Or pretty much wasted, I would think.

Atrios excerpts a few paragraphs from this WSJ story about the useless anti-terrorist database:

Among the flaws in the database, which was quickly built by Lockheed Martin Corp. in the wake of the Sept. 11, 2001, terrorist attacks, is its inability to do key-word searches.

Their story is entitled “Flaws Found In Watch List For Terrorists”, but that's not a flaw. It's an indication that the thing is no good and probably wasn't designed to be any good. It's like designing an automobile without an engine. Sure, it looks nice. People could sit inside and listen to tunes and drink their coffee. But they couldn't really go anywhere.

Just to give some idea of the situation, this database supposedly “includes an estimated 400,000 people and as many as 1 million names”. “One million,” you may say, “how does it contend with this fantastic bulk of messy data?!” For instance, the Pennsylvania Academic Library Consortium (one of many such) has a fully-searchable automated union catalogue that contains over thirty million items. You can search by author, title, subject, publisher; you can combine or exclude those terms; you can limit your search by date(s) of publication, format, language. And that has been standard for a good ten or fifteen years now all over the world. So there seems to be something wrong.

Labels: Costly Data Problems