[Wolves] Semantic web viability

[David Goodwin]

Luke Redpath wrote:
> Sorry, I wasn't able to make it last night - lots of stuff going on.
> 
> Aq, perhaps you could elaborate then? My understanding of the term 
> "semantic web" as it is commonly used has always been the use of 
> correctly structured XHTML or XML (usually XHTML) to markup documents 
> appropriately. Something I certainly don't think is too much extra work 
> and should always be encouraged over "tag soup", as it were.


Aq said something along the lines of that the eventual goal is for 
machine readable/understandable data (i..e well structured etc) which 
could be used to make inferences from.

Eventually it would lead to the situation where an "agent" (e.g pocket 
pc) could organise a doctors appointment for you, or order medicine for 
you using information from numerous sources (e.g. Drs waiting list, your 
credit card, your availability etc)


His arguments were that :

1) Data can't be trusted
2) People don't like maintaining meta-data (yet alone their own web pages)
3) People lie with metadata (e.g. html keywords) in an effort to boost 
page impressions/profit etc.
4) People don't always know what they're doing etc.
5) Security is an issue (i.e. if I trust you, and you trust him, so I 
therefore trust him?)


Blah



-- 
David Goodwin
w: http://www.clocksoft.co.uk
e: david.goodwin at clocksoft.com
t: 0121 313 3850