DL query with "NOT" doesnot work

classic Classic list List threaded Threaded
19 messages Options
Reply | Threaded
Open this post in threaded view
|

DL query with "NOT" doesnot work

Tran Binh
Hi all,
I create a very simple owl with 3 classes:
- SatelliteImage
- UrbanArea
- RuralArea
and property: hasCoverArea

SatelliteImage has two individuals: A and B
UrbanArea has two individuals: C and D

A hasCoverArea C
B hasCoverArea D
 
I run reasoner HermiT 1.3.3 on protege4.1. I query

SatelliteImage and not (hasCoverArea some RuralArea)


and protege return no individuals value while I expect it return A,B.

Any ideas for this problem?


Thanks,
Binh
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Timothy Redmond

Based on what you have said the reasoner is correct.   The simplest
reason is that it doesn't look like you have said that C and D are not
members of RuralArea.  In addition, you have not said that UrbanArea and
RuralArea are disjoint which would have allowed the reasoner to infer
that C and D are not members of RuralArea.

Even if you had made UrbanArea and RuralArea disjoint then the reasoner
would still not make the conclusion.  You have also not stated that
there are no other individuals E such that "A hasCoverArea E".  In
particular hasCoverArea has not been declared functional.

This is a standard issue and it is known as the open world assumption.

-Timothy


On 04/11/2011 09:25 AM, Tran Binh wrote:

> Hi all,
> I create a very simple owl with 3 classes:
> - SatelliteImage
> - UrbanArea
> - RuralArea
> and property: hasCoverArea
>
> SatelliteImage has two individuals: A and B
> UrbanArea has two individuals: C and D
>
> A hasCoverArea C
> B hasCoverArea D
>
> I run reasoner HermiT 1.3.3 on protege4.1. I query
>
> SatelliteImage and not (hasCoverArea some RuralArea)
>
> and protege return no individuals value while I expect it return A,B.
>
> Any ideas for this problem?
>
>
> Thanks,
> Binh
>
>
> --
> View this message in context: http://protege-ontology-editor-knowledge-acquisition-system.136.n4.nabble.com/DL-query-with-NOT-doesnot-work-tp3442476p3442476.html
> Sent from the Protege OWL mailing list archive at Nabble.com.
> _______________________________________________
> protege-owl mailing list
> [hidden email]
> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>
> Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Tran Binh
Thank you Timothy,
In that case, how can I assign that SatelliteImage just cover such area that have declare in owl file?
Cheers,
Binh


On Mon, Apr 11, 2011 at 6:32 PM, Timothy Redmond <[hidden email]> wrote:

Based on what you have said the reasoner is correct.   The simplest reason is that it doesn't look like you have said that C and D are not members of RuralArea.  In addition, you have not said that UrbanArea and RuralArea are disjoint which would have allowed the reasoner to infer that C and D are not members of RuralArea.

Even if you had made UrbanArea and RuralArea disjoint then the reasoner would still not make the conclusion.  You have also not stated that there are no other individuals E such that "A hasCoverArea E".  In particular hasCoverArea has not been declared functional.

This is a standard issue and it is known as the open world assumption.

-Timothy



On 04/11/2011 09:25 AM, Tran Binh wrote:
Hi all,
I create a very simple owl with 3 classes:
- SatelliteImage
- UrbanArea
- RuralArea
and property: hasCoverArea

SatelliteImage has two individuals: A and B
UrbanArea has two individuals: C and D

A hasCoverArea C
B hasCoverArea D

I run reasoner HermiT 1.3.3 on protege4.1. I query

SatelliteImage and not (hasCoverArea some RuralArea)

and protege return no individuals value while I expect it return A,B.

Any ideas for this problem?


Thanks,
Binh


--
View this message in context: http://protege-ontology-editor-knowledge-acquisition-system.136.n4.nabble.com/DL-query-with-NOT-doesnot-work-tp3442476p3442476.html
Sent from the Protege OWL mailing list archive at Nabble.com.
_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03



--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Timothy Redmond

I made the two areas disjoint and made the hasCoverArea property functional (which may not be what you want).  Then there is an argument that leads to the conclusion that you want.

