W3C home > Mailing lists > Public > public-vocabs@w3.org > April 2014

Re: Socialnetworks of a person or organization

From: Martin Hepp <martin.hepp@unibw.de>
Date: Fri, 11 Apr 2014 09:41:11 +0200
Cc: W3C Web Schemas Task Force <public-vocabs@w3.org>
Message-Id: <655002B8-F6EA-4A07-8BB5-0F45DAE1AD03@unibw.de>
To: Mike Bergman <mike@mkbergman.com>
My context is building Web vocabularies that hit a sweet spot between 

1. the ease of consumption of data (e.g. granularity, disambiguity, conceptual clarity), 
2. adoption at scale, and 
3. data quality.

This goal only partly overlaps with the agenda of promoting the Linked Open Data vision and "best" practices.

> If your context is implementation, then you really have to be in the structured data camp where it does not matter a fig how your relations/attributes are stated.

If you look at the Web ecosystem as a whole, even tiny implementation details (e.g. naming or the direction of a relationship type) can matter in terms of 1-3 from above.

Martin

On 11 Apr 2014, at 09:19, Mike Bergman <mike@mkbergman.com> wrote:

> Hi Martin,
> 
> Your point is a poster child for context.
> 
> If your context is tag adoption, which I think is important to you, your comments are spot on.
> 
> If your context is context, then you want a TBox representation, as Kingsley so often usefully highlights.
> 
> If your context is implementation, then you really have to be in the structured data camp where it does not matter a fig how your relations/attributes are stated.
> 
> What kind of context are you?
> 
> Best, Mike
> 
> On 4/11/2014 2:10 AM, martin.hepp@ebusiness-unibw.org wrote:
>> Kingsley:
>> My main point is that the theroretically nice abstractions that form the basis for the Linked Data Paradigm (entity vs. representation of a description etc.) are not necessary and not appropriate for schema.org.
>> 
>> As you know from www.productontology.org, I know quite exactly what I am talking of (it has all the best practices built in), but
>> 
>> - they were very complex to implement
>> - they are practically useless (99% of www.productontology.org users just need an URI for indicating a type of product)
>> - the HTTP content negotiation between the document URIs and the entity URIs alone consumes 98 % of all server resources and costs me ca. 1000 EUR / year for no actual benefit.
>> 
>> I am planning to flatten the implementation in www.productontology.org so that there will be just one URI for the page and the type entity, essentially removing all linked data stuff in the near future.
>> 
>> Best wishes / Mit freundlichen Grüßen
>> 
>> Martin Hepp
>> 
>> -------------------------------------------------------
>> martin hepp
>> e-business & web science research group
>> universitaet der bundeswehr muenchen
>> 
>> e-mail:  martin.hepp@unibw.de
>> phone:   +49-(0)89-6004-4217
>> fax:     +49-(0)89-6004-4620
>> www:     http://www.unibw.de/ebusiness/ (group)
>>          http://www.heppnetz.de/ (personal)
>> skype:   mfhepp
>> twitter: mfhepp
>> 
>> Check out GoodRelations for E-Commerce on the Web of Linked Data!
>> =================================================================
>> * Project Main Page: http://purl.org/goodrelations/
>> 
>> 
>> 
>> 
>> On 10 Apr 2014, at 16:15, Kingsley Idehen <kidehen@openlinksw.com> wrote:
>> 
>>> 
>>> Conclusion:
>>> 
>>> The basis have been covered for a long time re., DBpedia. I don't see what applying "application/json+schema" or "text/plain" on all resources brings to the table, bar ambiguity and the resultant out-of-band processing which is antithetical to the fundamental goals of Linked Open Data.
>> 
>> 
>> 
>> 
> 
> 
Received on Friday, 11 April 2014 07:41:41 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:29:39 UTC