Re: ISSUE-108: Names for Profiles

On 1 May 2008, at 08:39, Ivan Herman wrote:
> Bijan Parsia wrote:
>> OWL EDL (for EL++ DL)
>> OWL RDL (for relational DL)
>> OWL DLP (for description logic programs)
>
> Actually, these are not bad at all...
>
> Can someone solve a mystery for me? Where does the 'EL++'  
> terminology come from? I can relate, technically, both to RDL and  
> DLP, but not to that EDL because of that (I am not knowledgable of  
> Description Logic...)

Carsten could say more, I imagine, but my understanding is that "EL"  
is modeled on "FL" (early and very weak, but also not very robust)  
and "AL" (the core of ALC, i.e., Attribute Logic with Complement) and  
the calligraphic E which indicates Existentials. (e.g., you can have  
the logic ALE as well) Early DLs usually included universal  
quantification as a minimal feature. The EL family turns that on its  
head and takes existentials as key. Perhaps surprisingly (but very  
cooly) it turns out that EL can be much more robustly extended than  
AL/FL like langauges while remain tractible. The ++ indicates that  
there are additional features over basic EL.

See:
	http://dli.iiit.ac.in/ijcai/IJCAI-2003/PDF/048.pdf

Cheers,
Bijan.

Received on Thursday, 1 May 2008 07:57:21 UTC