[foaf-dev] foaf:interest seems not very useful, targeting URLs rather than URIs

Kingsley Idehen kidehen at openlinksw.com
Wed Dec 17 00:45:55 CET 2008


On 12/16/08 6:18 PM, Damian Steer wrote:
> On 16 Dec 2008, at 22:03, Ted Thibodeau Jr wrote:
>
>    
>> Brief discussion on #foaf --
>>
>>      
>
> <snip>
>
>    
>> 16:46:17 I understand the thinking.
>> 16:46:21 I suggest it's broken thinking.
>> 16:46:38 URIs rather than URLs...  it's a simple alternative
>>      
>
> You suggest there's a difference between URIs and URLs?
>
>    
>> 16:47:25 if a page lacks a declared primary topic -- I cannot
>> reasonably
>>           use it for my foaf:interest, even if it *is* about that topic
>>      
>
> 1) Missing != broken. 2) it doesn't need to be explicit to be useful.
>
> I think the story around foaf:interest goes something like this: what
> can use to identify interests? We could build an library of
> identifiers (see PSIs in the old days), but they seem to move slowly.
> On the other hand when something is of potential interest there's a
> fighting chance at least one obsessive has written a web page
> (primarily) about that interest. So use those pages, which often have
> high google visibility, to indirectly identify those interests.
> Wikipedia is ideal, of course.
>
> So it's by no means broken, but I'd dearly love a name for the direct
> relation between person and interest topic. topic_interest was
> supposed to do that, but it isn't recommended.
>
> Damian
>    
Damian,

I've just changed all my foaf:interest property values from 
non-document-thing URIs to Web Page URIs (or URI/URLs to be precise).

I've also created new foaf:topic_interest values using 
non-document-things URIs as per:
http://xmlns.com/foaf/0.1/topic_interest, which I can live with :-)

I don't think foaf:topic_interest should go away, especially now that  
taxonomies with vocabulary cross-linkage  (e.g UMBEL) are taking shape 
within the Linked Data Web. I (and I assume others) would like to 
explicitly identify interests using URIs via within  profile oriented 
data spaces (e.g. a FOAF based profile graph).

Kingsley

>
>
>
> _______________________________________________
> foaf-dev mailing list
> foaf-dev at lists.foaf-project.org
> http://lists.foaf-project.org/mailman/listinfo/foaf-dev
>
>    


-- 


Regards,

Kingsley Idehen	      Weblog: http://www.openlinksw.com/blog/~kidehen
President&  CEO
OpenLink Software     Web: http://www.openlinksw.com






More information about the foaf-dev mailing list