[protege-owl] using wordnet

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

[protege-owl] using wordnet

Thakker, Dhavalkumar
Dear list members,

We want to use already available lexical reference system from the WORDNET(http://wordnet.princeton.edu/). Are there alrady available WORDNET equivalent OWL ontologies . We know of SUMO OWL(http://ontology.teknowledge.com/), which however has limited semantics comapare to WORDNET or SUMO in KIF.

any comments?

many thanks

Dhavalkumar Thakker(kumar)
PhD candidate
School of Computing & Informatics
Nottingham Trent University, Nottingham, UK
(0) 0044 115 848 8387
http://clarinet.doc.ntu.ac.uk:2000/dysec/dthakker.html


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

[protege-owl] Re: using wordnet

Andreas Langegger
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,

I once played around with the OWL version of wordnet published by
http://truesense.net/#Wordnet and Jena. Maybe you can use it.

Regards,
Andy

Thakker, Dhavalkumar wrote:

> Dear list members,
>
> We want to use already available lexical reference system from the WORDNET(http://wordnet.princeton.edu/). Are there alrady available WORDNET equivalent OWL ontologies . We know of SUMO OWL(http://ontology.teknowledge.com/), which however has limited semantics comapare to WORDNET or SUMO in KIF.
>
> any comments?
>
> many thanks
>
> Dhavalkumar Thakker(kumar)
> PhD candidate
> School of Computing & Informatics
> Nottingham Trent University, Nottingham, UK
> (0) 0044 115 848 8387
> http://clarinet.doc.ntu.ac.uk:2000/dysec/dthakker.html
>
>
> -------------------------------------------------------------------------
> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
>

- --
- ----------------------------------------------------------------------
Dipl.-Ing.(FH) Andreas Langegger
Institute for Applied Knowledge Processing
Johannes Kepler University Linz
A-4040 Linz, Altenberger Straße 69
> http://www.faw.at
> http://www.langegger.at
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.2 (GNU/Linux)

iD8DBQFE34LVKk9SuaNc5+IRAp8hAKCbt5S40b6syfoE8gPTwDxUFe//JwCfRtwL
VRrk9azUR9jmHHXQwVbsqPk=
=u3cv
-----END PGP SIGNATURE-----
-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

[protege-owl] Re: using wordnet

Thakker, Dhavalkumar
Thanks Andreas

It is very useful.

best regards,

Dhavalkumar Thakker(kumar)
PhD candidate
School of Computing & Informatics
Nottingham Trent University, Nottingham, UK
(0) 0044 115 848 8387
http://clarinet.doc.ntu.ac.uk:2000/dysec/dthakker.html




-----Original Message-----
From: [hidden email] on behalf of Andreas Langegger
Sent: Sun 13/08/2006 20:51
To: [hidden email]
Subject: [protege-owl] Re: using wordnet
 
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,

I once played around with the OWL version of wordnet published by
http://truesense.net/#Wordnet and Jena. Maybe you can use it.

Regards,
Andy

Thakker, Dhavalkumar wrote:

> Dear list members,
>
> We want to use already available lexical reference system from the WORDNET(http://wordnet.princeton.edu/). Are there alrady available WORDNET equivalent OWL ontologies . We know of SUMO OWL(http://ontology.teknowledge.com/), which however has limited semantics comapare to WORDNET or SUMO in KIF.
>
> any comments?
>
> many thanks
>
> Dhavalkumar Thakker(kumar)
> PhD candidate
> School of Computing & Informatics
> Nottingham Trent University, Nottingham, UK
> (0) 0044 115 848 8387
> http://clarinet.doc.ntu.ac.uk:2000/dysec/dthakker.html
>
>
> -------------------------------------------------------------------------
> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
>
- --
- ----------------------------------------------------------------------
Dipl.-Ing.(FH) Andreas Langegger
Institute for Applied Knowledge Processing
Johannes Kepler University Linz
A-4040 Linz, Altenberger Straße 69
> http://www.faw.at
> http://www.langegger.at
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.2 (GNU/Linux)

iD8DBQFE34LVKk9SuaNc5+IRAp8hAKCbt5S40b6syfoE8gPTwDxUFe//JwCfRtwL
VRrk9azUR9jmHHXQwVbsqPk=
=u3cv
-----END PGP SIGNATURE-----
-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html



winmail.dat (4K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

[protege-owl] Re: using wordnet

Steve Wartik
In reply to this post by Andreas Langegger
Anyone ever get Protege to load this ontology? I just tried, and Protege ran out of memory even after I doubled the maximum stack size, somewhere around the 250,000th triple. If you know the magic number, please pass it on. Thanks.

