20011103: Berners-Lee, Machines and the Web

Contact:cdent@burningchrome.com


Berners-Lee, T., & Fischetti, M. (1999). Chapter 13: Machines and the
     Web. In _Weaving the Web: the original design and ultimate
     destiny of the World Wide Web by its inventor_. San Francisco:
     HaperSanFrancisco.

Discussion of the Semantic Web, a network of information sharing
between machines based on the presence of metadata represented in RDF.
The network is called semantic because an inference layer will allow
entities on the network to reach a consensual understanding. Such
understanding is supposed to enhance the generation of knowledge. 

-=-=-

There are so many ways to criticize this chapter (Berners-Lee has a
bad case of second system syndrome; he's up on a soapbox which is
distorting his view somewhat; he equates, to some extant, traversal of
a thesaurus as meaning acquistion; essentially he suggest that
computers can categorize, I don't think so: they classify) but who
wants to. This is such a nice pretty vision of the future I'm inclined
to support it despite the flaws. 

One area where this discussion makes a big win is in its understanding
of the power of brute force methods in granting computers some
semblance of intelligence. This was recently (200111) discussed on the
unrev-ii@yahoogroups.com mailing list: IBM has released some systems
that do self diagnosis and healing using techniques utilized for the
deep blue chess system. IBM discovered that to make deep blue smart
their best approach was to provide the computer with as much info as
possible from whlch it could do pattern matching. They call this the
brute force approach. Unrevvers liken this to the approach used with
cyc. 

The semantic web is the same: using the entire web as the brute
information force. 


Back to the Index