failed exporting inferred axioms

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

failed exporting inferred axioms

Federico Pianzola
I tried to export the inferred axioms obtained with Pellet but if I check the output file they are not visible.
I followed this guide https://www.michaeldebellis.com/post/export-inferred-axioms but gave up ticking the last two boxes because it was taking way too much time.

Is anybody willing to give it a try? Here is the original ontology: https://drive.google.com/file/d/1yPdIUpIKkxxoj9Y0d6ejmGqhLGq7hUTn/view?usp=sharing

Federico Pianzola

--
Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"
University of Milano-Bicocca

Senior Researcher – School of Media, Arts and Science
Sogang University, Seoul

Managing editor of Enthymema
International journal of literary criticism, literary theory, and philosophy of literature

_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: failed exporting inferred axioms

Campbell, James R

Federico

I found that the export of an inferred view by Protégé was very sensitive to the options chosen for what is to be exported.  I tried to create an inferred view several times before being successful.  I do not have the platform in front of me now but I will review in the morning and try to give more explicit advice

Jim Campbell

University of nebraska

 

Sent from Mail for Windows 10

 

From: [hidden email]
Sent: Tuesday, April 14, 2020 6:00 PM
To: [hidden email]
Subject: [protege-user] failed exporting inferred axioms

 

Non-UNMC email

I tried to export the inferred axioms obtained with Pellet but if I check the output file they are not visible.

I followed this guide https://www.michaeldebellis.com/post/export-inferred-axioms [michaeldebellis.com] but gave up ticking the last two boxes because it was taking way too much time.

 

Is anybody willing to give it a try? Here is the original ontology: https://drive.google.com/file/d/1yPdIUpIKkxxoj9Y0d6ejmGqhLGq7hUTn/view?usp=sharing [drive.google.com]

 

Federico Pianzola

 

--

Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"

University of Milano-Bicocca

 

Senior Researcher – School of Media, Arts and Science

Sogang University, Seoul

 

Managing editor of Enthymema

International journal of literary criticism, literary theory, and philosophy of literature

http://riviste.unimi.it/index.php/enthymema/index [riviste.unimi.it]

 


The information in this e-mail may be privileged and confidential, intended only for the use of the addressee(s) above. Any unauthorized use or disclosure of this information is prohibited. If you have received this e-mail by mistake, please delete it and immediately contact the sender.
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: failed exporting inferred axioms

Federico Pianzola
Thanks for your reply, James.
I have many object properties and owl:sameAs that have to be inferred for individuals but the reasoning with Pellet does not take more than a couple of minutes.

I will appreciate any help.

Federico

On Wed, 15 Apr 2020 at 10:47, Campbell, James R <[hidden email]> wrote:

Federico

I found that the export of an inferred view by Protégé was very sensitive to the options chosen for what is to be exported.  I tried to create an inferred view several times before being successful.  I do not have the platform in front of me now but I will review in the morning and try to give more explicit advice

Jim Campbell

University of nebraska

 

Sent from Mail for Windows 10

 

From: [hidden email]
Sent: Tuesday, April 14, 2020 6:00 PM
To: [hidden email]
Subject: [protege-user] failed exporting inferred axioms

 

Non-UNMC email

I tried to export the inferred axioms obtained with Pellet but if I check the output file they are not visible.

I followed this guide https://www.michaeldebellis.com/post/export-inferred-axioms [michaeldebellis.com] but gave up ticking the last two boxes because it was taking way too much time.

