W3C home > Mailing lists > Public > public-vocabs@w3.org > October 2013

RE: SKOS for schema.org proposal for discussion

From: Evain, Jean-Pierre <evain@ebu.ch>
Date: Wed, 9 Oct 2013 08:48:04 +0000
To: Guha <guha@google.com>, Thad Guidry <thadguidry@gmail.com>
CC: Aaron Bradley <aaranged@gmail.com>, Bernard Vatant <bernard.vatant@mondeca.com>, PublicVocabs <public-vocabs@w3.org>
Message-ID: <A71D9E70FCEE5E40A5EAA598864E605705E13088@maildrs.gva.ebu.ch>
I am fine with EnumConcept (or terminologyConcept)

I t would be nice to start discussing the implementation.

Somebody mentionned in the thread that e.g. enumConcept shoudl be declared as a subclass of SKOS concept. Why not as we woudn't need bothering about defining any additional properties that would be automatically inherited... provided schema.org imports skos. Any comment on this or suggestions?
________________________________
From: Guha [guha@google.com]
Sent: 09 October 2013 01:28
To: Thad Guidry
Cc: Aaron Bradley; Bernard Vatant; PublicVocabs
Subject: Re: SKOS for schema.org proposal for discussion

Come on folks ... let's continue the discussion ... we are well on our way to the first schema.org<http://schema.org> centi-thread.

Just joking. After all this, can everyone live with EnumConcept?

guha


On Tue, Oct 8, 2013 at 12:16 PM, Thad Guidry <thadguidry@gmail.com<mailto:thadguidry@gmail.com>> wrote:
The Google Knowledge Graph is well aware of LoC's IDs and the Concept of "Sustainable Agriculture".  You just do not see the results of the efforts on any Google Search Results page YET.  But it is coming...patience.

Bernard, you can get a sense of what Google's Knowledge Graph "knows" and will eventually show about that particular concept by looking at the JSON output from the graph for that LoC ID of sh87004216 here:  https://www.googleapis.com/freebase/v1/search?query=sh87004216&filter(all%20type:/type/object/key)&indent=true&output=((all%20all))

the (all all) outputs everything the Knowledge Graph knows outputting 2 ply for the results.. if you only want 1 ply... then just change the output=(all)

If you wanted to see only the Broader's and their Peer_of relationships... then like this output=((narrower_than peer_of))  like so:
https://www.googleapis.com/freebase/v1/search?query=sh87004216&filter(all%20type:/type/object/key)&indent=true&output=((narrower_than%20peer_of))

(full docs for the API are here if you want to play more: https://developers.google.com/freebase/ )

What your seeing is all gooble-gook, in JSON formating when you click on those links, but ....

Google is in the midst of "soonish" providing faceting and sub-searching tools to help visualize results with Broader / Narrower / Related / etc... based on SKOS Concepts.  I happen to be one of the folks involved in populating the Graph itself and linking SKOS Concepts.



On Tue, Oct 8, 2013 at 12:39 PM, Aaron Bradley <aaranged@gmail.com<mailto:aaranged@gmail.com>> wrote:
Bernard, to provide a general answer to both of your questions parenthetically prefaced "from a SEO point of view," I've yet to see any evidence that any search engine has ever used any external URI provided in schema.org<http://schema.org> markup in order to produce a rich snippet, change the ranking order of web pages or return a resource in the SERPs based solely on such a reference.

This doesn't mean that the search engines aren't ingesting and using these data, only that there's no observed evidence that they're doing so.

(I don't know whether Martin Hepp or anyone else has observed a demonstrable impact in the SERPs as a result of referencing productontology.org<http://productontology.org> URIs via additionalType - I'd certainly be interested in hearing of such cases if they exist!)

But this lack of evidence is not unimportant from an SEO practitioner's point of view, because as a result there's no incentive to employ such mechanisms:  the demonstrated "added value" is zero.



On Tue, Oct 8, 2013 at 10:06 AM, Bernard Vatant <bernard.vatant@mondeca.com<mailto:bernard.vatant@mondeca.com>> wrote:
Let me make my point differently.

Maybe this is obvious for all users of schema.org<http://schema.org>, please point me to the relevant resources if it's the case.

I want to say that my content (page/section) is about "Sustainable agriculture".

I have a skos:Concept for this, defined in a good reference vocabulary http://id.loc.gov/authorities/subjects/sh87004216

I can already use in the current state of affairs, the schema.org/about<http://schema.org/about> property to mark my page with this URI, right?

Q1. What is the current added value (from a SEO point of view) to mark with this URI vs marking with the string "Sustainable agriculture"? And particularly what is the added value of having this URI being defined as a skos:Concept in a most authoritative Concept Scheme (LCSH), instead of any other URI such as
http://dbpedia.org/page/Sustainable_agriculture
http://www.actu-environnement.com/ae/dictionnaire_environnement/definition/agriculture_durable.php4

Q2. What would be the added value (always from a SEO point of view) to add a schema.org<http://schema.org> type (whatever its name) to this URI indicating in the markup that this URI is indeed a skos:Concept belonging to a skos:ConceptScheme, namely LCSH (which you can discover by dereferencing the URI anyway, but do search engines follow their nose in the markup)?

(Thinking about it I have the same question for the use of any reference URI, be it a skos:Concept or not. What do you gain if any by using http://id.insee.fr/geo/departement/05 instead of the string "Hautes-Alpes" in a schema.org/Place<http://schema.org/Place> description?)

Bernard













2013/10/8 Vicki Tardif Holland <vtardif@google.com<mailto:vtardif@google.com>>

On Tue, Oct 8, 2013 at 10:41 AM, Karen Coyle <kcoyle@kcoyle.net<mailto:kcoyle@kcoyle.net>> wrote:
I actually read the discussion differently. It's not so much that people want to express topics in the KOS sense, but that they want to refer to controlled lists within their data, and SKOS covers that. SKOS gives you a way to define a finite list with a few useful relationships. I think it's the mechanism of SKOS that people are looking for, more than the KOS value.

I had the same interpretation.

I know that controlled vocabularies are sometimes seen as a nuisance outside of the library realm, but they are useful in the cases where programmers want an enumeration. SKOS is even better than a flat enumeration, because the vocabulary can have a hierarchy, allowing for inheritance.

As an example, we have been working through a proposal to support civic services in schema.org<http://schema.org>. One of the properties of a service is "serviceType". It would be nice to be able to encourage people to use something like openelegibility.org<http://openelegibility.org>'s taxonomy so that we have some hope of sorting out the services automatically.

- Vicki


Vicki Tardif Holland | Ontologist | vtardif@google.com<mailto:vtardif@google.com>




--
Bernard Vatant
Vocabularies & Data Engineering
Tel :  + 33 (0)9 71 48 84 59
Skype : bernard.vatant
Blog : the wheel and the hub<http://bvatant.blogspot.com>
Linked Open Vocabularies : lov.okfn.org <http://lov.okfn.org>
--------------------------------------------------------
Mondeca
3 cité Nollez 75018 Paris, France
www.mondeca.com<http://www.mondeca.com/>
Follow us on Twitter : @mondecanews<http://twitter.com/#%21/mondecanews>
----------------------------------------------------------




--
-Thad
Thad on Freebase.com<http://www.freebase.com/view/en/thad_guidry>
Thad on LinkedIn<http://www.linkedin.com/in/thadguidry/>

------------------------------------------------------------------------------

**************************************************
This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed.
If you have received this email in error, please notify the system manager. This footnote also confirms that this email message has been swept by the mailgateway
**************************************************
Received on Wednesday, 9 October 2013 08:49:00 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:29:32 UTC