W3C home > Mailing lists > Public > semantic-web@w3.org > December 2008

Re: units and durations

From: Story Henry <henry.story@bblfish.net>
Date: Wed, 10 Dec 2008 16:33:38 +0100
Cc: georgi.kobilarov@gmx.de, <dbpedia-discussion@lists.sourceforge.net>, <public-lod@w3.org>, Semantic Web <semantic-web@w3.org>
Message-Id: <720A2AA9-4FA6-4026-B263-505E10ADE11E@bblfish.net>
To: Toby A Inkster <tai@g5n.co.uk>


On 10 Dec 2008, at 15:43, Toby A Inkster wrote:

>
> Georgi Kobilarov wrote:
>
>> 1. lengths:
>> metre, kilometre, centimetre, km^2, etc.
>> dbpedia:Rhine <http://dbpedia.org/ontology/length>
>> "1320"^^<http://dbpedia.org/ontology/kilometre>
>
>
> See:
> http://idi.fundacionctic.org/muo/muo-vocab.html

Very thorough work but it makes the mistakes described here

http://esw.w3.org/topic/InterpretationProperties

So for example his example

http://forge.morfeo-project.org/wiki_en/index.php/Units_of_measurement_ontology

:Spain :area [ muo:numericalValue "504782"^^xsd:double ;
                muo:measuredIn :Sq_Km ] .


will when merged with the following which represents the same area

:Spain :area [ muo:numericalValue "504782000000"^^xsd:double ;
                muo:measuredIn :Sq_m ] .
give the useless
:Spain :area [ muo:numericalValue "504782"^^xsd:double ;
                muo:numericalValue "504782000000"^^xsd:double ;
                muo:measuredIn :Sq_m;
                muo:measuredIn :Sq_Km ] .

if the :area relation is considered to be functional, which it should  
very well be considered to be.
A better solution is found here:
http://www.w3.org/2007/ont/unit


Use cwm to read it
>> 2. runtimes
>> seconds, minutes, hours, days...
>> dbpedia:Yellow_Submarine_(song) <http://dbpedia.org/ontology/runtime>
>> "2:38"^^<http://dbpedia.org/ontology/minute>
>> dbpedia:The_Departed <http://dbpedia.org/ontology/runtime>
>> "151"^^<http://dbpedia.org/ontology/minute>
>>
>> 3. durations
>> dbpedia:Thirty_Years'_War <http://dbpedia.org/ontology/date>
>> "1618-1648"^^???
>
> In terms of ISO 8601, "durations" are defined as abstract floating  
> lengths of time, such as "three and a half minutes". When you fix  
> the duration to a particular time, such as "three and a half  
> minutes, starting now", then it is classed as an "interval".
>
> For durations, there is there is xsd:duration. e.g.
>
> 	dbpedia:Yellow_Submarine_(song)
> 	  ont:runtime
> 	    "PT2M38S"^^xsd:duration .
>
> For intervals, XSD doesn't offer a datatype, but ISO 8601 at least  
> offers a machine readable standard syntax for them - intervals are  
> written, slash-separated, as a start-time/end-time pair, a start- 
> time/duration pair, or a duration/end-time pair. Personally, I tend  
> to represent this in RDF as:
>
> 	dbpedia:Thirty_Years'_War
> 	  ont:date
> 	    "1618/1648"^^<urn:iso:std:iso:8601#timeInterval> .

There are temporal ontologies that would work better here. Think of  
time slices.

http://www.w3.org/TR/owl-time/


>
>
> -- 
> Toby A Inkster
> <mailto:mail@tobyinkster.co.uk>
> <http://tobyinkster.co.uk>
>
>
>
>
Received on Wednesday, 10 December 2008 15:34:29 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 21:45:26 GMT