Steve Wartik

Andreas Langegger wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,

I once played around with the OWL version of wordnet published by
http://truesense.net/#Wordnet and Jena. Maybe you can use it.

Regards,
Andy

Thakker, Dhavalkumar wrote:
  
Dear list members,

We want to use already available lexical reference system from the WORDNET(http://wordnet.princeton.edu/). Are there alrady available WORDNET equivalent OWL ontologies . We know of SUMO OWL(http://ontology.teknowledge.com/), which however has limited semantics comapare to WORDNET or SUMO in KIF.

any comments?

many thanks

Dhavalkumar Thakker(kumar)
PhD candidate
School of Computing & Informatics
Nottingham Trent University, Nottingham, UK
(0) 0044 115 848 8387
http://clarinet.doc.ntu.ac.uk:2000/dysec/dthakker.html


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

    

- --
- ----------------------------------------------------------------------
Dipl.-Ing.(FH) Andreas Langegger
Institute for Applied Knowledge Processing
Johannes Kepler University Linz
A-4040 Linz, Altenberger Straße 69
  
http://www.faw.at
http://www.langegger.at
    
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.2 (GNU/Linux)

iD8DBQFE34LVKk9SuaNc5+IRAp8hAKCbt5S40b6syfoE8gPTwDxUFe//JwCfRtwL
VRrk9azUR9jmHHXQwVbsqPk=
=u3cv
-----END PGP SIGNATURE-----
-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html


  

Reply | Threaded
Open this post in threaded view
|

[protege-owl] Re: using wordnet

Jennifer Vendetti-2
Hello Steven,

I'm using 3.2 beta, build 331.  I modified the following settings in the Protege.lax file as follows:

lax.nl.java.option.java.heap.size.initial=100000000
lax.nl.java.option.java.heap.size.max=1000000000

It took a while - but I was able to successfully load "wordnet.01.owl".  See console output for details:

Loaded 675000 triples
[TripleChangePostProcessor] Completed lists after 0 ms
[TripleChangePostProcessor] Completed anonymous classes after 16 ms
[TripleChangePostProcessor] Completed deprecated classes after 0 ms
[TripleChangePostProcessor] Completed properties after 47 ms
[TripleChangePostProcessor] Completed named classes after 2296 ms
... Loading completed after 1805883 ms

Jennifer

Steven Wartik wrote:
Anyone ever get Protege to load this ontology? I just tried, and Protege ran out of memory even after I doubled the maximum stack size, somewhere around the 250,000th triple. If you know the magic number, please pass it on. Thanks.

Steve Wartik
------------------------------------------------------------------------- To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
Reply | Threaded
Open this post in threaded view
|

[protege-owl] Re: using wordnet

Andreas Langegger
In reply to this post by Steve Wartik
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I also tried with 1GB of heap size - Protégé didn't respond, killed it.
I think it's not a problem of memory, also of cpu power.
Andy

Steven Wartik wrote:

> Anyone ever get Protege to load this ontology? I just tried, and Protege
> ran out of memory even after I doubled the maximum stack size, somewhere
> around the 250,000th triple. If you know the magic number, please pass
> it on. Thanks.
>
> Steve Wartik
>
> Andreas Langegger wrote:
> Hi,
>
> I once played around with the OWL version of wordnet published by
> http://truesense.net/#Wordnet and Jena. Maybe you can use it.
>
> Regards,
> Andy
>
> Thakker, Dhavalkumar wrote:
>  
>>>> Dear list members,
>>>>
>>>> We want to use already available lexical reference system from the WORDNET(http://wordnet.princeton.edu/). Are there alrady available WORDNET equivalent OWL ontologies . We know of SUMO OWL(http://ontology.teknowledge.com/), which however has limited semantics comapare to WORDNET or SUMO in KIF.
>>>>
>>>> any comments?
>>>>
>>>> many thanks
>>>>
>>>> Dhavalkumar Thakker(kumar)
>>>> PhD candidate
>>>> School of Computing & Informatics
>>>> Nottingham Trent University, Nottingham, UK
>>>> (0) 0044 115 848 8387
>>>> http://clarinet.doc.ntu.ac.uk:2000/dysec/dthakker.html
>>>>
>>>>
>>>> -------------------------------------------------------------------------
>>>> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
>>>>
>>>>    
>
> --
> ----------------------------------------------------------------------
> Dipl.-Ing.(FH) Andreas Langegger
> Institute for Applied Knowledge Processing
> Johannes Kepler University Linz
> A-4040 Linz, Altenberger Straße 69
>  
>>>> http://www.faw.at
>>>> http://www.langegger.at
>>>>    
- -------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
>>
>>