 

Is anybody willing to give it a try? Here is the original ontology: https://drive.google.com/file/d/1yPdIUpIKkxxoj9Y0d6ejmGqhLGq7hUTn/view?usp=sharing [drive.google.com]

 

Federico Pianzola

 

--

Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"

University of Milano-Bicocca

 

Senior Researcher – School of Media, Arts and Science

Sogang University, Seoul

 

Managing editor of Enthymema

International journal of literary criticism, literary theory, and philosophy of literature

http://riviste.unimi.it/index.php/enthymema/index [riviste.unimi.it]

 


The information in this e-mail may be privileged and confidential, intended only for the use of the addressee(s) above. Any unauthorized use or disclosure of this information is prohibited. If you have received this e-mail by mistake, please delete it and immediately contact the sender.
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user


--
Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"
University of Milano-Bicocca

Senior Researcher – School of Media, Arts and Science
Sogang University, Seoul

Managing editor of Enthymema
International journal of literary criticism, literary theory, and philosophy of literature

_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: failed exporting inferred axioms

samsontu
By “they are not visible” you meant that the inferred “same individuals as” axioms are not in the output file, right? That’s the result I got.

In the "Export inferred ontology" wizard, there is a “Select axioms to export” step that lists 12 types of axioms that can be exported. “Same individual as” is not in the list. Maybe that’s why the output file doesn’t contain this kind of inferred axioms.

With best regards,
Samson
 

On Apr 14, 2020, at 8:06 PM, Federico Pianzola <[hidden email]> wrote:

Thanks for your reply, James.
I have many object properties and owl:sameAs that have to be inferred for individuals but the reasoning with Pellet does not take more than a couple of minutes.

I will appreciate any help.

Federico

On Wed, 15 Apr 2020 at 10:47, Campbell, James R <[hidden email]> wrote:

Federico

I found that the export of an inferred view by Protégé was very sensitive to the options chosen for what is to be exported.  I tried to create an inferred view several times before being successful.  I do not have the platform in front of me now but I will review in the morning and try to give more explicit advice

Jim Campbell

University of nebraska

 

Sent from Mail for Windows 10

 

From: [hidden email]
Sent: Tuesday, April 14, 2020 6:00 PM
To: [hidden email]
Subject: [protege-user] failed exporting inferred axioms

 

Non-UNMC email

I tried to export the inferred axioms obtained with Pellet but if I check the output file they are not visible.

I followed this guide https://www.michaeldebellis.com/post/export-inferred-axioms [michaeldebellis.com] but gave up ticking the last two boxes because it was taking way too much time.

 

Is anybody willing to give it a try? Here is the original ontology: https://drive.google.com/file/d/1yPdIUpIKkxxoj9Y0d6ejmGqhLGq7hUTn/view?usp=sharing [drive.google.com]

 

Federico Pianzola

 

--

Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"

University of Milano-Bicocca

 

Senior Researcher – School of Media, Arts and Science

Sogang University, Seoul

 

Managing editor of Enthymema

International journal of literary criticism, literary theory, and philosophy of literature

http://riviste.unimi.it/index.php/enthymema/index [riviste.unimi.it]

 


The information in this e-mail may be privileged and confidential, intended only for the use of the addressee(s) above. Any unauthorized use or disclosure of this information is prohibited. If you have received this e-mail by mistake, please delete it and immediately contact the sender.
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user


--
Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"
University of Milano-Bicocca

Senior Researcher – School of Media, Arts and Science
Sogang University, Seoul

Managing editor of Enthymema
International journal of literary criticism, literary theory, and philosophy of literature
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user


_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: failed exporting inferred axioms

samsontu
Also, you ontology has a lot of property assertion inferences. When you said you "gave up ticking the last two boxes” (as I did), you’ve decided not to export these property assertion inferences. Does your ontology have any inferred axiom that’s not a sameAs or a property assertion axiom? If not, then it’s not surprising that the inferred axiom export won’t contain any inferred axioms.

The reason exporting property assertions takes too long is that the number of inferred property assertions in your ontology is probably humongous. Given a single asserted property assertion “A participatesTo B”, if A is same as 10 other individuals and B is same as 10 individuals, then you’ll generate 100 property assertion inferences from this single property assertion axiom. You have close to one million logical axioms. Most of them are probably sameAs or property assertion axioms. The combinatorics gets really bad.

