[sc34wg3] Draft Reference Model

[Steven R. Newcomb]

"Graham Moore" <gdm@empolis.co.uk> writes:

> I have a few comments and offerings

> 1. 

> I agree with bernard, it is not clear in the RM how I
> model symmetrical relationships. Such as his sibling
> example.

Did my note about opposite.1 and opposite.2 make it
clearer?

> 2. 

> I think the role types are mandatory for 2 reasons,
> one general and one specific to tm models. The
> general one is that you may want to make some
> assertion between two things but not really know what
> that assertion is yet. A real world example would be
> to transforms someones folder structure into a
> tm. Folder structures are weak (non-existant) on what
> the relationship is, but the user probably has a good
> idea about it. Thus having empty role types and empty
> assoc type allows for the association to be refined
> later. A very useful property when dealing with
> iterative topicmap construction. (Personally, I like
> all my types defined and organised but appreciate
> that that isnt everyones world.)

Amen.

> The second reason for leaving types undefined is
> because that is the nature of scope. I think we came
> to some consensus in montreal that scope is no more
> than untyped associations between an association and
> a set of topics.

Well... that depends on what you mean by "untyped".

> 3. 

> This picks up on Bernard's point regarding Sets in
> RM. I appreciate that RM does not subscribe to any
> mathematical set model but i wonder about when do
> individuals become sets?

> I'm not sure I found it but if I have two assertions in different maps where:
> 
> graham worksfor empolis gmbh
> graham worksfor empolis uk
> 
> and I merge them what happens?
> 
> is it now true that 
> 
> graham worksfor [empolis gmbh, empolis uk]

No, not according to the draft RM.

> through merging, or are the two assertions seperate?

Yes, they remain distinct, because they do not have
exactly the same role players playing the same roles.
(I'm assuming that "empolis gmbh" and "empolis uk" are
two distinct subjects.)

Of course, the TM Application that defines the
"worksfor" assertion type is free to define situation
features, SIDP values and merging rules such that the
merger you are proposing would happen.  But it won't
happen under the global merging rule for assertions
that is defined in the draft RM4TM.  That rule only
eliminates totally redundant assertions.

> How does this differ from the example bernard gave
> regarding his set of children?

I'm at a loss.  How is it similar????

-- Steve

Steven R. Newcomb, Consultant
srn@coolheads.com

Coolheads Consulting
http://www.coolheads.com

voice: +1 972 359 8160
fax:   +1 972 359 0270

1527 Northaven Drive
Allen, Texas 75002-1648 USA