W3C home > Mailing lists > Public > public-esw-thes@w3.org > August 2010

RE: FOAF spec revised - addtion of foaf:focus, a skos extension linking topical and factual information

From: Tudhope D S (AT) <dstudhope@glam.ac.uk>
Date: Thu, 12 Aug 2010 09:42:06 +0000
To: "public-esw-thes@w3.org" <public-esw-thes@w3.org>
Message-ID: <EB99149E0DAB564782E935BD3F19F37A2333A35A@MailDB1.uni.glam.ac.uk>
Hi all

I wonder if  completely avoiding overlap with words that are 'taken' is unnecessarily restrictive? 'About', 'Subject', Concept', etc build on a long tradition and have been mentioned as relevant in the discussion. Is there any mileage in considering further qualifying any one of these in a longer term, as with Al's skos:conceptFor ?



From: public-esw-thes-request@w3.org [public-esw-thes-request@w3.org] on behalf of Antoine Isaac [aisaac@few.vu.nl]
Sent: 12 August 2010 09:22
To: public-esw-thes@w3.org
Subject: Re: FOAF spec revised - addtion of foaf:focus, a skos extension        linking topical and factual information

Hi everyone,

How to resist such discussion?? At least I won't try to add stuff, but instead have a shoot at some of the proposals around :p

1. I think we have to be careful with whatever carries the idea of "name" or "term". There might be collision with other ways of looking at (term-based) KOSs (e.g., http://nkos.slis.kent.edu/FRSAR/index.html). Which includes in fact SKOS(XL): a "termFor" would rather apply for something skosxl:Label, I think. "conceptFor" is less dangerous!

2. Among Alistair's proposals there are some which sound like (soft) similarity (simile, analogue) at least in my ear of non-native. Maybe this is not ideal if the entities linked do not belong to the same level (KOS vs. "real-world").
"incarnation" stuff sounds much better, Actually it reminds me of Bernard's "praesens", which I like a lot (esp. for the interesting collision with the cool and modern "foaf" namespace ;-) ).

Otherwise "proxy" is used by ORE (http://www.openarchives.org/ore/1.0/datamodel#Proxies) and in the Europeana context we almost routinely use SKOSification to refer to the transformation of legacy KOS representation to brand new SKOS resoources ;-) so these ones are taken.!