Protege is not optimized for the kind of individual-heavy ontology you have. (Your ontology has 10 classes, 5 object properties, and > 33,000 individuals.) You may try something like Stardog, which has more optimizations for graphs with large number of individuals. For example, after performing “sameAs” reasoning, you can work with “canonical” individuals that represent sets of individuals that are the same [1], thus avoiding the combinatorics resulting from large number of individuals that are the same as each other.

With best regards,
Samson


On Apr 14, 2020, at 10:32 PM, Samson Tu <[hidden email]> wrote:

By “they are not visible” you meant that the inferred “same individuals as” axioms are not in the output file, right? That’s the result I got.

In the "Export inferred ontology" wizard, there is a “Select axioms to export” step that lists 12 types of axioms that can be exported. “Same individual as” is not in the list. Maybe that’s why the output file doesn’t contain this kind of inferred axioms.

With best regards,
Samson
 

On Apr 14, 2020, at 8:06 PM, Federico Pianzola <[hidden email]> wrote:

Thanks for your reply, James.
I have many object properties and owl:sameAs that have to be inferred for individuals but the reasoning with Pellet does not take more than a couple of minutes.

I will appreciate any help.

Federico

On Wed, 15 Apr 2020 at 10:47, Campbell, James R <[hidden email]> wrote:

Federico

I found that the export of an inferred view by Protégé was very sensitive to the options chosen for what is to be exported.  I tried to create an inferred view several times before being successful.  I do not have the platform in front of me now but I will review in the morning and try to give more explicit advice

Jim Campbell

University of nebraska

 

Sent from Mail for Windows 10

 

From: [hidden email]
Sent: Tuesday, April 14, 2020 6:00 PM
To: [hidden email]
Subject: [protege-user] failed exporting inferred axioms

 

Non-UNMC email

I tried to export the inferred axioms obtained with Pellet but if I check the output file they are not visible.

I followed this guide https://www.michaeldebellis.com/post/export-inferred-axioms [michaeldebellis.com] but gave up ticking the last two boxes because it was taking way too much time.

 

Is anybody willing to give it a try? Here is the original ontology: https://drive.google.com/file/d/1yPdIUpIKkxxoj9Y0d6ejmGqhLGq7hUTn/view?usp=sharing [drive.google.com]

 

Federico Pianzola

 

--

Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"

University of Milano-Bicocca

 

Senior Researcher – School of Media, Arts and Science

Sogang University, Seoul

 

Managing editor of Enthymema

International journal of literary criticism, literary theory, and philosophy of literature

http://riviste.unimi.it/index.php/enthymema/index [riviste.unimi.it]

 


The information in this e-mail may be privileged and confidential, intended only for the use of the addressee(s) above. Any unauthorized use or disclosure of this information is prohibited. If you have received this e-mail by mistake, please delete it and immediately contact the sender.
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user


--
Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"
University of Milano-Bicocca

Senior Researcher – School of Media, Arts and Science
Sogang University, Seoul

Managing editor of Enthymema
International journal of literary criticism, literary theory, and philosophy of literature
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user



_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: failed exporting inferred axioms

Lorenz Buehmann

That is correct. Nobody want's to compute the TC of owl:sameAs given that it's worst case size is what N^2 with N being the number of entities - like each entity is equivalent to each other in the data. Most tools tend to keep a single node per equivalence class (resp. owl:sameAs clique)

On 15.04.20 08:26, Samson Tu wrote:
Also, you ontology has a lot of property assertion inferences. When you said you "gave up ticking the last two boxes” (as I did), you’ve decided not to export these property assertion inferences. Does your ontology have any inferred axiom that’s not a sameAs or a property assertion axiom? If not, then it’s not surprising that the inferred axiom export won’t contain any inferred axioms.

The reason exporting property assertions takes too long is that the number of inferred property assertions in your ontology is probably humongous. Given a single asserted property assertion “A participatesTo B”, if A is same as 10 other individuals and B is same as 10 individuals, then you’ll generate 100 property assertion inferences from this single property assertion axiom. You have close to one million logical axioms. Most of them are probably sameAs or property assertion axioms. The combinatorics gets really bad.

