Re: [Tracker] Proposal for contact ontology (NCO)

[Martyn Russell <martyn lanedo com>]

On 16/02/10 22:59, THEVENET Florent wrote:
> Hello,

Hello,

> I'm very interested in NCO ontology and I saw the modifications you made
> from the original Nepomuk ontology.
> I must say that of not very fond of some of these changes. First because
> you seems to assume that only IM clients and mail apps will use NCO.
> Then because with the IMContact class you lost semantic meaning. Finally
> because I don't understand the need for an MetaContact thing, while
> Nepomuk's ontology has a powerful ContactMedium class.
>
> Let me explain my thoughts and expose my proposal.
>
> In my mind IMContact is evil; it's both a Contact and a ContactMedium so
> you lose a lot of meaning here. What's more as far as I understand, the
> Real Person should be represented by MetaContact; and then you link that
> MetaContact with IMContacts, which can have different names, genders...
> Does a real person have many names ? I don't think so.

I think it is possible to have multiple nick names and there are names 
which YOU have for yourself and which other people may want to have for 
you. So if you consider that in a protocol like XMPP, the nickname you 
have on the server (for yourself) can be different to the nickname that 
is used on a friend's contact's list for you. I am not sure how much 
sense it makes to differentiate here, but the ontology should be rich 
enough to describe it. I don't think it is likely you would have 
different genders though ;)

> Then you care only about IM things : where are all the hypes things like
> SocialNetwork and MicroBlogging from where you can gather a LOT of
> semantic informations about a contact ?

Not sure exactly what you're asking here?

[snip]

I will let Juergbi comment on the other changes you had in mind since he 
was more centrally involved in the ontology updates in this area.

--
Regards,
Martyn