Jon Udell says SpamBayes Rocks, and goes on to give examples of why. The best thing about SpamBayes, to me, isn't anything Jon mentioned explicitly; it's at the end of the article where he shows the analysis generated by Mark Hammond's plugin. That's what I love about the Outlook plugin and what frustrates me about all the other implementations I've used; they're completely opaque. I can't tell why Mozilla Mail thinks something's spam (incidentally, in the last week, the Mozilla Mail spam filter seems to have lost its effectiveness), or why Oddpost thinks something's spam, because it won't give me the report. SpamBayes, via the Outlook plugin, gives me that. Incidentally, there's more genius here: the settings are mine, not some third party's, not the choices of the other users of my mail service. Yahoo! Mail, for example, used to classify Chris Sells' newsletter as spam, presumably because somebody reported it as spam.

At the last meeting of the Denver Area Lisp Users' Group, Larry Hunter from CU Denver gave a talk on his work with Bioinformatics. Part of the presentation was a Powerpoint deck that he presents to new students describing why he uses Lisp for his work. Larry's talk was really interesting, the problem domain he's working with is huge. The way Larry described it, essentially genetics is a field where 10 years ago, a single dedicated individual could be an expert in their field and keep track of all the data related to their particular specialty. Now that sequencing a genome is relatively easy, the amount of data is overwhelming. So Larry and his students are creating a system that can answer questions like "What research has been done on this gene?", or "What differences are there between the genomes of these two species, or these two individuals?". One other interesting fact is that comparing genomes is basically a string matching problem on very big strings, so all that old, boring string matching algorithm research is finding new life in genetic computing. It's very humbling hearing about people solving problems like that. "Putting butts in airplane seats" is a piece of cake.