14 October 2007


Human-computer hybrids, in indexing

I recently completed an (as-of-yet unreviewed) article for Information - Wissenschaft & Praxis (IWP), the premier German journal on information science. The topic was the intersection of computer-based indexing and human indexing, and how these two approaches to indexing are unequal but in many ways compatible.

The biggest challenge in writing the article comes from the simple fact that I'm a human indexer, and that I believe that automatic indexing fails every important quality test. On the other hand, since it's unlikely we're going to have people typing away to index the World Wide Web (see my entry "A needle in a haystack with 100,000,000 blades"), it seems we're going to need something faster than human fingers and brains to get the job done.

I'm not going to repeat the article's ideas here, except to say that I tried to give an even-handed view of automatic indexing -- even as I tend to rip it to shreds in this blog when I can. The distinction is that people constantly overestimate when it's necessary. Automatic indexing is overused and misused.

Still, I thought my loyal readers might knowing that even on this, there are two valid opinions.

Labels: , , ,

This page is powered by Blogger. Isn't yours?