Protege is not optimized for the kind of individual-heavy ontology you have. (Your ontology has 10 classes, 5 object properties, and > 33,000 individuals.) You may try something like Stardog, which has more optimizations for graphs with large number of individuals. For example, after performing “sameAs” reasoning, you can work with “canonical” individuals that represent sets of individuals that are the same [1], thus avoiding the combinatorics resulting from large number of individuals that are the same as each other.

With best regards,
Samson


On Apr 14, 2020, at 10:32 PM, Samson Tu <[hidden email]> wrote:

By “they are not visible” you meant that the inferred “same individuals as” axioms are not in the output file, right? That’s the result I got.

In the "Export inferred ontology" wizard, there is a “Select axioms to export” step that lists 12 types of axioms that can be exported. “Same individual as” is not in the list. Maybe that’s why the output file doesn’t contain this kind of inferred axioms.

With best regards,
Samson
 

On Apr 14, 2020, at 8:06 PM, Federico Pianzola <[hidden email]> wrote:

Thanks for your reply, James.
I have many object properties and owl:sameAs that have to be inferred for individuals but the reasoning with Pellet does not take more than a couple of minutes.

I will appreciate any help.

Federico

On Wed, 15 Apr 2020 at 10:47, Campbell, James R <[hidden email]> wrote:

Federico

I found that the export of an inferred view by Protégé was very sensitive to the options chosen for what is to be exported.  I tried to create an inferred view several times before being successful.  I do not have the platform in front of me now but I will review in the morning and try to give more explicit advice

Jim Campbell

University of nebraska

 

Sent from Mail for Windows 10

 

From: [hidden email]
Sent: Tuesday, April 14, 2020 6:00 PM
To: [hidden email]
Subject: [protege-user] failed exporting inferred axioms

 

Non-UNMC email

I tried to export the inferred axioms obtained with Pellet but if I check the output file they are not visible.

I followed this guide https://www.michaeldebellis.com/post/export-inferred-axioms [michaeldebellis.com] but gave up ticking the last two boxes because it was taking way too much time.

 

Is anybody willing to give it a try? Here is the original ontology: https://drive.google.com/file/d/1yPdIUpIKkxxoj9Y0d6ejmGqhLGq7hUTn/view?usp=sharing [drive.google.com]

 

Federico Pianzola

 

--

Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"

University of Milano-Bicocca

 

Senior Researcher – School of Media, Arts and Science

Sogang University, Seoul

 

Managing editor of Enthymema

International journal of literary criticism, literary theory, and philosophy of literature

http://riviste.unimi.it/index.php/enthymema/index [riviste.unimi.it]

 


The information in this e-mail may be privileged and confidential, intended only for the use of the addressee(s) above. Any unauthorized use or disclosure of this information is prohibited. If you have received this e-mail by mistake, please delete it and immediately contact the sender.
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user


--
Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"
University of Milano-Bicocca

Senior Researcher – School of Media, Arts and Science
Sogang University, Seoul

Managing editor of Enthymema
International journal of literary criticism, literary theory, and philosophy of literature
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user



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

_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: failed exporting inferred axioms

Federico Pianzola
Thank you everybody for your replies, I learned useful things.
I think I wouldn't use Stardog because it's not free, I'll try to run queries using the reasoner included in the R package SPARQL.

Federico

On Wed, 15 Apr 2020 at 15:53, Lorenz Buehmann <[hidden email]> wrote:

That is correct. Nobody want's to compute the TC of owl:sameAs given that it's worst case size is what N^2 with N being the number of entities - like each entity is equivalent to each other in the data. Most tools tend to keep a single node per equivalence class (resp. owl:sameAs clique)

On 15.04.20 08:26, Samson Tu wrote:
Also, you ontology has a lot of property assertion inferences. When you said you "gave up ticking the last two boxes” (as I did), you’ve decided not to export these property assertion inferences. Does your ontology have any inferred axiom that’s not a sameAs or a property assertion axiom? If not, then it’s not surprising that the inferred axiom export won’t contain any inferred axioms.

