[topicmapmail] Temporality in Topic Maps

[Graham Moore]

>> So basically using occurrences as binary associations for events.

Not really. We are simply using the fact that topics can have occurrence
data and we are choosing the put the start and end time of the event in
those occurrences.

The main difference is that rather than the association being the nexus of
the event, we create a topic to be the event and because of this we can
easily assigned occurrence data. An alternative would be to reify the
association and add the event start and end things onto that. 

If you have a model where a topic is a time period it gets very messy when
you have lots of high granularity events. Its like having '1' as a topic.
Shudder. But for lower granularity maybe a topic would work just fine.

The other thing to point out is how this information will be used. I'm
thinking that you'll want to say give me all people who worked for Company X
while some other person also worked there. Having start end data as a 'real'
date means query operations will work on it.

With this approach there is no abuse of TMDM but as with all modelling
choices, it is a choice, takes you in a certain direction and has
consequences.

Cheers, 

Graham

--
Graham Moore, Founder, NetworkedPlanet
Editor XTM 1.0, ISO13250 (TopicMaps) -1,-2, TMCL
e: graham.moore at networkedplanet.com
w: www.networkedplanet.com
m: +47 45271713
p: +44 (0)1865 811131
 
-----Original Message-----
From: topicmapmail-bounces at infoloom.com
[mailto:topicmapmail-bounces at infoloom.com] On Behalf Of Alexander Johannesen
Sent: 04 June 2006 13:42
To: Graham Moore
Cc: topicmapmail at infoloom.com
Subject: Re: [topicmapmail] Temporality in Topic Maps

Hi Graham,

> I would also consider doing the following:
>
> topic "some person"
> topic "some organisation"
> topic "some event" : "Person Employment Event"
>         - occurrence "start date/time"
>         - occurrence "end date/time"

So basically using occurrences as binary associations for events.
Yeah, the pragmatist in me thinks this looks fine. Doesn't this scrape
against the core of the TM model, though? :) (Not to say that occurrence
itself isn't a short-hand for a binary association) I wish there was some
better semantic notion for these things. Has there been any talk about
binary assocations being embedable in a topic? (which again I guess goes
against the TM ideal)


Alex
--
"Ultimately, all things are known because you want to believe you know."
                                                         - Frank Herbert __
http://shelter.nu/ __________________________________________________
_______________________________________________
topicmapmail mailing list
topicmapmail at infoloom.com
http://www.infoloom.com/mailman/listinfo/topicmapmail