RE: [SEM-GRD] Semantic Grid workshop at GGF9

["Pouchard, Line Catherine" <pouchardlc@ornl.gov>]

Title: Message

Hello:

 

I entirely agree with Mike about the application holder as a desireable audience (in fact, I raised the issue at the workshop).  Appendices like the one he suggested are very important because they provide a concrete example that readers can latch onto.   For ESG and earth sciences users, semantic grid concepts will mostly help in the search, discovery, and publishing into archives of very large datasets created by simulation.    Interoperability becomes necessary, when users want to perform searches across various data grids  where data is stored with different indexing schemas and ontologies (for instance, searching for a dataset containing certain parameters under ESG and a NASA archive).    In the sense that there is no user-defined service composition, and that services invoked are pre-determined by developers in ESG does not mean that there is no space for semantic grid tools and ontologies.  Tools for inserting metadata values with minimum user input are very needed.  Schema mappings based on ontological reconciliation (rather than ad hoc decisions by a developer) are another one.    These needs may fall in the scope of domain-specific ontologies, but in the long run, without domain users semantic grids may not get much momentum.    Did I misunderstand a comment from Carole at the workshop that some of these issues are more akin to the semantic web than semantic grids.  What differentiates from the semantic web in my opinion are the small user community, the size of the datasets and the computations, the absolute necessity of collaboration between people, and the requirements to share big instruments.

 

Line

 

-----Original Message-----
From: Michael Niemi [mailto:niemi@us.ibm.com]
Sent: Tuesday, October 07, 2003 1:56 PM
To: sem-grd@gridforum.org
Subject: Re: [SEM-GRD] Semantic Grid workshop at GGF9

Dave and Carole,

I also enjoyed the workshop and would like to help with the primer.

A couple thoughts:

        - I think it was decided in the workshop that the target audience of the primer is to be the grid infrastructure developer.  But it was also felt that the "application holder" was a desireable secondary audience.  How about including application-orientated perspectives in appendices?  E.g. an appendix for "A day in the life of the bioinformatician using the Semantic Grid".  That would describe the problem and benefits in the vocabulary of the bioinformatician, referring them then to common concepts, tools, etc in the main body of the primer.  
        It could also have little case study vignettes of lessons learned in that application domain - like "XYZ ontology appears to be doing a good job at addressing problem such-and-such.  The developers have found that curating the ontology in such-an-such a way is working particularly well.  This is probably due to factors blah-blah-blah ...".

        - It's no doubt a topic for a roadmap document (not the Primer), but how about a proposal (including screen mockups) of what people might feel is an ideal, no-holds-barred tool or suite of tools for developing, maintaining, and using ontologies and semantic markup?  As the tools&methodology section of the Primer is developed, people might want to also try to accumulate this sort of wishlist on the side.

Mike

----- Forwarded by Michael Niemi/Raleigh/IBM on 10/07/2003 12:07 PM -----

Carole Goble <carole@cs.man.ac.uk> Sent by: owner-sem-grd@gridforum.org 10/07/2003 11:43 AM

To:        "Pouchard, Line Catherine" <pouchardlc@ornl.gov>         cc:        David De Roure <dder@ecs.soton.ac.uk>, carole.goble@cs.man.ac.uk, "Dean N. Williams" <williams13@llnl.gov>, don@ucar.edu, "Bernholdt, David E." <bernholdtde@ornl.gov>, sem-grd@gridforum.org         Subject:        Re: [SEM-GRD] Semantic Grid workshop at GGF9

Line

>David and Carole:
>
>Thank you for the thought-provoking workshop on Sunday.
>
Thanks!

>I am interested
>to contribute to the GGF primer.
>
Hurrah!

>When you have the chance to post it,
>I'll look at the outline slide that Carole presented to have a better
>idea of where I fit.
>
Attached. It is also appearing on the web site

>
>Would it be  possible/suitable to add the Earth System Grid (ESG), as as
>a semantic grid project on the list?  I started exploring ontologies for
>it, and our discovery mechanisms use concepts loosely based on
>semantics.   We do not have ontology services but we are using Globus
>Metadata Catalog Service, Replica Locator Service, and are now OGSA- and
>OGSA DAI-compliant.
>
>
>ESG is at www.earthsystemgrid.org.
>
Dave will do this :-)

Carole

>Line
>
>
>
>>-----Original Message-----
>>From: David De Roure [mailto:dder@ecs.soton.ac.uk]
>>Sent: Sunday, October 05, 2003 10:49 AM
>>To: sem-grd@gridforum.org
>>Subject: [SEM-GRD] Semantic Grid workshop at GGF9
>>
>>
>>Hello,
>>
>>The documents for today's workshop are on
>>
>>        http://www.semanticgrid.org/GGF/ggf9/
>>
>>These will evolve as other material (such as the talks)
>>becomes available
>>throughout the day.  Once I have all the papers into one
>>proceedings document I'll make this available there too.
>>
>>Regards
>>
>>-- Dave
>>
>>
>>
>
>
>



**** Attachment discussion.ppt has been removed from this note on October 07, 2003 by Michael Niemi ****