The reason exporting property assertions takes too long is that the number of inferred property assertions in your ontology is probably humongous. Given a single asserted property assertion “A participatesTo B”, if A is same as 10 other individuals and B is same as 10 individuals, then you’ll generate 100 property assertion inferences from this single property assertion axiom. You have close to one million logical axioms. Most of them are probably sameAs or property assertion axioms. The combinatorics gets really bad.

Protege is not optimized for the kind of individual-heavy ontology you have. (Your ontology has 10 classes, 5 object properties, and > 33,000 individuals.) You may try something like Stardog, which has more optimizations for graphs with large number of individuals. For example, after performing “sameAs” reasoning, you can work with “canonical” individuals that represent sets of individuals that are the same [1], thus avoiding the combinatorics resulting from large number of individuals that are the same as each other.

With best regards,
Samson


On Apr 14, 2020, at 10:32 PM, Samson Tu <[hidden email]> wrote:

By “they are not visible” you meant that the inferred “same individuals as” axioms are not in the output file, right? That’s the result I got.

In the "Export inferred ontology" wizard, there is a “Select axioms to export” step that lists 12 types of axioms that can be exported. “Same individual as” is not in the list. Maybe that’s why the output file doesn’t contain this kind of inferred axioms.

With best regards,
Samson
 

On Apr 14, 2020, at 8:06 PM, Federico Pianzola <[hidden email]> wrote:

Thanks for your reply, James.
I have many object properties and owl:sameAs that have to be inferred for individuals but the reasoning with Pellet does not take more than a couple of minutes.

I will appreciate any help.

Federico

On Wed, 15 Apr 2020 at 10:47, Campbell, James R <[hidden email]> wrote:

Federico

I found that the export of an inferred view by Protégé was very sensitive to the options chosen for what is to be exported.  I tried to create an inferred view several times before being successful.  I do not have the platform in front of me now but I will review in the morning and try to give more explicit advice

Jim Campbell

University of nebraska

 

Sent from Mail for Windows 10

 

From: [hidden email]
Sent: Tuesday, April 14, 2020 6:00 PM
To: [hidden email]
Subject: [protege-user] failed exporting inferred axioms

 

Non-UNMC email

I tried to export the inferred axioms obtained with Pellet but if I check the output file they are not visible.

I followed this guide https://www.michaeldebellis.com/post/export-inferred-axioms [michaeldebellis.com] but gave up ticking the last two boxes because it was taking way too much time.

 

Is anybody willing to give it a try? Here is the original ontology: https://drive.google.com/file/d/1yPdIUpIKkxxoj9Y0d6ejmGqhLGq7hUTn/view?usp=sharing [drive.google.com]

 

Federico Pianzola

 

--

Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"

University of Milano-Bicocca

 

Senior Researcher – School of Media, Arts and Science

Sogang University, Seoul

 

Managing editor of Enthymema

International journal of literary criticism, literary theory, and philosophy of literature

http://riviste.unimi.it/index.php/enthymema/index [riviste.unimi.it]

 


The information in this e-mail may be privileged and confidential, intended only for the use of the addressee(s) above. Any unauthorized use or disclosure of this information is prohibited. If you have received this e-mail by mistake, please delete it and immediately contact the sender.
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user


--
Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"
University of Milano-Bicocca

Senior Researcher – School of Media, Arts and Science
Sogang University, Seoul

Managing editor of Enthymema
International journal of literary criticism, literary theory, and philosophy of literature
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user



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


--
Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"
University of Milano-Bicocca

Senior Researcher – School of Media, Arts and Science
Sogang University, Seoul

Managing editor of Enthymema
International journal of literary criticism, literary theory, and philosophy of literature

_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: failed exporting inferred axioms

Michael DeBellis-2
I just saw this thread. The initial message got marked as spam by my overly aggressive gmail spam detector. I think you've already received good answers but just a few comments. 

