Using Protege with web content

[j***@dmc.dit.ie]

Hello all,

I am beginning a research project which will use descriptive power of
ontologies to drive more intelligent web content (suprisingly :-)). I was
hoping to use protege, which is an excellant ontology editor, only I wish
to store the ontology in a data-store of some sort using protege to edit
the ontology and an associated rules engine to alter the rules. Then i
could make use of protege API to interact with the web content, probably a
J2ee app. However, when using protege before it required a file to be
loaded to make any amendments and when stored in a database the file was
not optimised simply stored as a table, although i understand the
reasoning behind this i am inquiring to see has there been any headway in
storing the ontology in a data-store and providing access through an API
such as JENA?

I hope I made sense and would be greatfull for any help.

Thanks in advance

Rgds

j

[Holger Knublauch]

If you are interested in OWL/RDF or Jena, you should
ask your question on the protege-owl list. There is
a plugin for Protege-OWL that generates Jena DB from
the OWL project.

http://semweb.krasu.ru/

Also, you can easily extract a Jena OntModel from a
Protege OWLModel and then do the conversion with a few
lines of code.

Holger

-----Original Message-----
Sent: Wednesday, February 16, 2005 12:57 AM
Subject: [protege-discussion] Using Protege with web content
Hello all,
I am beginning a research project which will use descriptive
power of ontologies to drive more intelligent web content
(suprisingly :-)). I was hoping to use protege, which is an
excellant ontology editor, only I wish to store the ontology
in a data-store of some sort using protege to edit the
ontology and an associated rules engine to alter the rules.
Then i could make use of protege API to interact with the web
content, probably a J2ee app. However, when using protege
before it required a file to be loaded to make any amendments
and when stored in a database the file was not optimised
simply stored as a table, although i understand the reasoning
behind this i am inquiring to see has there been any headway
in storing the ontology in a data-store and providing access
through an API such as JENA?
I hope I made sense and would be greatfull for any help.
Thanks in advance
Rgds
j
--------------------------------------------------------------
----------------
To remove yourself from this list go to
http://protege.stanford.edu/lists.html

[j***@dmc.dit.ie]

Thanks Holger,

Sorry for cross posting,
Thanks for your help.

John

Post by Holger Knublauch
If you are interested in OWL/RDF or Jena, you should
ask your question on the protege-owl list. There is
a plugin for Protege-OWL that generates Jena DB from
the OWL project.
http://semweb.krasu.ru/
Also, you can easily extract a Jena OntModel from a
Protege OWLModel and then do the conversion with a few
lines of code.
Holger

-----Original Message-----
Sent: Wednesday, February 16, 2005 12:57 AM
Subject: [protege-discussion] Using Protege with web content
Hello all,
I am beginning a research project which will use descriptive
power of ontologies to drive more intelligent web content
(suprisingly :-)). I was hoping to use protege, which is an
excellant ontology editor, only I wish to store the ontology
in a data-store of some sort using protege to edit the
ontology and an associated rules engine to alter the rules.
Then i could make use of protege API to interact with the web
content, probably a J2ee app. However, when using protege
before it required a file to be loaded to make any amendments
and when stored in a database the file was not optimised
simply stored as a table, although i understand the reasoning
behind this i am inquiring to see has there been any headway
in storing the ontology in a data-store and providing access
through an API such as JENA?
I hope I made sense and would be greatfull for any help.
Thanks in advance
Rgds
j
--------------------------------------------------------------
----------------
To remove yourself from this list go to
http://protege.stanford.edu/lists.html

------------------------------------------------------------------------------
To remove yourself from this list go to
http://protege.stanford.edu/lists.html

[Ray Fergerson]

Post by j***@dmc.dit.ie
Hello all,
I am beginning a research project which will use descriptive power of
ontologies to drive more intelligent web content (suprisingly :-)). I was
hoping to use protege, which is an excellant ontology editor, only I wish
to store the ontology in a data-store of some sort using protege to edit
the ontology and an associated rules engine to alter the rules. Then i
could make use of protege API to interact with the web content, probably a
J2ee app. However, when using protege before it required a file to be
loaded to make any amendments and when stored in a database the file was
not optimised simply stored as a table,

Have you tried access through the Protege API to this information in a
database and found it to be too slow for your application? Many people
access information though the Protege API in web applications and find
the performance reasonable. Many others just assume it won't work
without trying it...

Ray

[YI Ping Wong]

I am new to protégé, I am trying to load the owl examples (http://protege.stanford.edu/plugins/owl/owl-library/) I have tried to load, import, and build but so far I have not been able to do so. How can I load one of those files in to protégé, step by step. Thank you all.



---------------------------------
Do you Yahoo!?
Yahoo! Search presents - Jib Jab's 'Second Term'

[Holger Knublauch]

Could you please post this question yet another time,
this time to the protege-owl list? protege-discussion
is about everything non-OWLish.

Thanks
Holger

-----Original Message-----
Of YI Ping Wong
Sent: Wednesday, February 16, 2005 10:34 AM
Subject: [protege-discussion] load owl
I am new to protégé, I am trying to load the owl examples
(http://protege.stanford.edu/plugins/owl/owl-library/) I have
tried to load, import, and build but so far I have not been
able to do so. How can I load one of those files in to
protégé, step by step. Thank you all.
---------------------------------
Do you Yahoo!?
Yahoo! Search presents - Jib Jab's 'Second Term'
--------------------------------------------------------------
----------------
To remove yourself from this list go to
http://protege.stanford.edu/lists.html

[j***@dmc.dit.ie]

I have tried before but protege requires the complete dataset to be loaded
into the editor before you can use the API. This is my problem, I wish to
use Protege to build and maintain the ontology and then use it's API to
access the data without loading the complete dataset into the application
server. I forsee problems here, so would prefer access the data with an
api (that provides reasoning) that uses a transactional datastore. An all
in one colution would be great but i am not sure is that possible.
Protege uses the JENA API to allow for OWL, however i am not sure does
protege load the complete dataset in on startup?

