http://kuurort.com/imke/2007/10/29/semantic-web/ interactive media and knowledge environments Tue, 06 Jan 2009 10:46:17 +0000 http://wordpress.org/?v=2.6.2 http://kuurort.com/imke/2007/10/29/semantic-web/#comment-6 Mauri Sun, 04 Nov 2007 08:24:45 +0000 http://kuurort.com/imke/2007/10/29/semantic-web/#comment-6 I am equally curious to see Twine in work, but from the theoretical point of view I'm a bit doubtful. Having the cognitive science background, I think there is too much similarity between Semantic Web and the dream of artificial intelligence in the 50's and 60's. Both share the idea of some fixed and 'true' hierarchy of representing the world. I rather like the idea that for all things there are multiple equally good ways to see them and that the ontology of things evolves and gets shaped all the time due to people's contributions, that is, 'soft ontology'. The problem to which 50's AI dream stumbled was the complexity of rules. In order to cover all intelligence, the project ended up coding not clear rules, but endless chains of exceptions. Same way I predict that the hierarchic tree of the Semantic Web will typically become too complex to be useful. But I may be wrong too... I am equally curious to see Twine in work, but from the theoretical point of view I’m a bit doubtful. Having the cognitive science background, I think there is too much similarity between Semantic Web and the dream of artificial intelligence in the 50’s and 60’s. Both share the idea of some fixed and ‘true’ hierarchy of representing the world. I rather like the idea that for all things there are multiple equally good ways to see them and that the ontology of things evolves and gets shaped all the time due to people’s contributions, that is, ’soft ontology’.
The problem to which 50’s AI dream stumbled was the complexity of rules. In order to cover all intelligence, the project ended up coding not clear rules, but endless chains of exceptions. Same way I predict that the hierarchic tree of the Semantic Web will typically become too complex to be useful. But I may be wrong too…

]]>