Four Pillars: Some musings on enterprise search

Have you been following what’s been happening at WWW 2006 in Edinburgh recently? One of my favourite cities, wish I could have been there.

It’s worth checking the site out. Some interesting papers and discussions. Incidentally, there’s a footnote to each paper:

Copyright is held by the International World Wide Web Conference Committee (IW3C2). Distribution of these papers is limited to classroom use, and personal use by others.

Ironic, isn’t it? To see a condition like that placed on research about the web, on the web, relating to a conference attended by Sir Tim Berners-Lee? Oh well.

But I digress. Some interesting papers. Like this one, on Using Annotations in Enterprise Search. The gist of the argument appears to be that intranets are different from the Internet, and that we can use annotation and similar feedback loops to improve the enterprise search experience. Nothing particularly earth-shattering, but worth a read. We can always learn.
I was more intrigued by some of the assertions made in the paper. Examples:

  • “….company employees cannot freely create their own Web pages in the intranet”
  • “….algorithms based on link structure analysis….do not apply to intranets the same way as they apply to the Internet”
  • “….the amount of anchor text, one of the major factors used by Internet search engines….is very limited in intranets”
  • “….One such characteristic is the absence of spam in intranets”

I wonder. Don’t blogs and wikis lower the barriers to web page entry? Can’t we improve the quality and quantity of anchor text by using such tools? Sure, I believe that annotation, tagging, feedback loops and collaborative filtering all need to migrate from the world outside to the enterprise. But it’s more than that. All this is not worth doing unless we’ve lowered the barriers to conversation, to publishing, in the first place.

I wonder. Isn’t spam nothing more than an act of commission or omission that clogs up arteries? Everyone understands spam in an e-mail context. Spam exists all over the place in organisations. Yes there is internal e-mail spam as well. There’s “Meeting-Agenda” spam. There’s “Things-that-need-deciding” spam. There’s “Things-that-need-prioritising” spam. Fundamentally there’s spam in every workflow channel; spam covers a multitude of techniques used by professional organisational men to deny service, particularly when it comes to prioritisation or even decision.

I wonder.

Amongst other papers of interest are:

Collaborative exchange of news feeds

Visualising tags over time

Semantic Wikipedia

The impact of online music services on the demand for stars in the music industry

I don’t agree with everything that’s in those papers. In fact, for some of the papers, the disagree bit outweighs the agree bit. But that’s what makes it worth my while, seeing other points of view and figuring out what the differences mean. As Gregory Bateson said about information, it is “a difference that makes a difference”.

Let me know what you think

This site uses Akismet to reduce spam. Learn how your comment data is processed.