First, thanks for the feedback on my blog post, I'll have to update it to make it clearer that what I say about exporting is meant for small to medium sized ontologies not ontologies with tens of thousands of individuals. 

Second, you might want to try the free version of the AllegroGraph triple store product from Franz Inc. Here is the link:  https://franz.com/agraph/downloads/?ui=new  

This is a great triple store that is designed to handle thousands of instances. Someone at Franz told me that even this free version can handle up to a few million triples. You can also use it with their Gruff product which is a great visualization tool, and also designed for viewing very large ontologies.  I think you can do SPARQL queries with Gruff as well. The only thing is there are a few OWL2 features that the reasoner in AllegroGraph doesn't support. Primarily SWRL is not supported and I think a few others as well but for the project I used it on we found it did everything we needed and the performance for a very large ontology was very good. 

Michael

On Wed, Apr 15, 2020 at 3:28 AM Federico Pianzola <[hidden email]> wrote:
Thank you everybody for your replies, I learned useful things.
I think I wouldn't use Stardog because it's not free, I'll try to run queries using the reasoner included in the R package SPARQL.

Federico

On Wed, 15 Apr 2020 at 15:53, Lorenz Buehmann <[hidden email]> wrote:

That is correct. Nobody want's to compute the TC of owl:sameAs given that it's worst case size is what N^2 with N being the number of entities - like each entity is equivalent to each other in the data. Most tools tend to keep a single node per equivalence class (resp. owl:sameAs clique)

On 15.04.20 08:26, Samson Tu wrote:
Also, you ontology has a lot of property assertion inferences. When you said you "gave up ticking the last two boxes” (as I did), you’ve decided not to export these property assertion inferences. Does your ontology have any inferred axiom that’s not a sameAs or a property assertion axiom? If not, then it’s not surprising that the inferred axiom export won’t contain any inferred axioms.

The reason exporting property assertions takes too long is that the number of inferred property assertions in your ontology is probably humongous. Given a single asserted property assertion “A participatesTo B”, if A is same as 10 other individuals and B is same as 10 individuals, then you’ll generate 100 property assertion inferences from this single property assertion axiom. You have close to one million logical axioms. Most of them are probably sameAs or property assertion axioms. The combinatorics gets really bad.

Protege is not optimized for the kind of individual-heavy ontology you have. (Your ontology has 10 classes, 5 object properties, and > 33,000 individuals.) You may try something like Stardog, which has more optimizations for graphs with large number of individuals. For example, after performing “sameAs” reasoning, you can work with “canonical” individuals that represent sets of individuals that are the same [1], thus avoiding the combinatorics resulting from large number of individuals that are the same as each other.

With best regards,
Samson


On Apr 14, 2020, at 10:32 PM, Samson Tu <[hidden email]> wrote:

By “they are not visible” you meant that the inferred “same individuals as” axioms are not in the output file, right? That’s the result I got.

In the "Export inferred ontology" wizard, there is a “Select axioms to export” step that lists 12 types of axioms that can be exported. “Same individual as” is not in the list. Maybe that’s why the output file doesn’t contain this kind of inferred axioms.

With best regards,
Samson
 

On Apr 14, 2020, at 8:06 PM, Federico Pianzola <[hidden email]> wrote:

Thanks for your reply, James.
I have many object properties and owl:sameAs that have to be inferred for individuals but the reasoning with Pellet does not take more than a couple of minutes.

I will appreciate any help.

Federico

On Wed, 15 Apr 2020 at 10:47, Campbell, James R <[hidden email]> wrote:

Federico

I found that the export of an inferred view by Protégé was very sensitive to the options chosen for what is to be exported.  I tried to create an inferred view several times before being successful.  I do not have the platform in front of me now but I will review in the morning and try to give more explicit advice

Jim Campbell

University of nebraska

 

Sent from Mail for Windows 10

 

From: [hidden email]
Sent: Tuesday, April 14, 2020 6:00 PM
To: [hidden email]
Subject: [protege-user] failed exporting inferred axioms

 

Non-UNMC email

I tried to export the inferred axioms obtained with Pellet but if I check the output file they are not visible.