Thanks for your help

Regards
john

Post by Ray Fergerson

Post by j***@dmc.dit.ie
Hello all,
I am beginning a research project which will use descriptive power of
ontologies to drive more intelligent web content (suprisingly :-)). I was
hoping to use protege, which is an excellant ontology editor, only I wish
to store the ontology in a data-store of some sort using protege to edit
the ontology and an associated rules engine to alter the rules. Then i
could make use of protege API to interact with the web content, probably a
J2ee app. However, when using protege before it required a file to be
loaded to make any amendments and when stored in a database the file was
not optimised simply stored as a table,

Have you tried access through the Protege API to this information in a
database and found it to be too slow for your application? Many people
access information though the Protege API in web applications and find
the performance reasonable. Many others just assume it won't work
without trying it...
Ray
------------------------------------------------------------------------------
To remove yourself from this list go to
http://protege.stanford.edu/lists.html

[Holger Knublauch]

Once Protege has created a database, only a fraction of the whole
dataset is loaded into memory at startup. Memory consumption is
reduced to a minimum. However, in order to first create the database,
it is in most cases necessary to first load a (OWL) file into memory.
If this is your bottleneck, then you may need to generate the database
contents using a script, or hack the JenaLoader to operate on a
Jena database incrementally.

Holger

-----Original Message-----
Sent: Wednesday, February 16, 2005 8:20 PM
Subject: [protege-discussion] Re: Using Protege with web content
I have tried before but protege requires the complete dataset
to be loaded into the editor before you can use the API.
This is my problem, I wish to use Protege to build and
maintain the ontology and then use it's API to access the
data without loading the complete dataset into the
application server. I forsee problems here, so would prefer
access the data with an api (that provides reasoning) that
uses a transactional datastore. An all in one colution would
be great but i am not sure is that possible.
Protege uses the JENA API to allow for OWL, however i am not
sure does protege load the complete dataset in on startup?
Thanks for your help
Regards
john

Post by Ray Fergerson

Post by j***@dmc.dit.ie
Hello all,
I am beginning a research project which will use

descriptive power of

Post by Ray Fergerson

Post by j***@dmc.dit.ie
ontologies to drive more intelligent web content

(suprisingly :-)).

Post by Ray Fergerson

Post by j***@dmc.dit.ie
I was hoping to use protege, which is an excellant

ontology editor,

Post by Ray Fergerson

Post by j***@dmc.dit.ie
only I wish to store the ontology in a data-store of some

sort using

Post by Ray Fergerson

Post by j***@dmc.dit.ie
protege to edit the ontology and an associated rules

engine to alter

Post by Ray Fergerson

Post by j***@dmc.dit.ie
the rules. Then i could make use of protege API to

interact with the

Post by Ray Fergerson

Post by j***@dmc.dit.ie
web content, probably a J2ee app. However, when using

protege before

Post by Ray Fergerson

Post by j***@dmc.dit.ie
it required a file to be loaded to make any amendments and when
stored in a database the file was not optimised simply stored as a
table,

Have you tried access through the Protege API to this

information in a

Post by Ray Fergerson
database and found it to be too slow for your application? Many
people access information though the Protege API in web

applications

Post by Ray Fergerson
and find the performance reasonable. Many others just

assume it won't

Post by Ray Fergerson
work without trying it...
Ray

----------------------------------------------------------------------

Post by Ray Fergerson
-------- To remove yourself from this list go to
http://protege.stanford.edu/lists.html

--------------------------------------------------------------
----------------
To remove yourself from this list go to
http://protege.stanford.edu/lists.html

[Ray Fergerson]

Post by j***@dmc.dit.ie
I have tried before but protege requires the complete dataset to be loaded
into the editor before you can use the API.

This is not the case. Information is loaded from the database as it is
needed. When it is no longer needed it is flushed from memory.

Post by j***@dmc.dit.ie
This is my problem, I wish to
use Protege to build and maintain the ontology and then use it's API to
access the data without loading the complete dataset into the application
server. I forsee problems here, so would prefer access the data with an
api (that provides reasoning) that uses a transactional datastore.

The Protege API does not itself provide reasoning but reasoning systems
have been layered on top of this API (JessTab, Algernon, etc). These
will benefit from the "load on demand" functionality of the API. Even
if you write a rule in such a system that requires traversal of the
entire kb this does not mean that the entire kb needs to be in memory at
any one time.

Ray

[j***@dmc.dit.ie]

Thanks Ray,

Thats all i needed to clarify.

Regards
John

Post by Ray Fergerson

Post by j***@dmc.dit.ie
I have tried before but protege requires the complete dataset to be loaded
into the editor before you can use the API.

This is not the case. Information is loaded from the database as it is
needed. When it is no longer needed it is flushed from memory.

Post by j***@dmc.dit.ie
This is my problem, I wish to
use Protege to build and maintain the ontology and then use it's API to
access the data without loading the complete dataset into the
application
server. I forsee problems here, so would prefer access the data with an
api (that provides reasoning) that uses a transactional datastore.

The Protege API does not itself provide reasoning but reasoning systems
have been layered on top of this API (JessTab, Algernon, etc). These
will benefit from the "load on demand" functionality of the API. Even
if you write a rule in such a system that requires traversal of the
entire kb this does not mean that the entire kb needs to be in memory at
any one time.
Ray
------------------------------------------------------------------------------
To remove yourself from this list go to
http://protege.stanford.edu/lists.html