Bayesian logic implementations of OWL

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

Bayesian logic implementations of OWL

Trina Myers
Hi All,

Would anyone know of any further work implemented in Protege using different
forms of logics. I would like to use some form of probabilistic logic (Bayesian,
causal, etc) to run sensitivity analyses, as opposed to classifying and
inferring knowledge given a set of classes/instances as is possible with DL and
Horn Logic.

I have briefly looked into PR-OWL and KEEPER, but there doesn't seem to be any
implementation further then the initial papers.

Or is it even possible??

Cheers
Trina

_______________________________________________
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: Bayesian logic implementations of OWL

adam.saltiel
I am very interested in the answer to this question.
I have done some research in the area, but don't have a complete answer in terms of what has been implemented.
One direction I have looked in is in the use of logics on top of OWL. There is no one favoured way of using logics and there is a lot of work in Prolog that is of relevance. What I have found is that there have been advances in the understanding of how to partition problems logically so that the relationship between subsections is controlled by that logic. The subsections themselves may use various techniques to calculate results including Hidden Markov Chains (Models = HMM) and Bayesian.
The pointers I have for this are e.g. David Poole
http://www.cs.ubc.ca/labs/lci/CIspace/index.html
XObjects from a research team at SEU China
http://ebi.seu.edu.cn/
PROFILE (freiburg university)
http://www.informatik.uni-freiburg.de/~kersting/profile/
Only the implementation from CLSpace is available (with some searching) in source code form and I am not sure of the licence. Anyway you need to interface with prolog, possibly a particular flavour, to see results.
I would be interested in what you might come up with and if there is anything actually open source.
Adam

On 29/03/07, Trina <[hidden email]> wrote:
Hi All,

Would anyone know of any further work implemented in Protege using different
forms of logics. I would like to use some form of probabilistic logic (Bayesian,
causal, etc) to run sensitivity analyses, as opposed to classifying and
inferring knowledge given a set of classes/instances as is possible with DL and
Horn Logic.

I have briefly looked into PR-OWL and KEEPER, but there doesn't seem to be any
implementation further then the initial papers.

Or is it even possible??

Cheers
Trina

_______________________________________________
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: Bayesian logic implementations of OWL

Doug Holmes
In reply to this post by Trina Myers
Hi Trina,
        About a week ago, there was a nice discussion of this topic on the  
Ontolog forum; Protégé was one of the featured editors.  PR-OWl was  
discussed in some detail. The notes [and slides] from the  
presentations are at: http://ontolog.cim3.net/cgi-bin/wiki.pl?
ConferenceCall_2007_03_29
Doug

On Mar 29, 2007, at 2:58 PM, Trina wrote:

> Hi All,
>
> Would anyone know of any further work implemented in Protege using  
> different
> forms of logics. I would like to use some form of probabilistic  
> logic (Bayesian,
> causal, etc) to run sensitivity analyses, as opposed to classifying  
> and
> inferring knowledge given a set of classes/instances as is possible  
> with DL and
> Horn Logic.
>
> I have briefly looked into PR-OWL and KEEPER, but there doesn't  
> seem to be any
> implementation further then the initial papers.
>
> Or is it even possible??
>
> Cheers
> Trina
>
> _______________________________________________
> 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