I followed this guide https://www.michaeldebellis.com/post/export-inferred-axioms [michaeldebellis.com] but gave up ticking the last two boxes because it was taking way too much time.

 

Is anybody willing to give it a try? Here is the original ontology: https://drive.google.com/file/d/1yPdIUpIKkxxoj9Y0d6ejmGqhLGq7hUTn/view?usp=sharing [drive.google.com]

 

Federico Pianzola

 

--

Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"

University of Milano-Bicocca

 

Senior Researcher – School of Media, Arts and Science

Sogang University, Seoul

 

Managing editor of Enthymema

International journal of literary criticism, literary theory, and philosophy of literature

http://riviste.unimi.it/index.php/enthymema/index [riviste.unimi.it]

 


The information in this e-mail may be privileged and confidential, intended only for the use of the addressee(s) above. Any unauthorized use or disclosure of this information is prohibited. If you have received this e-mail by mistake, please delete it and immediately contact the sender.
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user


--
Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"
University of Milano-Bicocca

Senior Researcher – School of Media, Arts and Science
Sogang University, Seoul

Managing editor of Enthymema
International journal of literary criticism, literary theory, and philosophy of literature
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user



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


--
Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"
University of Milano-Bicocca

Senior Researcher – School of Media, Arts and Science
Sogang University, Seoul

Managing editor of Enthymema
International journal of literary criticism, literary theory, and philosophy of literature
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user

_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: failed exporting inferred axioms

Federico Pianzola
Thanks for the feedback, Michael. I'll try AllegroGraph, too.

Federico

On Wed, 15 Apr 2020 at 23:56, Michael DeBellis <[hidden email]> wrote:
I just saw this thread. The initial message got marked as spam by my overly aggressive gmail spam detector. I think you've already received good answers but just a few comments. 

First, thanks for the feedback on my blog post, I'll have to update it to make it clearer that what I say about exporting is meant for small to medium sized ontologies not ontologies with tens of thousands of individuals. 

Second, you might want to try the free version of the AllegroGraph triple store product from Franz Inc. Here is the link:  https://franz.com/agraph/downloads/?ui=new  

This is a great triple store that is designed to handle thousands of instances. Someone at Franz told me that even this free version can handle up to a few million triples. You can also use it with their Gruff product which is a great visualization tool, and also designed for viewing very large ontologies.  I think you can do SPARQL queries with Gruff as well. The only thing is there are a few OWL2 features that the reasoner in AllegroGraph doesn't support. Primarily SWRL is not supported and I think a few others as well but for the project I used it on we found it did everything we needed and the performance for a very large ontology was very good. 

Michael

On Wed, Apr 15, 2020 at 3:28 AM Federico Pianzola <[hidden email]> wrote:
Thank you everybody for your replies, I learned useful things.
I think I wouldn't use Stardog because it's not free, I'll try to run queries using the reasoner included in the R package SPARQL.

Federico

On Wed, 15 Apr 2020 at 15:53, Lorenz Buehmann <[hidden email]> wrote:

That is correct. Nobody want's to compute the TC of owl:sameAs given that it's worst case size is what N^2 with N being the number of entities - like each entity is equivalent to each other in the data. Most tools tend to keep a single node per equivalence class (resp. owl:sameAs clique)

On 15.04.20 08:26, Samson Tu wrote:
Also, you ontology has a lot of property assertion inferences. When you said you "gave up ticking the last two boxes” (as I did), you’ve decided not to export these property assertion inferences. Does your ontology have any inferred axiom that’s not a sameAs or a property assertion axiom? If not, then it’s not surprising that the inferred axiom export won’t contain any inferred axioms.

