Strange Behavior in Protege 5.2

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

Strange Behavior in Protege 5.2

Michael DeBellis-2
I'm using the Windows version of 5.2.0. I'm creating a small model to make a point about something that George Lakoff wrote in his book Women, Fire, and Dangerous Things. He describes various different views of "Mother" and claims that standard logic can't represent these various views. As I read them I kept thinking "I could model that in Protege" and have started to do so. 

I defined the basic concept of Mother as a defined class with: "Female and isMotherOf some Human" as the definition. Then for one of the subclasses of Mother I have BirthMother. Also, as a sub-property of isMotherOf I have isBirthMotherOf (with domain BirthMother and range Human, isMotherOf has domain Mother). So I added the definition "Female and isBirthMotherOf some Human" to the class BirthMother. I was planning on making BirthMother a defined class. But when I made this change Protege seems to have changed the definition of Mother to be the same as BirthMother, it went from  "Female ans isMotherOf some Human" to  "Female and isBirthMotherOf some Human" and then it made Mother and BirthMother equivalent classes. 

I tried this for another similar property (isStepMotherOf) and got the same behavior. Is there something I'm missing, some reason it would do this? Attached is the ontology with the original definition of Mother (2-14-19) and then the same ontology after I made the change to BirthMother which changed the definition of Mother (2-14-19a).

Michael


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

MotherOfAllModels-2-14-19.owl (31K) Download Attachment
MotherOfAllModels-2-14-19a.owl (31K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Strange Behavior in Protege 5.2

samsontu
Hi Michael,

When I added “Female and isBirthMotherOf some Human” as the definition of BirthMother in your MotherOfAllModels-2-14-19.owl ontology, I didn’t get the conflation of Mother and BirthMother that you described in your message.

When I loaded MotherOfAllModels-2-14-19a.owl, I saw that Mother has the definition "Female and (isBirthMotherOf some Human)”.

I’ve never seen the behavior you described, and I can’t reproduce it in your MotherOfAllModels-2-14-19.owl ontology (in Protege 5.2).

But when I made this change Protege seems to have changed the definition of Mother to be the same as BirthMother,

With best regards,
Samson


On Feb 14, 2019, at 4:30 PM, Michael DeBellis <[hidden email]> wrote:

I'm using the Windows version of 5.2.0. I'm creating a small model to make a point about something that George Lakoff wrote in his book Women, Fire, and Dangerous Things. He describes various different views of "Mother" and claims that standard logic can't represent these various views. As I read them I kept thinking "I could model that in Protege" and have started to do so. 

I defined the basic concept of Mother as a defined class with: "Female and isMotherOf some Human" as the definition. Then for one of the subclasses of Mother I have BirthMother. Also, as a sub-property of isMotherOf I have isBirthMotherOf (with domain BirthMother and range Human, isMotherOf has domain Mother). So I added the definition "Female and isBirthMotherOf some Human" to the class BirthMother. I was planning on making BirthMother a defined class. But when I made this change Protege seems to have changed the definition of Mother to be the same as BirthMother, it went from  "Female ans isMotherOf some Human" to  "Female and isBirthMotherOf some Human" and then it made Mother and BirthMother equivalent classes. 

I tried this for another similar property (isStepMotherOf) and got the same behavior. Is there something I'm missing, some reason it would do this? Attached is the ontology with the original definition of Mother (2-14-19) and then the same ontology after I made the change to BirthMother which changed the definition of Mother (2-14-19a).

Michael

<MotherOfAllModels-2-14-19.owl><MotherOfAllModels-2-14-19a.owl>_______________________________________________
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

smime.p7s (1K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Strange Behavior in Protege 5.2

Michael DeBellis-2
Samson, it seemed very odd to me. I tried it multiple times and the "a" version of the ontology was just after I made that change to BirthMother without editing Mother and it changed Mother  for some reason.  Perhaps there was something about my environment that was corrupted or I was just really having a bad day and kept typing the wrong thing but I'm almost sure that last wasn't the case.  At least I know that it is not supposed to be happening,  I'll try some more debugging options here or perhaps just starting from a new ontology.  BTW, I'm working in Windows 10 if that makes a difference. Thanks for trying to replicate the error.

Michael

On Sat, Feb 16, 2019 at 11:29 AM Samson Tu <[hidden email]> wrote:
Hi Michael,

When I added “Female and isBirthMotherOf some Human” as the definition of BirthMother in your MotherOfAllModels-2-14-19.owl ontology, I didn’t get the conflation of Mother and BirthMother that you described in your message.

When I loaded MotherOfAllModels-2-14-19a.owl, I saw that Mother has the definition "Female and (isBirthMotherOf some Human)”.

