Saturday, August 27, 2005

Bag the Web

Okay, lets face it, the web is useless.

All I want is to find some software, preferably free, that will read a text member and tell me what non-trivial words predominate in the text, so they can be used as keywords.

Nothing shows up in the first two pages of a Google search. Well, okay, articles about stochastic, bayesian, random multi-linear non-chaotic fivverpudlian search algorithms and the like, all of which are interesting, and a couple are even comprehensible to me. I've been interested for quite some time, in a layman's way, in the general concept of text classification, so I have a vague sense of how hard it can be to do it well. But I don't want it done well -- more accurately, I don't need it done well. What I want is something along the lines of Text Keyword Generation for Dummies.

Do I expect too much?

(You betcha.)

No comments: