W3C home > Mailing lists > Public > public-owl-wg@w3.org > September 2008

Re: Review: RDF Mapping

From: Peter F. Patel-Schneider <pfps@research.bell-labs.com>
Date: Mon, 08 Sep 2008 20:37:36 -0400 (EDT)
Message-Id: <20080908.203736.26503337.pfps@research.bell-labs.com>
To: evren@clarkparsia.com
Cc: public-owl-wg@w3.org

From: "Evren Sirin" <evren@clarkparsia.com>
Subject: Review: RDF Mapping
Date: Mon, 8 Sep 2008 19:20:49 -0400

> 
> I've finished my review of the RDF mapping document and added my
> comments to the document (see the diff [1]). I think the document is
> in quite good shape and most of my comments are about improving
> readability or about minor details.
> 
> Regards,
> Evren
> 
> [1] http://www.w3.org/2007/OWL/wiki/index.php?title=Mapping_to_RDF_Graphs&diff=12443&oldid=12437

I've addressed the minor glitches in the tables, and a few other
comments in

http://www.w3.org/2007/OWL/wiki/index.php?title=Mapping_to_RDF_Graphs&diff=12449&oldid=12443

I also fixed a type (OPE instead of DPE) in Table 15.


I didn't make changes for two comments:

Review comment from EvrenSirin 23:14, 8 September 2008 (UTC) Table 1 can
be made much more compact by combining different rows that duplicate the
same transformation for object and datatype properties. For example,
there are two transformation rules involving PropertyDomain where the
resulting triple in both cases is identical. Instead we can have one
rule for PropertyRange(PE CE) that would generate that triple output. I
believe the reason for having duplicated entries is because earlier
functional syntax was typed and we had ObjectPropertyRange and
DataPropertyRange. Since this is not the case any more we can eliminate
multiple entries for many constructs and make this table shorter and
more readable. 
PeterPatel-Schneider Yes this would make Table 1 more
compact, but at the price of diverging it from the OWL 2 syntax. I don't
think that this would be a good trade-off.

Review comment from EvrenSirin 23:14, 8 September 2008 (UTC) It might be
better to move Table 13 right after Table 11 so all DataRange related
transformations are together. I understand that keeping all
backward-compatibility stuff together at the end is good structurally
but I think keeping the mapping rules for same constructs together makes
it easier to understand what is going on. 
PeterPatel-Schneider 00:24, 9 September 2008 (UTC) There are things to
say for this organization [the current one] as well (including inertia.)

peter
Received on Tuesday, 9 September 2008 00:38:20 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:42:06 UTC