Hi Gong,
Gong Cheng wrote:
Hi Mike and All,
Very interesting work! Class-level mappings are quite important to
domain-independent data integration, e.g., search engines like Falcons.
Yes, we agree. Falcons is consistently showing the right way to
search and present this stuff. I hope you see some new ideas
about class mappings moving forward! :)
And I still have some questions to be clarified. I've downloaded
"class_level_lod_constellation.csv", which gives relations between
ontologies rather than between classes. Are there any class-level mappings
available? And, how do you obtain such mappings? Is it just based on
explicitly stated axioms in ontologies, or based on any ontology matching
algorithms?
Good questions and points. I'm sure Fred will comment back as
well in the morning and I hope others do as well on this ML.
You are correct that we do not have a comprehensive inventory of
specific class mappings. However, we *do* for those that involve
UMBEL:
http://www.umbel.org/ontology/umbel_external_ontologies_linkage.n3
As for the others shown on the LOD constellation diagram, we have
ascertained there *are* class-level mappings but have not yet
compiled the specific class-level assertions those non-UMBEL
sources make. That enumeration should be done.
Class-level mappings, because of their *generalizability* and
*power*, need to be carefully done. We have looked at automatic
and semi-automatic methods for these mappings (an important
personal and business focus), but in the end such class-level
assertions are pivotal to quality and believability. This
current LOD constellation diagram is simply reporting the
linkages with no comment on provenance or accuracy.
My own view is that, while some alignment or mapping algorithms
might be helpful for screening or identification purposes,
actionable mappings need to be QA/QC'ed by humans with provenance
made clear. Axioms and algorithms can help to identify and rank
candidates, but should not be determinative.
However, saying that, our own UMBEL mappings have not taken that
long and once made are easily reused with leverage. IMO, it is a
much less onerous task to do class-level mappings than instance
ones. (While a single class assignment may require more time and
thought than an instance one, there are fortunately many fewer
structural -- that is, class -- than instance relationships
between most datasets.)
Of further caution is that once one moves beyond specific
individual sameAs assertions to those involving entire classes
the consequences of getting it wrong are greater.
But what we are seeing is that a few class connections can bring
some really powerful inferencing. Just getting a few right can
be really powerful. . . .
Regards, Mike
Regards,
Gong
-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Mike Bergman
Sent: Monday, October 06, 2008 10:13 AM
To: [email protected]
Subject: LOD Constellation
Hi LODsters,
We are pleased to provide a LOD 'constellation' diagram [1] of
class-level mappings within open linked datasets. It is meant to
accompany the LOD cloud diagram [2] that our community has been
using for more than a year to great effect to educate the public
about linked data.
Fred Giasson did all of the initial analysis behind the creation
of this LOD 'constellation' diagram. The input files behind this
diagram are also available for download [1].
I have also provided a bit more background information and
description of the importance of class-level mappings in a
related blog post [3].
As that post tries to stress, we now have all of the essential
infrastructure pieces for effectively promoting the linked data
vision.
Pretty cool! and thanks to all of you for publishing the datasets
and bringing linked data to this current inflection point.
Regards, Mike
[1] http://www/umbel.org/lod_constellation.html
[2]
http://www4.wiwiss.fu-berlin.de/bizer/pub/lod-datasets_2008-09-18.html
[3] http://www.mkbergman.com/?p=457
