Re: toString and fresnel

From: David Huynh <dfhuynh_at_csail.mit.edu>
Date: Tue, 22 Mar 2005 10:21:39 -0500

Eric Miller wrote:

> Yes. In the case of not knowing the 'label', applications should
> defaulting to some sort of additional helpful display. While this
> default will be application specific in the case of Longwell /
> PiggyBank / SemBank a linked version of the above is preferable to
> 'anonymous'.

Longwell 2.0 looks for
- any property that is rdfs:subPropertyOf dc:title or rdfs:label on the
class(es) of the resource in question, and tries to retrieve any such
property from the resource
- failing that, it queries for all property of the resource and selects
whichever one that is rdfs:subPropertyof dc:title or rdfs:label
(actually, this is commented out at the moment as it's too expensive)
- failing that, it tries to retrieve rdfs:label and then dc:title from
the resource
- failing even that, if the resource is anonymous, it displays
"(anonymous item)"; otherwise, it attempts to abbreviate the URI by
taking the suffix of the URI following either "#" or "/". In fact, this
is how you are seeing the labels for most properties in LW/PB/SB 2.0 in
the absence of the appropriate ontologies... and you haven't complained
about them yet ;-)

I like David Karger's suggestion of "unnamed <rdfs:Class of resource>".
That's definitely more helpful than "(anonymous item)". I'll make the
change.

David
Received on Tue Mar 22 2005 - 15:22:33 EST