On 04/11/2011 09:38 AM, Tran Thai Binh wrote:
Thank you Timothy,
In that case, how can I assign that SatelliteImage just cover such area that have declare in owl file?
Cheers,
Binh


On Mon, Apr 11, 2011 at 6:32 PM, Timothy Redmond <[hidden email]> wrote:

Based on what you have said the reasoner is correct.   The simplest reason is that it doesn't look like you have said that C and D are not members of RuralArea.  In addition, you have not said that UrbanArea and RuralArea are disjoint which would have allowed the reasoner to infer that C and D are not members of RuralArea.

Even if you had made UrbanArea and RuralArea disjoint then the reasoner would still not make the conclusion.  You have also not stated that there are no other individuals E such that "A hasCoverArea E".  In particular hasCoverArea has not been declared functional.

This is a standard issue and it is known as the open world assumption.

-Timothy



On 04/11/2011 09:25 AM, Tran Binh wrote:
Hi all,
I create a very simple owl with 3 classes:
- SatelliteImage
- UrbanArea
- RuralArea
and property: hasCoverArea

SatelliteImage has two individuals: A and B
UrbanArea has two individuals: C and D

A hasCoverArea C
B hasCoverArea D

I run reasoner HermiT 1.3.3 on protege4.1. I query

SatelliteImage and not (hasCoverArea some RuralArea)

and protege return no individuals value while I expect it return A,B.

Any ideas for this problem?


Thanks,
Binh


--
View this message in context: http://protege-ontology-editor-knowledge-acquisition-system.136.n4.nabble.com/DL-query-with-NOT-doesnot-work-tp3442476p3442476.html
Sent from the Protege OWL mailing list archive at Nabble.com.
_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03



--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -
_______________________________________________ protege-owl mailing list [hidden email] https://mailman.stanford.edu/mailman/listinfo/protege-owl Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03


_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03

Region.owl (4K) Download Attachment
Region.png (87K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Tran Binh
Actually, It is not correct if we make hasCoverArea functional, because if it is functional so UrbanArea A will be UrbanArea B. I dont want it.

And I wonder why it works with functional property.

Binh


On Mon, Apr 11, 2011 at 6:49 PM, Timothy Redmond <[hidden email]> wrote:

I made the two areas disjoint and made the hasCoverArea property functional (which may not be what you want).  Then there is an argument that leads to the conclusion that you want.


On 04/11/2011 09:38 AM, Tran Thai Binh wrote:
Thank you Timothy,
In that case, how can I assign that SatelliteImage just cover such area that have declare in owl file?
Cheers,
Binh


On Mon, Apr 11, 2011 at 6:32 PM, Timothy Redmond <[hidden email]> wrote:

Based on what you have said the reasoner is correct.   The simplest reason is that it doesn't look like you have said that C and D are not members of RuralArea.  In addition, you have not said that UrbanArea and RuralArea are disjoint which would have allowed the reasoner to infer that C and D are not members of RuralArea.

Even if you had made UrbanArea and RuralArea disjoint then the reasoner would still not make the conclusion.  You have also not stated that there are no other individuals E such that "A hasCoverArea E".  In particular hasCoverArea has not been declared functional.

This is a standard issue and it is known as the open world assumption.

-Timothy



On 04/11/2011 09:25 AM, Tran Binh wrote:
Hi all,
I create a very simple owl with 3 classes:
- SatelliteImage
- UrbanArea
- RuralArea
and property: hasCoverArea

SatelliteImage has two individuals: A and B
UrbanArea has two individuals: C and D

A hasCoverArea C
B hasCoverArea D

I run reasoner HermiT 1.3.3 on protege4.1. I query

SatelliteImage and not (hasCoverArea some RuralArea)

and protege return no individuals value while I expect it return A,B.

Any ideas for this problem?


Thanks,
Binh


--
View this message in context: http://protege-ontology-editor-knowledge-acquisition-system.136.n4.nabble.com/DL-query-with-NOT-doesnot-work-tp3442476p3442476.html
Sent from the Protege OWL mailing list archive at Nabble.com.
_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03



--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -
_______________________________________________ protege-owl mailing list [hidden email] https://mailman.stanford.edu/mailman/listinfo/protege-owl Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03


_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03




--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Olivier Dameron
In reply to this post by Timothy Redmond
On 04/11/2011 06:32 PM, Timothy Redmond wrote:
> Even if you had made UrbanArea and RuralArea disjoint then the reasoner
> would still not make the conclusion.  You have also not stated that
> there are no other individuals E such that "A hasCoverArea E".  In
> particular hasCoverArea has not been declared functional.
>
> This is a standard issue and it is known as the open world assumption.

