SourceForge.net Logo
Summary
Forums
CVS
Download

Laird Breyer
Download
contents introduction tutorial spam fun man related
previous next

TREC

The (United States) National Institute of Standards and Technology organises an annual conference on text retrieval called TREC, which in 2005 began a new track on spam filtering. A goal of this conference is to develop over several years a set of standard methodologies for evaluating and comparing spam filtering systems.

For 2005, the initial goal is to compare spam filters in a laboratory environment, not directly connected to the internet. An identical stream of email messages addressed to a single person is shown in chronological order to all participating filters, which can learn them incrementally and must predict the type of each message as it arrives.

The spamjig is the automated system which performs this evaluation. You can download it yourself and run it with your own email collections to test any participating filters. Special instructions for dbacl can be found in the TREC subdirectory of the source package. Many other open source spam filters can also be tested in this framework.

previous next
contents introduction tutorial spam fun man related