How about comparing the non trivial inferences of different models.
Now there is a PhD topic if I ever saw one... Prunning, optimization
techniques, the possibilities are endless......

I'm sure there are some labs working on this.

N.



On 4/9/13, Uri Shani <[email protected]> wrote:
> Thanks,
> Indeed this is "just" a boolean, no description of the differences.
> My use case is that as one goes about changing a certain model, or an
> ontology, one may want to compare revisions to see clearly what has
> changed.
> So the model evolution is changing enough to make the new model different,
> yet how does it differ is of much value.
> - Uri
>
>
> From:   Andy Seaborne <[email protected]>
> To:     [email protected],
> Date:   09/04/2013 11:17 AM
> Subject:        Re: Comparing models
>
>
>
> On 09/04/13 08:50, Uri Shani wrote:
>> Hi,
>> Is there a solution for comparing two models? Specifically based on
> their
>> XML format?
>>
>> Here is a solution I made with some drawbacks:
>> I took the triples representation of both models, sorted them by
> subjects,
>> and made a similar attempt as diff does, although a bit more
>> RDF oriented than plain diff.
>> Some disruptions come from the fact (specifically when comparing
> ontology
>> OWL models) that there are sometimes triples with blank nodes
>> required to represent what appears "cleaner" in the xml format.
>> The difference of these triples in the two compared models seem to be
> due
>> more to cosmetics.
>> Perhaps a simple solution is to ignore blank nodes in the comparison.
>>
>> So again my Q at the top of this note.
>>
>> Thanks,
>> - Uri
>>
>
> Model.isIsomprphicWith(Model)
>
> Calculating the difference is somewhat harder.
>
>                  Andy
>
>
>

Reply via email to