CEAS
2006
Springer
14 years 3 months ago
2006
Springer
We describe a very simple technique for discriminatively training a spam filter. Our results on the TREC Enron spam corpus would have been the best for the Ham at .1% measure, and...