Re: status and problems on sematicweb.org

Certainly some curation would be great.

Having RDF on the data.semanticweb.org site that doesn't validate is Not a Good Idea.

I refer you to my earlier message, below.

Best
Hugh

Begin forwarded message:

> From: Hugh Glaser <hg@ecs.soton.ac.uk>
> Subject: Re: The Linked Data-a-thon starts NOW!
> Date: 2 October 2011 16:35:27 GMT+01:00
> To: Juan Sequeda <juanfederico@gmail.com>
> Cc: public-lod <public-lod@w3.org>, Semantic Web <semantic-web@w3.org>, <iswc2011-news@mailman.stanford.edu>
> 
> Hi there.
> Fun times!
> 
> If I had noticed earlier I would have got this addressed at source, but I have a feeling that quite a few people might be already starting on the work - sorry I didn't notice earlier.
> And the admin@data.semanticweb.org email address bounced, and I am not sure who actually runs it.
> I am guessing that many entrants will use the mandatory data, plus other data from semanticweb.org, and this may need a bit of a health warning.
> 
> If you are getting data from the dumps (I have done nothing to check the other sources).
> Some problems in the dumps that you probably want to sort out before you start using the data:
> 
> http://data.semanticweb.org/dumps/other/pref_labels.rdf
> is not valid RDF.
> The last three entries should close with </rdf:Description>, not </Organization>
> 
> www-2008-complete.rdf:        <rdfs:label></rdfs:label>
> fis-2010-complete.rdf:        <rdfs:label> </rdfs:label>
> aren't very nice (rather empty).
> One comes from
>    <Person rdf:about="http://data.semanticweb.org/person/">
>        <swc:holdsRole rdf:resource="http://data.semanticweb.org/conference/fis/2010/chair"/>
>        <rdfs:label> </rdfs:label>
>        <name> </name>
> which is suspect.
> And I am not sure that
> LDOW-2008-complete.rdf:   <rdfs:label><![CDATA[Linked Data Spaces & Data Portability]]></rdfs:label>
> is what you really want.
> 
> Then I got into real problems doing coref on the data:
> hg@cohen [2011-10-02T14:56:26] > grep "<name> </name>" models/*
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/eswc-2007-complete.rdf:        <name> </name>
> models/fis-2010-complete.rdf:        <name> </name>
> 
> http://data.semanticweb.org/conference/ICPW/2009/paper/main/8
> is in the icpw-2009-complete.rdf RDF, but 404s (both RDF & html)
> In fact a lot of icpw-2009-complete.rdf URIs 404.
> 
> There are other things to do with diversity of ontology (dc:title used for both terms and vanilla, I think), but they are all part of the rich tapestry of consuming RDF.
> 
> I think I'll stop for now - hope that helps.
> Cheers
On 13 Jan 2012, at 14:41, Tim Berners-Lee wrote:

> This is indeed an important issue for the community.
> 
> One community service that W3C has provided in the past has been to
> take over the service of a domain, like daml.org, which hosts 
> ontologies which may be used by all kinds of people, but whose
> development has stopped.
> 
> Also, we have come to agreements with some domain hosts which
> have sem web ontologies that if they wrap up the project or the organization
> we will take over the domain in the future.
> 
> W3C itself has a persistence policy which commits that we will do the same
> with another public interest hosting organization should we ever
> wrap up W3C.
> 
> I could ask folks whether we could do this for *.semanticweb.org domains 
> which are no longer being modified.
> 
> Tim
> 
> On 2012-01 -13, at 04:04, Michael Hausenblas wrote:
> 
>> 
>>> This is only weakly related to this wiki site discussion but should be easy to solve anyway. Short answers:
>>> 
>>> * Subdomains of semanticweb.org are managed independently and (mainly) on different servers run by different people. The responsibility is with the people who asked for the subdomain. Stefan Decker is managing the domain registration and should know who is responsible in cases where it is not clear from the web page.
>> 
>> 
>> Exactly.
>> 
>> 
>>> * Content of earlier versions of semanticweb.org has not been migrated in the last relaunch of the site a few years ago. As long as it is just static HTML content that does not conflict with new content (i.e., if there is nothing else under that URL now), it would be possible to put it back into its place. To do this, please send me an email with the archived HTML content and let me know where to put it. I do not have access to any old versions of the site, so I cannot help in finding the old content.
> 
> How did that happen?
> 
> 
>> 
>> 
>> Maybe we should also have a look at http://semanticweb.org/wiki/Special:PopularPages to figure what is relevant or not (modulo person-pages and long tail). Happy to invest some time content-wise ...
>> 
>> Cheers,
>> 	Michael
>> -[...]
> 

-- 
Hugh Glaser,  
             Web and Internet Science
             Electronics and Computer Science,
             University of Southampton,
             Southampton SO17 1BJ
Work: +44 23 8059 3670, Fax: +44 23 8059 3045
Mobile: +44 75 9533 4155 , Home: +44 23 8061 5652
http://www.ecs.soton.ac.uk/~hg/

Received on Friday, 13 January 2012 15:00:48 UTC