Subscription Based Service
eMail2Pop's built-in spam filter requires a subscription. Subscribers are
entitled to spam filter engine updates and filter definition files (FDF)
updates. When a filter update is available, eMail2Pop can notify users of
the update.
Subscriptions are available in 3 months, 6 months, 12 months, and unlimited
time periods. A subscription is not required if you do not wish to use
eMail2Pop's integrated spam filtering features. Subscriptions can be
purchased and extended at any time at the
license upgrade page.
Scanning Functionality - How It Works
If eMail2Pop has a valid spam subscription, spam filtering is
automatically activated. When a POP3 e-mail client requests AOL e-mail to
be downloaded, eMail2Pop will forward the e-mail to the spam filter for
processing before forwarding the AOL mail message to the POP3 client. The
filtering activity is completely transparent and does not require any user
intervention.
Message processing times vary depending upon computer processing speed and
internet connection throughput. Because the spam filter runs many
analytical tests on patterns and expressions in e-mail messages, a 1Ghz CPU and
256MB of ram should be the bare minimum for a computer running eMail2Pop.
By default, the spam filter will also query multiple online databases for
positive spam matches, so internet connection speed is also an important
factor. On average, each message takes 4-10 seconds to analyze on a
broadband internet connection.
Bayes Auto-learning Capability
Each e-mail message analyzed by eMail2Pop is broken down into individual words and phrases and logged into a database on the computer. Over time, the Bayes system will adjust to identify new patterns in both spam and non-spam e-mail. Words that are commonly shared in spam e-mails will be given a higher score, and words that are commonly associated with non-spam e-mails will give the message a lower score. Because the Bayes system requires a sizable database of words before it can produce accurate results, Bayes scoring is not activated until at least 50 high-scoring spam and 50 low-scoring non-spam e-mails are entered into the database.
Analyzed Message Output - Results
eMail2Pop's spam filter utilizes a scoring system. The filter analyzes e-mail messages with a list of expressions and rules. When a match is made, points are added to the message. In addition, points are added to the message when a positive match has been made with an online database. The Bayes auto-learning database will also assign an applicable score based upon the likelihood of the e-mail being spam through an analysis of the word content. If the overall score surpasses a threshold, it will be marked as spam.
If a message is marked as spam, eMail2Pop generates a new e-mail message with a content preview and a breakdown of scoring. The suspected spam message is attached to this e-mail. Suspected spam e-mails are sent to the POP3 e-mail client as an attachment to prevent the unwanted execution of malicious scripts and/or tracking devices commonly used by spammers to verify if an e-mail account is active.
Sample Analysis