W3C home > Mailing lists > Public > public-lod@w3.org > March 2011

Re: data schema / vocabulary / ontology / repositories

From: Kingsley Idehen <kidehen@openlinksw.com>
Date: Tue, 15 Mar 2011 12:39:38 -0400
Message-ID: <4D7F964A.5000206@openlinksw.com>
To: Renaud Delbru <renaud.delbru@deri.org>
CC: Keith Alexander <k.j.w.alexander@gmail.com>, Martin Hepp <martin.hepp@ebusiness-unibw.org>, Christopher Gutteridge <cjg@ecs.soton.ac.uk>, Dieter Fensel <dieter.fensel@sti2.at>, Linked Data community <public-lod@w3.org>, semantic-web <semantic-web@w3.org>
On 3/15/11 12:00 PM, Renaud Delbru wrote:
> Hi Keith,
> On 14/03/11 13:18, Keith Alexander wrote:
>> On Mon, Mar 14, 2011 at 8:28 AM, Martin Hepp
>> <martin.hepp@ebusiness-unibw.org>  wrote:
>>> Hi Dieter:
>>> There are several ontology repositories available on-line, but to my 
>>> knowledge they all suffer from two serious limitations:
>>> 1. They do not rate ontologies by quality/relevance/popularity, so 
>>> you do not get any hint whether foaf:Organization or 
>>> foo:Organization will be the best way to expose your data.
>> Schemacache[1] used to order results by the number of documents
>> Sindice found it it, but this wasn't terribly effective; what we want
>> is  something more like "number of individual publishers using term X"
>> rather than "number of individual documents using term X".
> The new Sindice search frontend provides a first solution towards this 
> problematic. Sindice allows you to group search results per domain. 
> See [1,2] as examples. It is not yet perfect, nor optimal, but this is 
> a first try, and this might be useful for your scenario.
> We are currently focussing our effort in Sindice towards what we call 
> dataset search. The dataset/domain grouping is a first step towards 
> this big picture. We will add additional features in the future, like 
> a more detailled summary of the datasets, e.g., its inter-relations 
> with other datasets.
> [1] http://sindice.com/search?q=foaf%3Aperson&nq=&fq=&facet.field=domain
> [2] http://sindice.com/search?q=owl%3Asameas&nq=&fq=&facet.field=domain

No golden answer (I sure someone's already made this comment). All you 
can do is offer access to a data space that let's the user Find what 
they seek via disambiguation oriented filters.


1. http://uriburner.com/fct/facet.vsp?cmd=load&fsq_id=81 -- Pattern: 
Person that may or may not be associated with a Class or Property
2. http://uriburner.com/fct/facet.vsp?cmd=load&fsq_id=82 -- Pattern: 
Person associated with a Class (explicitly)
3. http://uriburner.com/fct/facet.vsp?cmd=load&fsq_id=83 -- Pattern: 
Person associated with a Transitive Property.

In all cases above, when happy click on Entity1 or EntityN (too see and 
access descriptions of matching entities) depending on where you're at 
in your quest. Of course, you can switch "uriburner.com" for 
"lod.openlinksw.com" and do the same thing against an even larger data 
set etc..



Kingsley Idehen	
President&  CEO
OpenLink Software
Web: http://www.openlinksw.com
Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca: kidehen
Received on Tuesday, 15 March 2011 16:41:23 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:16:12 UTC