My knowledge of neural net software is limited to a cartoonish understanding

of the theory behind them, but as I understand it they might be a useful tool for identifying the author of a given text stream. Comments? Anyone on the list familiar with neural net implementation, or in need of some sort of

academic project? :)

-- Greg Broiles greg@goldenbear.com Baked, not fried.

Actually there is no need to use neural networks. There is quite a lot of software running around that does a statistical anaysis of two authors works and then calculates the similarity. You should be able to find copies in either linguistic or mathematical archives.