- --
- ----------------------------------------------------------------------
Dipl.-Ing.(FH) Andreas Langegger
Institute for Applied Knowledge Processing
Johannes Kepler University Linz
A-4040 Linz, Altenberger Straße 69
> http://www.faw.at
> http://www.langegger.at
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.2 (GNU/Linux)

iD8DBQFE4ZaxKk9SuaNc5+IRAtAFAJ9Nr3Oqbx2ND+khS6d1gTN97Jc9OwCfeetT
46Tpx6DkISbmnEb5Sv0PIfk=
=EV2Z
-----END PGP SIGNATURE-----
-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

[protege-owl] Re: using wordnet

Steve Wartik
In reply to this post by Jennifer Vendetti-2
Jennifer,

Thanks for the numbers -- a heap size of 1G worked for me. Did you time how long Protege took to crunch through all the triples? I first let it run for half an hour, then gave up and re-started it when I left work yesterday evening. This morning, there was the WordNet ontology in all its hugeness. It's one way to view data, if not exactly real-time response.

Steve Wartik

Jennifer Vendetti wrote:
Hello Steven,

I'm using 3.2 beta, build 331.  I modified the following settings in the Protege.lax file as follows:

lax.nl.java.option.java.heap.size.initial=100000000
lax.nl.java.option.java.heap.size.max=1000000000

It took a while - but I was able to successfully load "wordnet.01.owl".  See console output for details:

Loaded 675000 triples
[TripleChangePostProcessor] Completed lists after 0 ms
[TripleChangePostProcessor] Completed anonymous classes after 16 ms
[TripleChangePostProcessor] Completed deprecated classes after 0 ms
[TripleChangePostProcessor] Completed properties after 47 ms
[TripleChangePostProcessor] Completed named classes after 2296 ms
... Loading completed after 1805883 ms

Jennifer

Steven Wartik wrote:
Anyone ever get Protege to load this ontology? I just tried, and Protege ran out of memory even after I doubled the maximum stack size, somewhere around the 250,000th triple. If you know the magic number, please pass it on. Thanks.

Steve Wartik
------------------------------------------------------------------------- To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

[protege-owl] Wordnet and persistent storage

Andreas Langegger
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi Steve, hi Jennifer,

on my 1,6GHz Centrino, after loading the triples I was waiting for about
2 hours until I gave up and killed the process.

Jennifer, do you know, how the DB-backend is working. I guess Protégé
has to load all triples from the DB, hasn't it? Otherwise, maybe this
would be an alternative working with wordnet. Steve, maybe you could
have a look on Cyg also if you haven't already.

I'm still wondering why DIG reasoners are integrated over TCP - this
means that all concepts have to be transmitted to a from the reasoner
over sockets, although it may be a local linux sock, it will need extra
memory. Can anybody tell my why the DL impl group decided to do so?

Thanks in advance,
Andy

Steven Wartik wrote:

> Jennifer,
>
> Thanks for the numbers -- a heap size of 1G worked for me. Did you time
> how long Protege took to crunch through all the triples? I first let it
> run for half an hour, then gave up and re-started it when I left work
> yesterday evening. This morning, there was the WordNet ontology in all
> its hugeness. It's one way to view data, if not exactly real-time response.
>
> Steve Wartik
>
> Jennifer Vendetti wrote:
>> Hello Steven,
>>
>> I'm using 3.2 beta, build 331.  I modified the following settings in
>> the Protege.lax file as follows:
>>
>> lax.nl.java.option.java.heap.size.initial=100000000
>> lax.nl.java.option.java.heap.size.max=1000000000
>>
>> It took a while - but I was able to successfully load
>> "wordnet.01.owl".  See console output for details:
>>
>> Loaded 675000 triples
>> [TripleChangePostProcessor] Completed lists after 0 ms
>> [TripleChangePostProcessor] Completed anonymous classes after 16 ms
>> [TripleChangePostProcessor] Completed deprecated classes after 0 ms
>> [TripleChangePostProcessor] Completed properties after 47 ms
>> [TripleChangePostProcessor] Completed named classes after 2296 ms
>> ... Loading completed after 1805883 ms
>>
>> Jennifer
>>
>> Steven Wartik wrote:
>>> Anyone ever get Protege to load this ontology? I just tried, and
>>> Protege ran out of memory even after I doubled the maximum stack
>>> size, somewhere around the 250,000th triple. If you know the magic
>>> number, please pass it on. Thanks.
>>>
>>> Steve Wartik
>> -------------------------------------------------------------------------
>> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html 
>

