Anti spam filter reviews, junk mail advice and spam blocker user ratings
Anti spam filter reviews, junk mail advice and spam blocker user ratings from WhichSpamFilter.com
Spam Filter Reviews, Anti Spam Tips, Advice and Spam Filter Ratings
Home | Dealing With Spam | Types of Filters | Spam Filters | Our Reviews | Resources | Forum
Contents

Description
Verdict
Installation
Interface
Features/Operation
Accuracy
Conclusion
Back to Review Index


Newsletter

Sign up for our newsletter and receive the latest information on the best new filters as they release, special member-only offers and what is happening in the world of spam fighting!

E-mail Address:

We never share your e-mail address with any third party!
You can opt out of mailings at any time.
Anti spam filter reviews, junk mail advice and spam blocker user ratings from WhichSpamFilter.com

Outlook Spam Filter Review

Overall Rating:
Anti spam filter reviews, junk mail advice and spam blocker user ratings from WhichSpamFilter.com

 

Price: $25.00

Description

As the name would suggest, Outlook Spam Filter is an Outlook plug-in. It is a Bayesian filter, with whitelist/blacklist functionality and rudimentary content-based filtering.

 

Verdict

Outlook Spam Filter took a while to "settle down" and start returning decent capture rates, but in the end proved quite good at correctly identifying messages. Others have been more accurate, but good results nonetheless.

Nice clean interface with the minimum of interaction required by the user.

 

Installation

Installation went smoothly. During installation, it scans your sent items folder to "learn" what constitutes a good message.

Outlook Spam Filter's web site states that the filter is "pre-trained" from more than 30,000 spam letters, so should be working well from installation.

 

Interface

Familiar to previous users of Outlook add-in spam filters, Outlook Spam Filter adds its toolbar to the Outlook Interface:

The only other interface element that the user will occasionally need to interact with is the settings dialog, which presents a clear and intuitive interface:

 

 

Features and Operation

One of the big advantages to Bayesian-based filters is the lack of interaction required in daily use. It is mostly a simple case of using two buttons to correct wrongly identified messages (In this case by clicking either the "Add to Friends" or "Add to Enemies" buttons on the main toolbar).

In this case however, we did find a minor annoyance: whenever you click the "Add to Enemies" button, such as when a spam message makes it through to the inbox, you are presented (every time) with the following dialog:

Both of these questions would be better moved to an option in the settings dialog. The answer to both of these questions would most almost certainly be the same every time it is answered, but this dialog popped up every time.
We even tried to disable the enemies list in the settings dialog, but this just earned us an extra option in the message box asking us if we would like to enable it again! No big deal maybe, but we did find this getting on our nerves somewhat (maybe too much coffee and too little sleep?)

Aside from the Bayesian filtering, which is the main heart of Outlook Spam Filter, there is some rudimentary content-based filtering. Messages can be flagged as spam on certain header conditions and there is the ability to add some "Spam Recognition Keywords".

These are simple text only matches so will add little to the capture rate of the filter.

A click on the "Bounce" button and a forged undeliverable message will be sent to the sender of the currently selected message.

 

Accuracy

During our week-long test of Outlook Spam Filter, we noticed a definite increase in accuracy as it "learned" from the messages that were received.
We definitely didn't get the 97% accuracy that their web site promises "out of the box", but things did improve the longer the test went on.

Unusually, we had more false positives than false negatives which prompted us to pay more attention to searching the spam folder for incorrectly identified messages.

In the end, the accuracy returned was 93.61%. Not as good as we would have liked, but we feel this would definitely have improved with time.

Message Count Spam False Positives False Negatives Accuracy
1361 89.05% 3.53% 2.87% 93.61%

 

Conclusion

Outlook Spam Filter proved itself to be a stable and simple to use solution.

The overall average accuracy returned from our tests was not as good as similar filters we have tested, but it has to be considered that Bayesian filters only reach optimum accuracy after as much as a month of continuous use. We could see it improving toward the end of our tests.

Because of this accuracy figure though, it didn't quite make our 5 star rating.

Anti spam filter reviews, junk mail advice and spam blocker user ratings from WhichSpamFilter.com

Return to Review Index

 

Spam Filter Reviews, Anti Spam Tips, Advice and Spam Blocker Ratings