The reason exporting property assertions takes too long is that the number of inferred property assertions in your ontology is probably humongous. Given a single asserted property assertion “A participatesTo B”, if A is same as 10 other individuals and B is same as 10 individuals, then you’ll generate 100 property assertion inferences from this single property assertion axiom. You have close to one million logical axioms. Most of them are probably sameAs or property assertion axioms. The combinatorics gets really bad.

Protege is not optimized for the kind of individual-heavy ontology you have. (Your ontology has 10 classes, 5 object properties, and > 33,000 individuals.) You may try something like Stardog, which has more optimizations for graphs with large number of individuals. For example, after performing “sameAs” reasoning, you can work with “canonical” individuals that represent sets of individuals that are the same [1], thus avoiding the combinatorics resulting from large number of individuals that are the same as each other.

With best regards,
Samson


On Apr 14, 2020, at 10:32 PM, Samson Tu <[hidden email]> wrote:

By “they are not visible” you meant that the inferred “same individuals as” axioms are not in the output file, right? That’s the result I got.

In the "Export inferred ontology" wizard, there is a “Select axioms to export” step that lists 12 types of axioms that can be exported. “Same individual as” is not in the list. Maybe that’s why the output file doesn’t contain this kind of inferred axioms.

With best regards,
Samson
 

On Apr 14, 2020, at 8:06 PM, Federico Pianzola <[hidden email]> wrote:

Thanks for your reply, James.
I have many object properties and owl:sameAs that have to be inferred for individuals but the reasoning with Pellet does not take more than a couple of minutes.

I will appreciate any help.

Federico

On Wed, 15 Apr 2020 at 10:47, Campbell, James R <[hidden email]> wrote:

Federico

I found that the export of an inferred view by Protégé was very sensitive to the options chosen for what is to be exported.  I tried to create an inferred view several times before being successful.  I do not have the platform in front of me now but I will review in the morning and try to give more explicit advice

Jim Campbell

University of nebraska

 

Sent from Mail for Windows 10

 

From: [hidden email]
Sent: Tuesday, April 14, 2020 6:00 PM
To: [hidden email]
Subject: [protege-user] failed exporting inferred axioms

 

Non-UNMC email

I tried to export the inferred axioms obtained with Pellet but if I check the output file they are not visible.

I followed this guide https://www.michaeldebellis.com/post/export-inferred-axioms [michaeldebellis.com] but gave up ticking the last two boxes because it was taking way too much time.

 

Is anybody willing to give it a try? Here is the original ontology: https://drive.google.com/file/d/1yPdIUpIKkxxoj9Y0d6ejmGqhLGq7hUTn/view?usp=sharing [drive.google.com]

 

Federico Pianzola

 

--

Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"

University of Milano-Bicocca

 

Senior Researcher – School of Media, Arts and Science

Sogang University, Seoul

 

Managing editor of Enthymema

International journal of literary criticism, literary theory, and philosophy of literature

http://riviste.unimi.it/index.php/enthymema/index [riviste.unimi.it]

 


The information in this e-mail may be privileged and confidential, intended only for the use of the addressee(s) above. Any unauthorized use or disclosure of this information is prohibited. If you have received this e-mail by mistake, please delete it and immediately contact the sender.
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user


--
Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"
University of Milano-Bicocca

Senior Researcher – School of Media, Arts and Science
Sogang University, Seoul

Managing editor of Enthymema
International journal of literary criticism, literary theory, and philosophy of literature
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user



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


--
Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"
University of Milano-Bicocca

Senior Researcher – School of Media, Arts and Science
Sogang University, Seoul

Managing editor of Enthymema
International journal of literary criticism, literary theory, and philosophy of literature
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user


--
Marie Skłodowska Curie Research Fellow – Department of Human Sciences for Education "R. Massa"
University of Milano-Bicocca

Senior Researcher – School of Media, Arts and Science
Sogang University, Seoul

Managing editor of Enthymema
International journal of literary criticism, literary theory, and philosophy of literature

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