|
As the number of pages on the Internet is still growing exponentially and more and more people are realizing that knowledge (and the ability to learn) is the ultimate business asset and that data != information and information != knowledge, searching becomes the most prominent problem domain of the Internet. Even Microsoft seems to be interested in building its own internet-scaled search engine lately.
Several approaches to the semantic web claim to know how to build the ideal meta data based searching infrastructure. As Tom Bray's excellent series on searching told me, there is always a trade-off between having the desired set of meta data and actually getting people collecting these meta data (being able to search is the only reason you collect document meta data in case you didn't know :-) - given that navigation is a special searching discipline).
Google's major achievement are useful page ranks without having someone to collect explicit meta data.
However having a single giant index for the whole Internet seems to be flawed from the start. So I am happy to see people stepping forward suggesting distributed search architectures for the Internet. What I don't like with this proposal is that it uses SOAP. Immediately Tim Bray's proposal came to my mind. It somehow appears most suitable for that kind of architecture.
7:39:59 PM 
|