> Hi Alistair, I do not resist to hitchike in this galaxy:
> Dan's original definition:
> "foaf:focus links a skos:Concept to the thing that the concept stands for"
> This link being defined within the namespace of the representation
> system of the destination, it means that a similar relation term may be
> useful for other applications than FOAF and that this relation term may
> represent the SKOS contribution in the RDF network.
> So what "essence" SKOS brings to the representation of a person, of some
> thing? It does not "document", it does not "explain" (like OWL), it does
> not "link". For me, the SKOS concept provides the terminology to make
> the person or thing discoverable / identifiable by humans. It "names" (a
> very ambiguous term).
> It also "orders" things in a BT/NT pyramid. Something not so agreable
> for the FOAFed humans!
> Some proposals:
> foaf:namesFor
> foaf:termsFor
> foaf:controlledTermsFor
> foaf:thesaurusEntryFor
> foaf:skosConceptFor
> or skos:conceptFor ?
> The proof has been done again that people exchange much more knowledge
> when they have the excuse of discussing terminology than when (serious)
> topics are put directly on the table: at the end we may have finalized
> the way we want to relate SKOS concepts to their alter ego in other
> representation systems...
> Question: skos:conceptFor (reciprocal foaf:skosConcept) or skos:termsFor
> (reciprocal foaf:skosTerms) would not be easierFor:everybody?
> Good evening!
> Christophe
> Le 11/08/2010 18:19, Alistair Miles a écrit :
>> Hi all,
>> FWIW, I'm glad this is in FOAF, I think it's a useful pattern, and I
>> could live with "foaf:focus" given a few good examples in the
>> documentation.
>> If a decision is made to change the name, the only thing I'd suggest
>> is to steer clear of anything that might have any formal connotations
>> regarding language or logic, and to keep it informal, quirky and if
>> possible entirely novel.
>> The bottom line is this is a useful pattern (e.g., where two blogs
>> have a category for the Dalai Lama) and it's convenient from a
>> programming point of view to have three nodes in the graph (one for
>> each category and one for his holiness himself) and some triples
>> linking them together ... but if you start thinking too hard about
>> what your doing from a logical or philosophical point of view, well I
>> start feeling like I've crossed the streams and pointed them at my own
>> head - especially when I try to describe it to myself in terms of the
>> RDF semantics and the interpretation function I. (Question: Does
>> IEXT(I(foaf:focus)) describe the function I itself? Answer: 42. :-)
>> That said, I can't resist having a go at thinking up a better name, so
>> here's a few thoughts, some clearly won't work in FOAF because of
>> other connotations, but might suggest other ideas...
>> avatar
>> cast
>> quintessence
>> tti (the thing itself)
>> skosification / skosifies (my favourite, simply because it's novel)
>> incarnation
>> proxy
>> surrogate
>> delegate
>> deputy
>> portrait
>> simile
>> analogue
>> echo
>> reflection / reflects
>> frankie / benjy :-)
>> Cheers
>> Alistair
>> On Wed, Aug 11, 2010 at 12:36 PM, Helmut Nagy<h.nagy@semantic-web.at>
>> wrote:
>>> Hi,
>>> we are following the discussion with great interest and are having a
>>> discussion here to so i will add some thoughts:
>>> Found this in the foaf discussion: "focus is more explicit: this is
>>> what this category or concept is "about""
>>> One concern we have is that foaf:focus may be misused in another way
>>> e.g to link a person to topics they have a "focus on"
>>> skos:"Helmut Nagy" foaf:focus http://rdf.freebase.com/ns/en.foaf
>>> So we also think that focus may be the wrong term.
>>> Cheers
>>> Helmut
>>> | Helmut Nagy
>>> | Semantic Web Company GmbH
>>> | Lerchenfelder Guertel 43/5
>>> | A - 1160 Wien, Austria
>>> | http://www.semantic-web.at/ | http://www.i-semantics.at/ |
>>> http://blog.semantic-web.at/
>>> | h.nagy@semantic-web.at
>>> On 10.08.2010, at 19:51, Dan Brickley wrote:
>>>> On Tue, Aug 10, 2010 at 7:00 PM, Bernard Vatant
>>>> <bernard.vatant@mondeca.com> wrote:
>>>>> Welcome Antoine to the brainstorming
>>>>> Since the box is open, it's open :)
>>>> :)
>>>>> I like standsFor, but my latin culture would prefer a latin term,
>>>>> so why not
>>>>> "represents" or even simply "presents" [1]
>>>>> Well, I know, I will have the same remarks as for "referent" or
>>>>> "refersTo"
>>>> Thanks for saving me some typing ;)
>>>>> But I'm waiting for real good arguments against it. A concept is
>>>>> really a
>>>>> way for a thing to be made præsens, in the various meanings of the
>>>>> word such
>>>>> as "really there" and "efficient".
>>>> RDF and OWL are themselves a representational system, as is [at
>>>> another level] the Web itself. To use such a generic term, risks
>>>> constant confusion.
>>>> Another option discussed btw was 'about'; however both RDF/XML and
>>>> RDFa syntax use that already
>>>>> See http://www.perseus.tufts.edu/hopper/morph?l=praesens&la=la#lexicon
>>>>> ... or for not-so-young frenchies remembering their humanities
>>>>> years, the
>>>>> good old Gaffiot I just discovered on-line.
>>>>> http://www.lexilogos.com/latin/gaffiot.php?p=1225
>>>> I'll have a read!
>>>> Dan
>>>>> Bernard
>>>>> [1] Since no presentation is really new, any presentation is a
>>>>> representation (and vice-versa) See
>>>>> http://blog.hubjects.com/2009/11/representation-as-translation.html
>>>>> 2010/8/10 Antoine Isaac<aisaac@few.vu.nl>
>>>>>> Hi Dan,
>>>>>> I think I buy all the naming arguments below.
>>>>>> But since the Pandora box is re-opened, even though with strong