- --
- ----------------------------------------------------------------------
Dipl.-Ing.(FH) Andreas Langegger
Institute for Applied Knowledge Processing
Johannes Kepler University Linz
A-4040 Linz, Altenberger Straße 69
> http://www.faw.at
> http://www.langegger.at
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.2 (GNU/Linux)

iD8DBQFE4zC3Kk9SuaNc5+IRAgB6AJ4wSNSS2YUC9eRTBL3Zj1j6lohbRgCfW+Ti
Z+EpmKiM7jufyHbIII2ZWCY=
=vtHX
-----END PGP SIGNATURE-----
-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

[protege-owl] Re: Wordnet and persistent storage

Steve Wartik
Andreas,

I'm not brave enough to study Protégé's memory management strategies. I was however interested to see that the Java process on my computer grew to over 750 MB as the ontology was loading, then dropped down to about 140 MB after loading completed.

Steve Wartik

Andreas Langegger wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi Steve, hi Jennifer,

on my 1,6GHz Centrino, after loading the triples I was waiting for about
2 hours until I gave up and killed the process.

Jennifer, do you know, how the DB-backend is working. I guess Protégé
has to load all triples from the DB, hasn't it? Otherwise, maybe this
would be an alternative working with wordnet. Steve, maybe you could
have a look on Cyg also if you haven't already.

I'm still wondering why DIG reasoners are integrated over TCP - this
means that all concepts have to be transmitted to a from the reasoner
over sockets, although it may be a local linux sock, it will need extra
memory. Can anybody tell my why the DL impl group decided to do so?

Thanks in advance,
Andy

Steven Wartik wrote:
  
Jennifer,

Thanks for the numbers -- a heap size of 1G worked for me. Did you time
how long Protege took to crunch through all the triples? I first let it
run for half an hour, then gave up and re-started it when I left work
yesterday evening. This morning, there was the WordNet ontology in all
its hugeness. It's one way to view data, if not exactly real-time response.

Steve Wartik

Jennifer Vendetti wrote:
    
Hello Steven,

I'm using 3.2 beta, build 331.  I modified the following settings in
the Protege.lax file as follows:

lax.nl.java.option.java.heap.size.initial=100000000
lax.nl.java.option.java.heap.size.max=1000000000

It took a while - but I was able to successfully load
"wordnet.01.owl".  See console output for details:

Loaded 675000 triples
[TripleChangePostProcessor] Completed lists after 0 ms
[TripleChangePostProcessor] Completed anonymous classes after 16 ms
[TripleChangePostProcessor] Completed deprecated classes after 0 ms
[TripleChangePostProcessor] Completed properties after 47 ms
[TripleChangePostProcessor] Completed named classes after 2296 ms
... Loading completed after 1805883 ms

Jennifer

Steven Wartik wrote:
      
Anyone ever get Protege to load this ontology? I just tried, and
Protege ran out of memory even after I doubled the maximum stack
size, somewhere around the 250,000th triple. If you know the magic
number, please pass it on. Thanks.

Steve Wartik
        
-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html 
      

- --
- ----------------------------------------------------------------------
Dipl.-Ing.(FH) Andreas Langegger
Institute for Applied Knowledge Processing
Johannes Kepler University Linz
A-4040 Linz, Altenberger Straße 69
  
http://www.faw.at
http://www.langegger.at
    
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.2 (GNU/Linux)

iD8DBQFE4zC3Kk9SuaNc5+IRAgB6AJ4wSNSS2YUC9eRTBL3Zj1j6lohbRgCfW+Ti
Z+EpmKiM7jufyHbIII2ZWCY=
=vtHX
-----END PGP SIGNATURE-----
-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html