I’ve never seen the behavior you described, and I can’t reproduce it in your MotherOfAllModels-2-14-19.owl ontology (in Protege 5.2).

But when I made this change Protege seems to have changed the definition of Mother to be the same as BirthMother,

With best regards,
Samson


On Feb 14, 2019, at 4:30 PM, Michael DeBellis <[hidden email]> wrote:

I'm using the Windows version of 5.2.0. I'm creating a small model to make a point about something that George Lakoff wrote in his book Women, Fire, and Dangerous Things. He describes various different views of "Mother" and claims that standard logic can't represent these various views. As I read them I kept thinking "I could model that in Protege" and have started to do so. 

I defined the basic concept of Mother as a defined class with: "Female and isMotherOf some Human" as the definition. Then for one of the subclasses of Mother I have BirthMother. Also, as a sub-property of isMotherOf I have isBirthMotherOf (with domain BirthMother and range Human, isMotherOf has domain Mother). So I added the definition "Female and isBirthMotherOf some Human" to the class BirthMother. I was planning on making BirthMother a defined class. But when I made this change Protege seems to have changed the definition of Mother to be the same as BirthMother, it went from  "Female ans isMotherOf some Human" to  "Female and isBirthMotherOf some Human" and then it made Mother and BirthMother equivalent classes. 

I tried this for another similar property (isStepMotherOf) and got the same behavior. Is there something I'm missing, some reason it would do this? Attached is the ontology with the original definition of Mother (2-14-19) and then the same ontology after I made the change to BirthMother which changed the definition of Mother (2-14-19a).

Michael

<MotherOfAllModels-2-14-19.owl><MotherOfAllModels-2-14-19a.owl>_______________________________________________
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: Strange Behavior in Protege 5.2

Michael DeBellis-2
In reply to this post by samsontu
Samson, Just FYI, I don't know what was going on with my system. I tried it again and every time I made the change to the subclass BirthMother for some reason it changed the definition (and made it seem as if it were a user change and not a reasoner change) and it altered the DL definition of the Mother superclass to make it equivalent to BirthMother. I found a way around it though, I changed BirthMother so that it wasn't initially defined as a subclass of Mother, just as a subclass of owl:Thing and then I gave it the appropriate definition and made it a defined class and it all worked as expected, making it a subclass of Mother. It's working fine I've extended the ontology a lot now and no other strange behavior.

Michael

On Sat, Feb 16, 2019 at 11:29 AM Samson Tu <[hidden email]> wrote:
Hi Michael,

When I added “Female and isBirthMotherOf some Human” as the definition of BirthMother in your MotherOfAllModels-2-14-19.owl ontology, I didn’t get the conflation of Mother and BirthMother that you described in your message.

When I loaded MotherOfAllModels-2-14-19a.owl, I saw that Mother has the definition "Female and (isBirthMotherOf some Human)”.

I’ve never seen the behavior you described, and I can’t reproduce it in your MotherOfAllModels-2-14-19.owl ontology (in Protege 5.2).

But when I made this change Protege seems to have changed the definition of Mother to be the same as BirthMother,

With best regards,
Samson


On Feb 14, 2019, at 4:30 PM, Michael DeBellis <[hidden email]> wrote:

I'm using the Windows version of 5.2.0. I'm creating a small model to make a point about something that George Lakoff wrote in his book Women, Fire, and Dangerous Things. He describes various different views of "Mother" and claims that standard logic can't represent these various views. As I read them I kept thinking "I could model that in Protege" and have started to do so. 

I defined the basic concept of Mother as a defined class with: "Female and isMotherOf some Human" as the definition. Then for one of the subclasses of Mother I have BirthMother. Also, as a sub-property of isMotherOf I have isBirthMotherOf (with domain BirthMother and range Human, isMotherOf has domain Mother). So I added the definition "Female and isBirthMotherOf some Human" to the class BirthMother. I was planning on making BirthMother a defined class. But when I made this change Protege seems to have changed the definition of Mother to be the same as BirthMother, it went from  "Female ans isMotherOf some Human" to  "Female and isBirthMotherOf some Human" and then it made Mother and BirthMother equivalent classes. 

I tried this for another similar property (isStepMotherOf) and got the same behavior. Is there something I'm missing, some reason it would do this? Attached is the ontology with the original definition of Mother (2-14-19) and then the same ontology after I made the change to BirthMother which changed the definition of Mother (2-14-19a).

Michael

<MotherOfAllModels-2-14-19.owl><MotherOfAllModels-2-14-19a.owl>_______________________________________________
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