>>>>>> warnings,
>>>>>> I'll have one try :-)
>>>>>> How about standsFor? You're using it yourself in the announcement, in
>>>>>> fact...
>>>>>> Otherwise:
>>>>>>> (aside: a possibility here might be to declare foaf:focus a sub
>>>>>>> property of inverse of dcterms:isReferencedBy)
>>>>>> I'm not sure we should go that way: DC's property seems very
>>>>>> bibliography-style citation-oriented...
>>>>>> Cheers,
>>>>>> Antoine
>>>>>>> +cc: Leigh
>>>>>>> On Mon, Aug 9, 2010 at 3:26 PM, Simon Spero<ses@unc.edu> wrote:
>>>>>>>> Dan-
>>>>>>>> can i suggest using a different word than focus, as this is term
>>>>>>>> of art
>>>>>>>> in
>>>>>>>> controlled vocabularies. It is used when referring to
>>>>>>>> modified/specialized
>>>>>>>> "terms".
>>>>>>> Thanks for the feedback. It seems that words are like Internet
>>>>>>> domain
>>>>>>> names; all the good ones are taken!
>>>>>>> To understand the extent of the "it's already in use" problem,
>>>>>>> could I
>>>>>>> ask you to post a few sentences using 'focus' from the literature?
>>>>>>> Even one would help.
>>>>>>> Naming RDF terms is something of a nightmare, because RDF is
>>>>>>> designed
>>>>>>> to allow information to flow beyond its original comfort-zone;
>>>>>>> whatever we choose here will show up in all kinds of unexpected
>>>>>>> contexts, including the Web pages of various publishers.
>>>>>>> I originally liked the 'skos:it' (and skos:as inverse) since 'it'
>>>>>>> had
>>>>>>> the charm of being at least easy to spell and quick to type. However
>>>>>>> after bouncing 'it' around in discussions 'it' transpired that 'it'
>>>>>>> was a bit too clever for 'its' own good, as a name. The 'focus' name
>>>>>>> came from discussions with Leigh Dodds, who I Cc: here. Some of our
>>>>>>> notes are in http://wiki.foaf-project.org/w/term_focus (btw each
>>>>>>> FOAF
>>>>>>> term now has a Wiki page for annotations).
>>>>>>>> Possible labels that might work could be isReferredToBy ; SKOS
>>>>>>>> concepts
>>>>>>>> are
>>>>>>>> intentional-with-a-t, so reference is a natural label;
>>>>>>>> isFoafProxyForIntentionReferencedBySKOSConcept is awful
>>>>>>>> ComputerDeutch.
>>>>>>> So I see the logic behind 'isReferredToBy', however I'm cautious
>>>>>>> for a
>>>>>>> few reasons. Firstly the inverse direction adds a level of
>>>>>>> confusion,
>>>>>>> so we'd want to have 'references', eg. "skos_3 :references
>>>>>>> thing_23".
>>>>>>> And since we're operating in the context of RDF, not to mention
>>>>>>> hypertext, there are plenty of other contexts in which 'references'
>>>>>>> gets used - mainly with documents. Which puts us in the awkward
>>>>>>> situation of deciding whether to re-use an existing more general
>>>>>>> purpose term that talks about reference; eg.
>>>>>>> http://dublincore.org/documents/dcmi-terms/ has
>>>>>>> http://dublincore.org/documents/dcmi-terms/#terms-isReferencedBy
>>>>>>> already --- "A related resource that references, cites, or otherwise
>>>>>>> points to the described resource." ... or if we proceed with a term
>>>>>>> that is explicitly for use with skos:Concept, we should expect to
>>>>>>> see
>>>>>>> it accidentally misused by anyone who is fumbling around looking
>>>>>>> for a
>>>>>>> nice term to use when one thing references, mentions, or identifies
>>>>>>> another thing.
>>>>>>> (aside: a possibility here might be to declare foaf:focus a sub
>>>>>>> property of inverse of dcterms:isReferencedBy)
>>>>>>>> Foaf person "Paul The Octopus" isReferredTo by SKOS Concept
>>>>>>>> "#PTO1".
>>>>>>>> Where "#PTO1" isSubjectOf "#document" "Decideabity and
>>>>>>>> tractablity of
>>>>>>>> logical inference with binary serial octacles".
>>>>>>>> (The halting problem has time complexity PTO(1) but other tasks may
>>>>>>>> require
>>>>>>>> an infinite series of questions.)
>>>>>>> Saying that the concept *references* the real world entity seems a
>>>>>>> tiny bit strong anyway; I guess I'd say 'reference' with regard
>>>>>>> to the
>>>>>>> concept's documentation, or with regard to a use of the concept in
>>>>>>> some document. But at some level this is all metaphor anyhow;
>>>>>>> nothing
>>>>>>> is really 'focussing' either. I had hoped 'focus' was a word that
>>>>>>> came
>>>>>>> with relatively little baggage in this community and amongst Web
>>>>>>> technologists, since 'topic' and 'subject' are already heavily
>>>>>>> over-used.
>>>>>>> I think 'references' will prove too general/broad to use directly
>>>>>>> (people will immediately start applying it with document 'mentions'
>>>>>>> and hyperlinks), but I appreciate the feedback and suggestion. Same
>>>>>>> with Bernard's 'referent', even though yes the basic idea is that
>>>>>>> the
>>>>>>> concepts are proxying / standing in for / indirectly identifying /
>>>>>>> referring to some real world entities.
>>>>>>> cheers,
>>>>>>> Dan
>>>>>>> ps. Another terminology possible ingredient; in FOAF we have a
>>>>>>> property foaf:primaryTopic which points from a document to the thing
>>>>>>> the document is primarily about. It has an inverse, isPrimaryTopicOf
>>>>>>> too.
>>>>> --
>>>>> Bernard Vatant
>>>>> Senior Consultant
>>>>> Vocabulary& Data Engineering
>>>>> Tel: +33 (0) 971 488 459
>>>>> Mail: bernard.vatant@mondeca.com
>>>>> ----------------------------------------------------
>>>>> Mondeca
>>>>> 3, cité Nollez 75018 Paris France
>>>>> Web: http://www.mondeca.com
>>>>> Blog: http://mondeca.wordpress.com
>>>>> ----------------------------------------------------
Received on Thursday, 12 August 2010 09:43:29 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 12 August 2010 09:43:30 GMT