Sorry for being picky, but the open-world assumption is an OWL feature
and not an issue (although in Tran Binh's case this may well be :-)

olivier (yes, I'll take a breath)
_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Olivier Dameron
In reply to this post by Tran Binh
On 04/11/2011 06:58 PM, Tran Thai Binh wrote:
> And I wonder why it works with functional property.

b hasCoverArea d

if hasCoverArea is functional, b can have only 1 cover area (wich is an
instance of UrbanArea). Therefore, b cannot have other cover areas that
would be instances of RuralArea.

The "functional" axiom prevents your model from having other unspecified
cover areas associated with b

olivier
_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Tran Binh


On Mon, Apr 11, 2011 at 7:03 PM, Olivier Dameron <[hidden email]> wrote:

The "functional" axiom prevents your model from having other unspecified cover areas associated with b

But it also means that 1 satellite image cover exactly 1 urban area? am i right?
 



--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Tran Binh
Is there any way to set something like "Affirmative declaration" (I dont know how to call it) that means if I say satellite image A cover urbanarea X,Y,Z so it cover exactly 3 areas no more.

I know we use open world assumption but sometime we need some affirmative declaration.

Cheers,
Binh

On Mon, Apr 11, 2011 at 7:07 PM, Tran Thai Binh <[hidden email]> wrote:


On Mon, Apr 11, 2011 at 7:03 PM, Olivier Dameron <[hidden email]> wrote:

The "functional" axiom prevents your model from having other unspecified cover areas associated with b

But it also means that 1 satellite image cover exactly 1 urban area? am i right?
 



--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -



--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Thomas Russ
In reply to this post by Tran Binh

On Apr 11, 2011, at 10:07 AM, Tran Thai Binh wrote:

>
>
> On Mon, Apr 11, 2011 at 7:03 PM, Olivier Dameron <[hidden email]> wrote:
>
> The "functional" axiom prevents your model from having other unspecified cover areas associated with b
>
> But it also means that 1 satellite image cover exactly 1 urban area? am i right?

Correct.  It is not what you want.

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Thomas Russ
In reply to this post by Tran Binh

On Apr 11, 2011, at 10:12 AM, Tran Thai Binh wrote:

> Is there any way to set something like "Affirmative declaration" (I dont know how to call it) that means if I say satellite image A cover urbanarea X,Y,Z so it cover exactly 3 areas no more.
>
> I know we use open world assumption but sometime we need some affirmative declaration.

Yes, you can do this.

What you need to do is add a closure assertion to your knowledge base.

So if you have the following:

  A cover X
  A cover Y
  A cover Z

open world semantics allows for there to be other areas that are covered by the image.  To exclude them, you have to say that X,Y and Z are the only areas covered by this image.  You can do that with an allValuesFrom restriction ("only") that lists exactly the individuals that are allowed.  This would involve making A an instance of a new, possibly anonymous class definition:

  A rdf:type (cover only {X Y Z})




>
> Cheers,
> Binh
>
> On Mon, Apr 11, 2011 at 7:07 PM, Tran Thai Binh <[hidden email]> wrote:
>
>
> On Mon, Apr 11, 2011 at 7:03 PM, Olivier Dameron <[hidden email]> wrote:
>
> The "functional" axiom prevents your model from having other unspecified cover areas associated with b
>
> But it also means that 1 satellite image cover exactly 1 urban area? am i right?
>  
>
>
>
> --
> Trần Thái Bình
> GIS and Remote Sensing
> [hidden email]
> [hidden email]
> Phone: (848) 38247360
> mobile: 0913 194 917
> - Việt Nam -
>
>
>
> --
> Trần Thái Bình
> GIS and Remote Sensing
> [hidden email]
> [hidden email]
> Phone: (848) 38247360
> mobile: 0913 194 917
> - Việt Nam -
> _______________________________________________
> protege-owl mailing list
> [hidden email]
> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>
> Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Tran Binh

open world semantics allows for there to be other areas that are covered by the image.  To exclude them, you have to say that X,Y and Z are the only areas covered by this image.  You can do that with an allValuesFrom restriction ("only") that lists exactly the individuals that are allowed.  This would involve making A an instance of a new, possibly anonymous class definition:

 A rdf:type (cover only {X Y Z})

Please show me how to set it in protege 4.1
Thanks,
Binh

--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Tran Binh
Hi all,

I think I should decribe more clearly my case. I have some SatelliteImage, some of it cover only some UrbanArea, some of it cover only some RuralArea, and some cover both UrbanArea and RuralArea. I want to query for only SatelliteImage that cover only UrbanArea.

