Re: Thoughts in the prov primer

Hi Stephan,

Thanks, these are all good suggestions.

For point 4, I'm not clear what you are intending the section to
contain - aren't domain-specific attributes and qualifiers introduced
already as part of the worked example (as described in your points 1
to 3)? What further would you like to be said?

I'll send a follow-up email with a more general plan for the primer.

Thanks,
Simon

On 3 November 2011 03:24, Stephan Zednik <zednis@rpi.edu> wrote:
> Here are my thoughts on how we could change the primer to reflect the changes to prov-o discussed on Monday's prov-o telecon.
>
> For those not on Monday's telecon, we discussed replacing EntityInRole with Usage, Generation, Control, and Participation(?) classes.  Qualifiers can now be applied to individuals of these classes, which matches nicely with the description in the prov-DM spec dice these classes represent the relations we want to qualify.
>
> 1) It seems logical to introduce the Generation and Usage classes in the 'Use and Generation' working example section.  I think it would make sense to introduce at least one qualifier here as well, as the qualifiers justify the need for the modeling of these prov-DM relations as OWL classes.  Perhaps temporal qualifiers?
>
> 2) Control and Participant(?) classes could be introduced in the 'Agent' working example section.
>
> 3) the Roles section is very straightforward if Usage and Generation have already been presented.  It is now just introduction of Role class, explanation of roles as an extension point, and adding role qualifiers to already presented Usage and Generation encoding.
>
> 4) Do we have a general sections for attributes, qualifiers, and annotations?  These are described as important extensibility points, so I think we should provide examples of extending them for the worked example domain.
>
> --Stephan
>



-- 
Dr Simon Miles
Lecturer, Department of Informatics
Kings College London, WC2R 2LS, UK
+44 (0)20 7848 1166

Received on Monday, 7 November 2011 13:02:16 UTC