My current side-project is doing something like this. I have IM and IRC logs going back to 2001 and wanted them in a single format in a database.<p>I'm at the point where the logs can be loaded and searched, and am currently trying to decide what to do about merging the identities of different people. I've got about 2000 unique names and have been thinking about what sort of interface to use to unify and hide ones that don't matter (exactly the problem the author of this article is having).<p>I didn't think it was viable for a business, so I'll be curious to see what happens with them.<p>If anyone's interested in my source code, it's on github (<a href="http://github.com/jeffjenkins/cdb" rel="nofollow">http://github.com/jeffjenkins/cdb</a>). It handles Adium (HTML and XML), Miranda, Trillian, and Colloquy (and almost mIRC)<p>edit: there's also a fairly basic UI which I can put up somewhere if people are actually interested in using this.