Re: Wiki-page on EL++ and ISSUE-79

On 12 Dec 2007, at 17:25, Alan Ruttenberg wrote:

> Thanks Carsten!
> Regarding integrating into Bernardo's document - as long as it is  
> Bernardo's version OK. For the version the WG is working on we  
> would need to discuss it at a meeting. If that's what's proposed,  
> could you please submit an issue?

The existing Fragments document provides an overview of various  
interesting fragments and pointers into the literature. I don't think  
that it would make sense to add a much more detailed description of  
any one fragment, at least not until we make strategic decisions on  
fragments in general. It may make sense for Bernardo to make  
editorial fixes to the existing document in response to the points  
made by Carsten below.


> Thanks,
> Alan
> On Dec 12, 2007, at 9:07 AM, Carsten Lutz wrote:
>> Dear all,
>> as proposed by Jim, I have set up a Wiki page with details on the  
>> EL++
>> tractable fragment of OWL 1.1 DL. You can find it at
>> The version of EL++ given there is, in my opinion, the natural  
>> version
>> of EL++ that comes up as a fragment of OWL 1.1 DL. It differs from  
>> the
>> one given in Bernardo's "tractable fragments" document in a number of
>> (mostly minor) ways. Notably,
>> - some additional expressive power is added, such as reflexive  
>> properties,
>>   facts, and datatypes
>> - a nonstructural restriction is adopted to achieve tractability (the
>>   fragment reported in Bernardo's document doesn't have it, and turns
>>   out to be at least PSpace-hard).
>>> From my perspective, this settles ISSUE-79 (raised by myself),  
>>> which asks
>> for the appropriate version of EL++ that we should consider within  
>> this
>> group.
>> As an aside: note that my wiki page is quite long. If EL++ should
>> become part of the standard, I feel that the document should be even
>> more detailed. This indicates that standardizing a larger number of
>> tractable fragments will be quite a bit of work, and produce quite
>> long documents.
>> greetings,
>> 		Carsten
>> PS: Bernardo, if you want to integrate all or some of this in your
>>     document, please go ahead.
>> --
>> *      Carsten Lutz, Institut f"ur Theoretische Informatik, TU  
>> Dresden       *
>> *     Office phone:++49 351 46339171   mailto:lutz@tcs.inf.tu- 
>>     *

Received on Wednesday, 12 December 2007 17:45:04 UTC