How can I do that?

Please, thanks,
Binh

On Tue, Apr 12, 2011 at 9:17 AM, Tran Thai Binh <[hidden email]> wrote:

open world semantics allows for there to be other areas that are covered by the image.  To exclude them, you have to say that X,Y and Z are the only areas covered by this image.  You can do that with an allValuesFrom restriction ("only") that lists exactly the individuals that are allowed.  This would involve making A an instance of a new, possibly anonymous class definition:

 A rdf:type (cover only {X Y Z})

Please show me how to set it in protege 4.1
Thanks,
Binh


--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -



--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Timothy Redmond
On 04/12/2011 01:20 AM, Tran Thai Binh wrote:
Hi all,

I think I should decribe more clearly my case. I have some SatelliteImage, some of it cover only some UrbanArea, some of it cover only some RuralArea, and some cover both UrbanArea and RuralArea. I want to query for only SatelliteImage that cover only UrbanArea.

There is nothing wrong with the way you set up the query.  You just did not give enough facts in your original ontology to conclude that your satellite images did not include rural areas.  I added some detail to my previous example to give another standard way to define satellite images that only cover urban areas.  So in particular my definition of the satellite image, b, is as follows:

Individual: b
    Types: 
        hasCoverArea exactly 1 Area,
        SatelliteImage
    Facts:  
     hasCoverArea  d

Here I have said that the satellite image b only covers the one area, d.  Thus I am able to conclude that the satellite image b only covers urban areas.

-Timothy



How can I do that?

Please, thanks,
Binh

On Tue, Apr 12, 2011 at 9:17 AM, Tran Thai Binh <[hidden email]> wrote:

open world semantics allows for there to be other areas that are covered by the image.  To exclude them, you have to say that X,Y and Z are the only areas covered by this image.  You can do that with an allValuesFrom restriction ("only") that lists exactly the individuals that are allowed.  This would involve making A an instance of a new, possibly anonymous class definition:

 A rdf:type (cover only {X Y Z})

Please show me how to set it in protege 4.1
Thanks,
Binh


--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -



--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -
_______________________________________________ protege-owl mailing list [hidden email] https://mailman.stanford.edu/mailman/listinfo/protege-owl Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03


_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Timothy Redmond
In reply to this post by Tran Binh

Oops - here is the ontology.

-Timothy


On 04/12/2011 01:20 AM, Tran Thai Binh wrote:
Hi all,

I think I should decribe more clearly my case. I have some SatelliteImage, some of it cover only some UrbanArea, some of it cover only some RuralArea, and some cover both UrbanArea and RuralArea. I want to query for only SatelliteImage that cover only UrbanArea.

How can I do that?

Please, thanks,
Binh

On Tue, Apr 12, 2011 at 9:17 AM, Tran Thai Binh <[hidden email]> wrote:

open world semantics allows for there to be other areas that are covered by the image.  To exclude them, you have to say that X,Y and Z are the only areas covered by this image.  You can do that with an allValuesFrom restriction ("only") that lists exactly the individuals that are allowed.  This would involve making A an instance of a new, possibly anonymous class definition:

 A rdf:type (cover only {X Y Z})

Please show me how to set it in protege 4.1
Thanks,
Binh


--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -



--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -
_______________________________________________ protege-owl mailing list [hidden email] https://mailman.stanford.edu/mailman/listinfo/protege-owl Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03


_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03

Region.owl (8K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Tran Binh
In reply to this post by Timothy Redmond

Individual: b
    Types: 
        hasCoverArea exactly 1 Area,
        SatelliteImage
    Facts:  
     hasCoverArea  d

Here I have said that the satellite image b only covers the one area, d.  Thus I am able to conclude that the satellite image b only covers urban areas.

