "data  item" refer to 
>>>>something like a property or attribute (like 
>>>>"name" or  "age").  More specifically, 
>>>>they're used to seeing records as  containing 
>>>>multiple data items (or their values).  From 
>>>>that point of  view, the sentence "When you 
>>>>interpret the Web of Data as a set of 
>>>>interlinked databases, a data item would 
>>>>equal a record in a specific  database." 
>>>>looks particularly strange.  As I say, I 
>>>>understand the  inevitability of terminology 
>>>>conflicts, but ...?
>>>>On Jul 25, 2007, at 10:12 AM, Chris Bizer wrote:
>>>>>Hi Tim,
>>>>>>I can't think of a term for "the 
>>>>>>information which you get about the thing 
>>>>>>identified by it  when you look up a URI" 
>>>>>>which works for me.
>>>>>>It has of course the term "Representation" 
>>>>>>which connects an Information Resource and 
>>>>>>the (metadata, bits) pair which you get 
>>>>>>back, which is different.
>>>>>As we did not want to repeat the definition 
>>>>>all over the tutorial, we ended up with a 
>>>>>term called "data item".
>>>>>Within section 2.1 of the tutorial, we 
>>>>>define the term as: "The term data items 
>>>>>refers to the description of a 
>>>>>non-information resource that a client 
>>>>>obtains by dereferencing a specific URI that 
>>>>>identifies this non-information resource." 
>>>>>(http:// sites.wiwiss.fu- 
>>>>>Note that the definition is a bit more 
>>>>>specific than your sentence above, as it is 
>>>>>restricted to non-information resources and 
>>>>>not things in general (assuming that your 
>>>>>term "thing" refers to non- information 
>>>>>resources as well as information resources).
>>>>>We were also struggling to find a good word 
>>>>>that matches the concept and have chosen 
>>>>>"data item" in the end as it somehow relates 
>>>>>to the overall term "Linked Data" and as we 
>>>>>hope that people from the database community 
>>>>>will understand the second informal 
>>>>>definition of the term: "When you interpret 
>>>>>the Web of Data as a set of interlinked 
>>>>>databases, a data item would equal a record 
>>>>>in a specific database."

IHMC		(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.	(850)202 4416   office
Pensacola			(850)202 4440   fax
FL 32502			(850)291 0667    cell
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes
Received on Monday, 30 July 2007 18:10:22 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 07:41:58 UTC