Re: Time Ontology in OWL WD published

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: Time Ontology in OWL WD published

Bernard Vatant

Hello all

Some remarks on the Time Ontology WD I just start to explore.
I'm much interested by the rich modeling of time:Interval, and
singularly the time:ProperInterval, but am puzzled by all those
properties having it for both domain and range, without further
comments. Unless I miss something, there is no explanation about their
specific semantics and intended use, neither inside the RDF file, nor in
the presentation document. One can only guess from the URI what is the
semantics of intervalMeets vs intervalOverlaps or intervalEquals (the
latter I really don't have any guess on what it is intended to convey,
certainly equality, but which kind of equality?). And actually, I should
not even try to guess, since there should not be any semantics embedded
in those URI fragments, right? Are semantics defined in extra
documentation. If so, pointers, please!
How are people supposed to use those properties correctly? I suggest at
least some prose should be added either somewhere in the Draft, and/or
preferably as comments in the RDF file itself.
Moreover, maybe I'm wrong, but seems to me those different properties
are formally linked to each other by rules, like e.g., the following in
SPARQL (if I got intervalOverlaps semantics and SPARQL syntax correctly
...) BTW I find strange that time:inside has domain time:Interval and
range time:Instant. One would expect it the other way round ...

PREFIX time:    <http://www.w3.org/2006/time#>

CONSTRUCT   { ?x time:intervalOverlaps ?y }

WHERE       { ?y time:hasBeginning ?a.
                ?y time:hasEnd ?b.
                ?x time:inside ?a.
                ?x time:inside ?b.
            }

Cheers

Bernard

--

*Bernard Vatant
*Knowledge Engineering
----------------------------------------------------
*Mondeca**
*3, cité Nollez 75018 Paris France
Web:    www.mondeca.com <http://www.mondeca.com>
----------------------------------------------------
Tel:       +33 (0) 871 488 459
Mail:     [hidden email] <mailto:[hidden email]>
Blog:    Leçons de Choses <http://mondeca.wordpress.com/>


Reply | Threaded
Open this post in threaded view
|

Re: Time Ontology in OWL WD published

Feng Pan
Hi Bernard,

Thank you for your comments on the time ontology note.

The semantics and the rules/axioms for the properties are defined formally
in first-order logic, which can be found in [1,2]. [1] is linked from the
note. More comments can be added to the OWL code.

As for the domain and range of time:inside, it's in a similiar situation
as "begins" and "ends", as we have explained and discussed in [3]. To
conform to OWL convention, we have followed Guus' comments to change
"begins" to "hasBeginning" and "ends" to "hasEnd". However, if we would
have to do the same thing for "inside", it would be changed to
"hasInteriorPoint", but we would rather not to do so.

Thanks!

Feng Pan and Jerry Hobbs

[1] Jerry R. Hobbs and Feng Pan. 2004. An Ontology of Time for the
Semantic Web. ACM Transactions on Asian Language Processing
(TALIP): Special issue on Temporal Information Processing, Vol. 3, No. 1,
March 2004, pp. 66-85.
http://www.isi.edu/~pan/time/pub/hobbs-pan-TALIP04.pdf

[2] Jerry R. Hobbs. 2004. An OWL Ontology of Time
http://www.isi.edu/~pan/time/owl-time-july04.txt

[3] http://lists.w3.org/Archives/Public/public-swbp-wg/2006May/0064.html

On Fri, 10 Nov 2006, Bernard Vatant wrote:

>
> Hello all
>
> Some remarks on the Time Ontology WD I just start to explore.
> I'm much interested by the rich modeling of time:Interval, and singularly the
> time:ProperInterval, but am puzzled by all those properties having it for
> both domain and range, without further comments. Unless I miss something,
> there is no explanation about their specific semantics and intended use,
> neither inside the RDF file, nor in the presentation document. One can only
> guess from the URI what is the semantics of intervalMeets vs intervalOverlaps
> or intervalEquals (the latter I really don't have any guess on what it is
> intended to convey, certainly equality, but which kind of equality?). And
> actually, I should not even try to guess, since there should not be any
> semantics embedded in those URI fragments, right? Are semantics defined in
> extra documentation. If so, pointers, please!
> How are people supposed to use those properties correctly? I suggest at least
> some prose should be added either somewhere in the Draft, and/or preferably
> as comments in the RDF file itself.
> Moreover, maybe I'm wrong, but seems to me those different properties are
> formally linked to each other by rules, like e.g., the following in SPARQL
> (if I got intervalOverlaps semantics and SPARQL syntax correctly ...) BTW I
> find strange that time:inside has domain time:Interval and range
> time:Instant. One would expect it the other way round ...
>
> PREFIX time:    <http://www.w3.org/2006/time#>
>
> CONSTRUCT   { ?x time:intervalOverlaps ?y }
>
> WHERE       { ?y time:hasBeginning ?a.
> ?y time:hasEnd ?b.
> ?x time:inside ?a.
> ?x time:inside ?b.
>    }
> Cheers
>
> Bernard
>
> --
>
> *Bernard Vatant
> *Knowledge Engineering
> ----------------------------------------------------
> *Mondeca**
> *3, cité Nollez 75018 Paris France
> Web:    www.mondeca.com <http://www.mondeca.com>
> ----------------------------------------------------
> Tel:       +33 (0) 871 488 459
> Mail:     [hidden email] <mailto:[hidden email]>
> Blog:    Leçons de Choses <http://mondeca.wordpress.com/>
>
>
>
--
Feng Pan, Ph.D. Candidate
USC Information Sciences Institute (ISI)
Email: [hidden email]
Web: http://www.isi.edu/~pan/
Reply | Threaded
Open this post in threaded view
|

