English explanation generator for SWRL and OWL inferences?

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

English explanation generator for SWRL and OWL inferences?

Michael DeBellis-2
I'm working on a paper that shows an approach to integrating Machine Learning and OWL/SWRL. One of the things I want to explain are the advantages of having things in an ontology rather than say a matrix in Python. Explicit knowledge representation seems to me to be one such advantage although for some people (who haven't already "drunk the kool-aid" of semantic technology) I think that may be hard to grasp.  The fact that the reasoners can explain their inferences is one obvious benefit but while I find the explanations generated by Protege to usually be easy to understand I'm afraid for someone not familiar with them they won't look all that intuitive. 

I was wondering if anyone has done any work on taking the explanations generated by the reasoner and generating more readable English text? When I was working in Expert Systems there were several commercial and research systems (from ISI and a small company called Cogentext) that did this really well. I looked through the list of plugins but didn't see anything. If anyone has any prototypes that work with the current version of Protege please let me know, I would like to use it, I think it would add a lot to what we are saying in the paper. thanks. 

Michael

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

Re: English explanation generator for SWRL and OWL inferences?

Csongor Nyulas
Administrator
I remember the ACE view plug-in [1] could do some cool verbalization in Protege 4.0, and one could even write axioms in a controlled English. However I don't think it was integrated with the explanation plug-in.
Also, unfortunately, I don't think The ACEView plugin was ever updated to work with Protege 5.x. Which is a pity. :( It would be a cool thing if someone would do it.
This is the more generic tool that the ACEView plugin relies on, if I am not mistaken:
https://github.com/Kaljurand/owl-verbalizer

Csongor

[1] https://github.com/Kaljurand/aceview/tree/v1.3.1


On 7/5/19 3:04 PM, Michael DeBellis wrote:
I'm working on a paper that shows an approach to integrating Machine Learning and OWL/SWRL. One of the things I want to explain are the advantages of having things in an ontology rather than say a matrix in Python. Explicit knowledge representation seems to me to be one such advantage although for some people (who haven't already "drunk the kool-aid" of semantic technology) I think that may be hard to grasp.  The fact that the reasoners can explain their inferences is one obvious benefit but while I find the explanations generated by Protege to usually be easy to understand I'm afraid for someone not familiar with them they won't look all that intuitive. 

I was wondering if anyone has done any work on taking the explanations generated by the reasoner and generating more readable English text? When I was working in Expert Systems there were several commercial and research systems (from ISI and a small company called Cogentext) that did this really well. I looked through the list of plugins but didn't see anything. If anyone has any prototypes that work with the current version of Protege please let me know, I would like to use it, I think it would add a lot to what we are saying in the paper. thanks. 

Michael

_______________________________________________
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