InBoxer won every anti-spam competitive review it has been in. This new version makes InBoxer faster and even more accurate. It improves upon the innovative language technologies that beat the other methods for fighting spam.

BOXBOROUGH, MA April 14, 2004 –- An updated version of the InBoxer anti-spam filter, winner of every competitive review in which it was compared, was announced today by Audiotrieve, LLC. This new release includes many customer-driven improvements, is more accurate, significantly faster, and easier to use.

“InBoxer outsmarts the spammers because it doesn’t just block spam, it uses language technologies to determine what each user considers to be a good message,” said Roger Matus, Chief Executive.

InBoxer learns by analyzing the visible words and invisible parts of messages and builds a personalized email language model. The text of an incoming message and other information in the message header are used to determine a good or bad message. The user may define trusted senders and companies. Incoming spam messages are put aside, but not deleted. Should InBoxer make a mistake, one button click corrects it. InBoxer learns by example and usually will not make the same mistake again.

Some of the innovations in version 1.2 include “balancing” of the good and spam language models for higher accuracy, an auto-activation system that eliminates the need to type activation keys, a new self-repair feature that can fix a problem before the user sees it, and advanced fine-tuning adjustments to control CPU usage and that allow Microsoft Outlook rules more time to run. Users will also notice a significant increase in performance speed.