Re: [tlug] Effect of AI and data processing rift on the Semantic Web

[Edward Middleton <emiddleton@example.com>]

Stephen J. Turnbull wrote:
> Edward Middleton writes:
>
>  > Yes, and it is a shame the same couldn't be said for your commentary.
>
> Not until you've done a little thinking.  Then it will very likely
> start to grow on you. ;-)
>   

Well no, I am coming to the conclusion you left before the show started ;)

> The semantic web originally was supposed to build on top of the web as
> we know it, applying AI to extract semantic relationships.  But what
> they discovered almost immediately was that the semantics generally
> aren't there, that they have to be injected into the existing content.
> This led to attempts to define standards for including semantic
> relationships by committee, but that's wrong-headed.
>   

Getting semantics into web content isn't the real challenging problem of 
the semantic web.  If there is a perceived benefit people will do it.  
Most sites have RSS feeds because people see value in doing it.  Granted 
most current applications of RSS are pretty boring.  FOAF[1] 
applications are more interesting but they are still not very 
sophisticated, i.e. let all my friends and their friends post directly 
to my blog[2].

The article was about how the adoption of the PDP 6 by MIT and Stanford 
AI groups in 1964 and IBM's dominance in the data processing field lead 
to a rift between these two fields which is having negative impact on 
the development of the semantic web.  i.e. AI researchers involved in 
the development of the semantic web are not benefiting from the advances 
made by IBM and others in the data processing field.

Edward

1. http://www.foaf-project.org/
2. http://tlug.jp/ML/0804/msg00242.html