Re: A long but hopefully interesting introduction

On Mar 5, 2005, at 5:55 AM, Geoff Chappell wrote:

> Hi Ben,
> Sounds like a cool project.

Thanks!

[...]
> 	http://labs.intellidimension.com/ontag/

Very interesting...

> I think the wiki model will work if you expect that the community of 
> users
> is like minded enough so that the ontology will converge.

Ideally, yes. The "ideal" likn community is one that shares a 
specialized vocabulary and set of topics (ie, just about any group of 
people involved in a common enterprise). I expect there to be some 
debate over definitions and classes within any community, but that's a 
Good Thing. Examples of communities that would be prime candidates for 
likn: a graduate department at a university, an association of medical 
specialists, an open-source software project, a fan-site, etc.

>  I think a
> networked model, where nodes are connected to similar nodes and where
> information propagates according to a support threshold, would allow
> competing and divergent ontologies to develop and coexist. Sound like 
> you're
> talking about a bit of both?

Yes, I guess I'm trying to have my cake and eat it too; I like the idea 
of these ontologies being networked globally and in conversation with 
one another, but on the other hand, a likn colony is essentially a 
sheltered sub-network that knows a lot about itself but little about 
the outside world.

[...]
> It seems likely your ontology will converge to a semi-stable point 
> which
> will satisfy some. You could always publish stable snapshots for the 
> rest.

That's a possibility... I could always make the ontology publishing 
manual rather than periodic, so administrators could do it when they 
felt it was right.

> I doubt you could pull off what you describe and stay within the
> restrictions of owl-dl, so you might as well go full (and use classes 
> as
> individuals).

Sounds good -- thanks.

> One approach is to only disallow the import/merge if it creates an
> inconsistency (and then choose to either disallow the whole merge or 
> just
> the offending statements). I've worked out an automatic inconsistency
> resolution mechanism for this purpose.

That's a smart way to go about it. I may bring this back up when it's 
time for 1.5. :)

> I wondered about this too. It's tough because it probably drives you 
> to some
> form of probabilistic reasoning (which might be a good direction but
> probably expands the scope of your project :-)

Definitely. Right now likn uses a fairly crude system. At below 50% 
confidence, it ignores the statement, unless it is specifically asked 
about it, or the statement is the only assertion about class, in which 
case it will qualify it with something like "could be X, but it's 
unlikely." Otherwise, it talks about the assertion with words like 
might and probably, depending on the score. If the score is high 
enough, it drops the qualifier. Once a statement has been ignored, 
dependent statements fail, so if it's unsure

> Of course you could always split the ontologies (in effect give each 
> user
> their own ontology). That's more the way I've looked at it - i.e. each 
> user
> has a view on the world, and is connected to those with similar views. 
> New
> information is propagated through the network according to some 
> threshold
> mechanism (e.g. two out of three of my friends believe fido is a dog, 
> so I
> guess fido is a dog; if I believe fido is a cat and cats are disjoint 
> with
> dogs than an inconsistency if created that will be automatically 
> resolved
> using least change heuristic).

The only issue I have with that is that individual users wouldn't get 
the benefit of expanding on others' assertions. Every user would have 
to explicitly say "fido is a dog," or at least vote on the assertion, 
whereas if that "fact" were already established with a good confidence 
rating by others, you wouldn't have to weigh in on it.

Thanks Geoff!

- ben

Received on Saturday, 5 March 2005 13:14:57 UTC