Re: A long but hopefully interesting introduction

On Mar 5, 2005, at 7:01 AM, Danny Ayers wrote:

> Sounds great!
>
> There are a few bots around #swig that may be of interest, like julie.
> (Incidentally the sembot thing is another resident lurker on my own
> to-do list).

Thanks -- I've been meaning to check out the existing bots!

> Like Phil, I'll be interested to hear how you intend to do this.
(syndicate every node, with its semantic relationships)

You mean technically? This part is thankfully just about done. The wiki 
updates a queue when a node is edited or replied to, and the chatbot 
updates the same queue when a new statement is made (it sends all three 
of the subject, predicate and object into the queue, unless the object 
is a literal value, which is somewhat rare). Then, hourly, the RSS 
daemon whips up RSS files for every node in the queue. If the item in 
the queue is a node, it pulls out the most recent 20 edits, replies and 
triples that the node is associated with. Alternately, if the item in 
the queue is a user or group, it generates a list of the 20 most recent 
edits by that user/group. The final way to subscribe is by individual 
word, which generates 20 most recent nodes which contain the word in 
question. To keep from doing vast amounts of unnecessary work, none of 
this happens unless a user specifically requests that the system start 
syndicating the thing in question.


> For an application like this, agreed 100%. The ecosystem should
> support a whole spectrum of trustworthiness, including bits related to
> ontologies.

Cool. That's good to hear.

[...]
> Again I think there's a bit of app-specificity in that for some
> purposes rigidity is desirable to maintain some kind of correctness,
> (like avoiding the apparently common confusion between the terms
> "traveller" and "terrorist"), at other times it may just misplace
> something in their blog index (and not land anyone in jail).

That makes sense.

> Whatever,
> it's probably worth considering whether some kind of proof mechanism
> can be used, so if you do wind up with unexpected conclusions you can
> backtrack to their source.

Yeah. Where I run into problems is if a constraint is suggested when 
the data doesn't match, yet probably should. For example, there are no 
constraints in the beginning, but 100% of people describe themselves as 
having one birth mother. Then somebody asserts that they have 20 birth 
mothers. When you tell likn "a person has exactly one birth mother" it 
objects. I guess in a case like this you err on the side of the data 
when it comes down to data vs. constraint...

> I'm beginning to get the feeling that SemWeb development is a little
> hampered by assumptions from other languages, especially those around
> XML & relational DB schema.
[...]

Personally, I think my hesitation in being more expressive with RDF/OWL 
comes not from another language, but from a fear of doing it Wrong -- 
especially so early in the adoption of the SW. There aren't many 
examples out "in the wild," especially of the type that I'm interested 
in, and those that exist (or are described in the drafts and 
recommendations) are fairly strict and formal. Even Shelly Powers, who 
uses great informal examples from her personal site throughout her book 
Practical RDF, titled her chapter on OWL "Ontologies: RDF Business 
Models." Business models? This won't get anyone excited to be more 
expressive in RDF/OWL.

> It may take more convoluted inference
> to get answers, but I reckon that flexibility to facilitate model
> "fitness" is more important than religious direct reuse and/or bending
> things for the sake of performance - that seems like premature
> optimisation.

That's music to my ears. Thanks Danny.

[...]
> In the context of RDF/OWL, I don't think homebrew and interchange are
> mutually exclusive - if anything, hopefully they'll be complementary
> (my homebrew can connect to your homebrew, thus connecting the
> interwiki upper ontology you reference to the standard furry quadruped
> vocabulary I use...).

I suppose connecting any two ontologies will require a little 
finessing, regardless of how they're represented...

> Wow, I got deja vu on that one, I must have asked the same question
> myself in the recent past. It's not very explicit in that doc, but all
> things being equal it's not so much a philosophical question as a
> computational one. If you start treating classes as individuals then
> it makes inference that much more complex. This may not be a problem
> if you're just using the graph model aspect of RDF, or wiring up your
> own app-specific reasoner, but if you're wanting to plug in an
> off-the-shelf DL engine then you'll have to abide by the language's
> constraints.

Interesting. I get the feeling that most OWL reasoners only support up 
to DL -- is that the case? If so, should I ignore this fact and hope 
that we'll see more software support OWL/Full?

[...]
> If
> each of these is maintained in a separate namespace, then there's much
> more flexibility for interconnection.

The only issue is that it gets verbose in chat mode, when every term 
has to be given the context of what namespace it comes from ("WordNet 
says X")...

[...]
> I suppose the
> ontologies should be versioned and annotated as cleanly as possible,
> but until you need to hook into other caches/triplestores which
> remember your earlier assertions on the Web there shouldn't be too
> many problems (datestamped annotations are probably a good idea).

Or via reification? Along with the confidence in the statement, likn 
could export the last time the statement was verified...

> There's additional help when it comes to more, errm, humanish
> terminology (great for blogs and Wikis) in the form of SKOS, which
> allows for less tightly-bound relationships between terms without
> throwing away all the reasoning potential.

I hadn't seen SKOS before -- very interesting....

all the best,

- ben

Received on Sunday, 6 March 2005 02:02:37 UTC