Thanks Timothy, The idea is, if I understand correctly, you set the restriction "exactly" but it is not suitable in my case, because I use Mapping Master to import data into my owl file, and the individuals are so many, and they also have relationship with many region. It is impossible to count the cover area.

Binh
--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Thomas Russ

On Apr 12, 2011, at 9:01 AM, Tran Thai Binh wrote:

>
> Individual: b
>     Types:
>         hasCoverArea exactly 1 Area,
>         SatelliteImage
>     Facts:  
>      hasCoverArea  d
>
>
> Here I have said that the satellite image b only covers the one area, d.  Thus I am able to conclude that the satellite image b only covers urban areas.
>
> Thanks Timothy, The idea is, if I understand correctly, you set the restriction "exactly" but it is not suitable in my case, because I use Mapping Master to import data into my owl file, and the individuals are so many, and they also have relationship with many region. It is impossible to count the cover area.

If it is impossible to count the cover area, you will also have difficulty in enumerating the property filler values for producing a closure axiom that says only {this list of individuals} can be values of this property.  Which is what you will need to do in order to effectively have negation work with open world semantics.


_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Thomas Russ
In reply to this post by Tran Binh

On Apr 12, 2011, at 1:20 AM, Tran Thai Binh wrote:

> Hi all,
>
> I think I should decribe more clearly my case. I have some SatelliteImage, some of it cover only some UrbanArea, some of it cover only some RuralArea, and some cover both UrbanArea and RuralArea. I want to query for only SatelliteImage that cover only UrbanArea.
>
> How can I do that?
>
> Please, thanks,
> Binh
>
> On Tue, Apr 12, 2011 at 9:17 AM, Tran Thai Binh <[hidden email]> wrote:
>
> open world semantics allows for there to be other areas that are covered by the image.  To exclude them, you have to say that X,Y and Z are the only areas covered by this image.  You can do that with an allValuesFrom restriction ("only") that lists exactly the individuals that are allowed.  This would involve making A an instance of a new, possibly anonymous class definition:
>
>  A rdf:type (cover only {X Y Z})
>
> Please show me how to set it in protege 4.1
Go to the individuals panel.
Click the "+" button next to Types.
Select the "Class Expression Editor" tab.
Type in:
   hasCoverArea only {area-instance-1, area-instance-2, ...}
where the area-instances are the areas that this particular image covers.  You need to include all of them.


See attached.




_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03

image-example.owl (4K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: DL query with "NOT" doesnot work

Tran Binh
Thank you very much Thomas Russ, I got it.
My database is very dynamic and I use Mapping Master to import individuals into my OWL file.
Is there any way to set it with Mapping Master?

Cheers,
Binh

On Wed, Apr 13, 2011 at 1:09 AM, Thomas Russ <[hidden email]> wrote:

On Apr 12, 2011, at 1:20 AM, Tran Thai Binh wrote:

> Hi all,
>
> I think I should decribe more clearly my case. I have some SatelliteImage, some of it cover only some UrbanArea, some of it cover only some RuralArea, and some cover both UrbanArea and RuralArea. I want to query for only SatelliteImage that cover only UrbanArea.
>
> How can I do that?
>
> Please, thanks,
> Binh
>
> On Tue, Apr 12, 2011 at 9:17 AM, Tran Thai Binh <[hidden email]> wrote:
>
> open world semantics allows for there to be other areas that are covered by the image.  To exclude them, you have to say that X,Y and Z are the only areas covered by this image.  You can do that with an allValuesFrom restriction ("only") that lists exactly the individuals that are allowed.  This would involve making A an instance of a new, possibly anonymous class definition:
>
>  A rdf:type (cover only {X Y Z})
>
> Please show me how to set it in protege 4.1

Go to the individuals panel.
Click the "+" button next to Types.
Select the "Class Expression Editor" tab.
Type in:
  hasCoverArea only {area-instance-1, area-instance-2, ...}
where the area-instances are the areas that this particular image covers.  You need to include all of them.


See attached.




_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03




--
Trần Thái Bình
GIS and Remote Sensing
[hidden email]
[hidden email]
Phone: (848) 38247360
mobile: 0913 194 917
- Việt Nam -

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03