Re: Time Ontology in OWL WD published

Bernard Vatant

Hi folks

Feng Pan a écrit :
> Hi Bernard,
>
> Thank you for your comments on the time ontology note.
>
> The semantics and the rules/axioms for the properties are defined
> formally in first-order logic, which can be found in [1,2]. [1] is
> linked from the note.
Thanks a lot for that. Actually [1] is quoted in the References section
with bibliographical elements only, but without URI, hence my question.
I think it would be good to add this link in the note.
> More comments can be added to the OWL code.
That would be good and beyond comments, having the formal axioms defined
by [1] in the OWL code would be good. Although I'm not sure which axioms
e.g., linking instants and intervals, expressed as first-order logic
predicates in this document, can be expressed in OWL. Did you explore
this question? (I'm sure you have sorry if it has already been
discussed). If not in OWL, what do you think of using SPARQL CONSTRUCT
to express those axioms? Else which language?
> As for the domain and range of time:inside, it's in a similiar
> situation as "begins" and "ends", as we have explained and discussed
> in [3]. To conform to OWL convention, we have followed Guus' comments
> to change "begins" to "hasBeginning" and "ends" to "hasEnd". However,
> if we would have to do the same thing for "inside", it would be
> changed to "hasInteriorPoint", but we would rather not to do so.
OK. That was a minor point, and moreover by a non-native english speaker
:-)

>
> Thanks!
>
> Feng Pan and Jerry Hobbs
>
> [1] Jerry R. Hobbs and Feng Pan. 2004. An Ontology of Time for the
> Semantic Web. ACM Transactions on Asian Language Processing
> (TALIP): Special issue on Temporal Information Processing, Vol. 3, No.
> 1, March 2004, pp. 66-85.
> http://www.isi.edu/~pan/time/pub/hobbs-pan-TALIP04.pdf
>
> [2] Jerry R. Hobbs. 2004. An OWL Ontology of Time
> http://www.isi.edu/~pan/time/owl-time-july04.txt
>
> [3] http://lists.w3.org/Archives/Public/public-swbp-wg/2006May/0064.html
>
> On Fri, 10 Nov 2006, Bernard Vatant wrote:
>
>>
>> Hello all
>>
>> Some remarks on the Time Ontology WD I just start to explore.
>> I'm much interested by the rich modeling of time:Interval, and
>> singularly the time:ProperInterval, but am puzzled by all those
>> properties having it for both domain and range, without further
>> comments. Unless I miss something, there is no explanation about
>> their specific semantics and intended use, neither inside the RDF
>> file, nor in the presentation document. One can only guess from the
>> URI what is the semantics of intervalMeets vs intervalOverlaps or
>> intervalEquals (the latter I really don't have any guess on what it
>> is intended to convey, certainly equality, but which kind of
>> equality?). And actually, I should not even try to guess, since there
>> should not be any semantics embedded in those URI fragments, right?
>> Are semantics defined in extra documentation. If so, pointers, please!
>> How are people supposed to use those properties correctly? I suggest
>> at least some prose should be added either somewhere in the Draft,
>> and/or preferably as comments in the RDF file itself.
>> Moreover, maybe I'm wrong, but seems to me those different properties
>> are formally linked to each other by rules, like e.g., the following
>> in SPARQL (if I got intervalOverlaps semantics and SPARQL syntax
>> correctly ...) BTW I find strange that time:inside has domain
>> time:Interval and range time:Instant. One would expect it the other
>> way round ...
>>
>> PREFIX time:    <http://www.w3.org/2006/time#>
>>
>> CONSTRUCT   {     ?x time:intervalOverlaps ?y }
>>
>> WHERE       {     ?y time:hasBeginning ?a.
>>         ?y time:hasEnd ?b.
>>         ?x time:inside ?a.
>>         ?x time:inside ?b.
>>         } Cheers
>>
>> Bernard
>>
>> --
>>
>> *Bernard Vatant
>> *Knowledge Engineering
>> ----------------------------------------------------
>> *Mondeca**
>> *3, cité Nollez 75018 Paris France
>> Web:    www.mondeca.com <http://www.mondeca.com>
>> ----------------------------------------------------
>> Tel:       +33 (0) 871 488 459
>> Mail:     [hidden email] <mailto:[hidden email]>
>> Blog:    Leçons de Choses <http://mondeca.wordpress.com/>
>>
>>
>>
>
> --
> Feng Pan, Ph.D. Candidate
> USC Information Sciences Institute (ISI)
> Email: [hidden email]
> Web: http://www.isi.edu/~pan/
> ------------------------------------------------------------------------
>
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.1.409 / Virus Database: 268.14.5/533 - Release Date: 13/11/2006
>  

--

*Bernard Vatant
*Knowledge Engineering
----------------------------------------------------
*Mondeca**
*3, cité Nollez 75018 Paris France
Web:    www.mondeca.com <http://www.mondeca.com>
----------------------------------------------------
Tel:       +33 (0) 871 488 459
Mail:     [hidden email] <mailto:[hidden email]>
Blog:    Leçons de Choses <http://mondeca.wordpress.com/>