Re: [BioRDF] BioSIOC / aTag task

Hi Matthias,

My concern is not redundancy of content, or of technical  
implementation, but of potential semantic redundancy and/or mismatch  
between two formulations of the same basic idea.

This problem could arise precisely because the form in which the  
content is expressed - i.e. scientific assertions - would have  
different technical implementations, despite being semantically  
identical at a fundamental level.   Therefore I think your proposed  
work should it be carefully coordinated with SWAN-SIOC both before and  
during development to ensure alignment - if it is to be part of an  
official HCLS task.

When you initially proposed your idea on the Scientific Discourse  
call, I felt confident this coordination could occur, if and when the  
work was started, because it was being done in the same task group as  
the other Discourse tasks.  Now that you have proposed it again in  
BioRDF,  I am not confident this coordination will occur spontaneously  
unless we make it happen - therefore I suggest you and Kei and I spend  
some time exploring the ramifications of starting this task in another  
group and how to achieve alignment.  Perhaps Susie could lend a hand  
in this discussion as well.

If it later turns out I am wrong and it turns out, after discussion,  
that there is no need for any pre-alignment or coordination of your  
work with the Discourse tasks, we will still have had the chance to  
understand your ideas better, and have shared our thinking, which is  
all good.

Best

Tim

On Jan 7, 2009, at 4:33 AM, Matthias Samwald wrote:

>
> Dear Tim,
>
> The project I have in mind is not redundant with SWAN-SIOC, both in  
> its technical implementation and the biomedical content that will be  
> represented. The statements will not have discourse relationships  
> among themselves, and indeed such relationships would be better  
> represented through the vocabulary that SWAN can add to the basic  
> SIOC vocabulary.
> This project could also demonstrate the value of the alignment of  
> SWAN and SIOC, showing that information represented in basic SIOC  
> can be easily aligned with information represented in the more  
> expressive SWAN vocabulary, via the SWAN-SIOC alignment. This is  
> important for demonstrating the advantages gained (in terms of  
> interoperability) by the SWAN-SIOC alignment.
>
> Cheers,
> Matthias Samwald
>
>

Tim Clark
Director of Informatics, MassGeneral Institute for Neurodegenerative  
Disease
Instructor in Neurology, Harvard Medical School
617-947-7098 (mobile)

Received on Wednesday, 7 January 2009 11:15:06 UTC