W3C home > Mailing lists > Public > public-owl-wg@w3.org > December 2007

Re: [OWLWG-COMMENT] Example why current RDF mapping for QCRs might hurt OWL-1.1-Full

From: Jim Hendler <hendler@cs.rpi.edu>
Date: Mon, 17 Dec 2007 19:55:48 -0500
Message-Id: <AD402BB2-331B-4E69-8DEE-E1DBBD0CD133@cs.rpi.edu>
Cc: OWL Working Group WG <public-owl-wg@w3.org>
To: Bijan Parsia <bparsia@cs.man.ac.uk>
Bijan -
   I've discussed the users I feel that I am representing as often as  
you have, but here's another attempt: I feel I represent the Web 3.0  
innovators whose boards I sit on, I personally interact often with  
the people recognizing the emerging power of the Web of Data and the  
need for "A little semantics" (a term you all have heard from me  
often) to help manage it, and I represent personally an interest, and  
I believe a growing set of like-minded academics,  in thinking about  
the "database like" aspects of the Semantic Web, with concern for  
scaling, approximate reasoning, and ontologies viewed as part of a  
much larger eco-system of Semantic techniques and technologies [0],  
rather than solely as standalone components for single domain reasoners.
  This should come as no surprise, I wrote about this in my "Dark  
side of the Semantic Web" blogs and editorials [1]. Ora and I wrote  
about this in our "Embracing Web 3.0" article in Internet Computing  
[2], I have a discussion of it in a forthcoming IEEE Computer column,  
and it's been a mainstay of my research since, since, ... well really  
for my whole career, but certainly since I went to DARPA to try to  
help bring it into reality..
  Hope that helps explain to everyone where I'm coming from - you may  
not agree, but I think I've been consistent about this over the  
years, even back when we were first forming the joint committee that  
led to DAML+OIL
  JH

[0] http://members.deri.at/~ruzicap/ISWC2007-WS
[1] http://csdl.computer.org/comp/mags/ex/2007/01/x1002.pdf (free  
preprint available at http://www.mindswap.org/blog/2006/12/14/tales- 
from-the-dark-side-continued/)
[2] http://doi.ieeecomputersociety.org/10.1109/MIC.2007.52 (free  
preprint at http://www.mindswap.org/papers/2007/90-93.pdf)


On Dec 17, 2007, at 6:40 PM, Bijan Parsia wrote:

>
> On Dec 17, 2007, at 2:38 PM, Jeremy Carroll wrote:
>
>> Jim Hendler wrote:
>>> <flame on - but not at Matthew>
>>
>> without wishing to fan Jim's flames, ...
>>
>> one aspect I noted at the F2F was that there is this decidability  
>> litmus test there is some wiggle room, and the actual drivers for  
>> what was a compelling argument and what wasn't had to do with the  
>> use cases and customers who we each had in mind.
>
> This is just about always the case. *All* OWL 1.1 features were  
> driven by user needs and implementor realizability. This is hardly  
> an strange design stance at the W3C!
>
> However, you have made a very specific decidability argument, to  
> wit, that under just about any extension (e.g., of nary data  
> predicates), if I combine two files that meet all the decidability  
> restrictions then their union should also be decidable.
>
> *This* notion of a "decidability requirement" has not been taken as  
> a hard requirement if only because there's no evidence that it was  
> *ever* a requirement: it's not true of OWL DL (simple roles,  
> datatypes). Thus, it doesn't make us *worse off*.
>
> Data values are an explicit extensibility point of OWL.
>
> [snip]
>> So, as HP rep, I found the Oracle presentation compelling, much  
>> more compelling than most, because the Oracle customers and the HP  
>> customers are similar and doing similar things. However, none of  
>> the presentations were explicit in terms of customers,
>
> This isn't true, or perhaps at best narrowly true (I've not  
> reviewed all the presentations per se). I've several times  
> mentioned specific users for Rich Annotations and N-ary datatypes.  
> We brought into the meeting users of both. (And not just "academic"  
> users..both Alan and esp. Sebastian are working in collaboration  
> with Siemens on a commercial project.)
>
>> and we have made precious little advance on a use case and  
>> requirements document,
>
> UC&R documents at the W3C very rarely track back to specific users,  
> so I don't see this as an advance.
>
>> so that the hidden differences between us (the various members of  
>> this WG) in terms of what we are trying to do with OWL, for whom,  
>> and why, remain hidden.
> [snip]
>
> Nothing stops you (or Jim) from talking as explicitly about your  
> users as I do. Indeed, I encourage you to do so and have done so  
> frequently.
>
> Cheers,
> Bijan.
>

"If we knew what we were doing, it wouldn't be called research, would  
it?." - Albert Einstein

Prof James Hendler				http://www.cs.rpi.edu/~hendler
Tetherless World Constellation Chair
Computer Science Dept
Rensselaer Polytechnic Institute, Troy NY 12180
Received on Tuesday, 18 December 2007 00:56:33 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:13:29 GMT