|
|
For many email users, the number of SPAM messages received are a source of inconvenience, frustration and a means of lowering
productivity levels. Information and general techniques to follow in order to avoid being added to SPAM emailing lists may
be found by following this link.
BGNet email users are also susceptible to the effects of SPAM email messages. A recent one-hour sampling of inbound BGNet
email messages produced the following results:
- Inbound mail load during the test period was 18,104 messages
- During the test period, 9,767 messages were identified at the 100% SPAM confidence level, which equates to 53.95% of all inbound
mail
- During the test period, approximately 70% of all inbound mail was assigned a 99% or greater SPAM confidence level score.
- During the test period 4,748 unique users received a message identified at the 100% SPAM confidence level, which equates to
approximately 11% of all BGNet users
- The average number of messages identified at the 100% SPAM confidence level during the test period, among unique users who
received at least one, was two messages per user.
While this is not fully representative of BGNet mail traffic patterns across all situations, this sample date and time period
information is in line with statistical estimates of SPAM traffic volume and measurements of typical inbound message loads
for the past 12 months.
In October 2003, the Intelligent Inbox technology was introduced to BGNet users, allowing the filtering of email messages such as SPAM messages once activated and
configured by each account holder. Some account holders have activated this technology to assist in monitoring their incoming
email.
Currently all incoming BGNet messages that do not originate from a BGNet sender are scanned and evaluated for SPAM characteristics.
These messages are then assigned a confidence score. The header of each email message assigned a confidence score will include
a X-MASF:xxx% line. Users can choose to act on these scores by using the filters available through Intelligent Inbox or the
filtering features of their email client. Documentation in pdf format on using these features may be found at:
The product used to generate the confidence score has been and will continue to be upgraded. It continues to grow in sophistication
of the technology used to identify and assign a confidence score to email messages. Information about the techniques used
to assign the confidence score is available at: how a SPAM confidence score is determined
This sophistication has allowed confidence in the decision to filter SPAM by blocking email messages that have been assigned a 100% confidence score beginning January 19, 2005.
Some important points to note with this implementation:
- there will be no effect on messages sent between BGNet account
- mail with a 100% confidence score wil no longer be delivered to a "junk" mailbox, these messages will be blocked rather than
being delivered
- each BGNet account holder can identify messages assigned a confidence score by referencing the X-MASF:xxx% header line of
these email messages
- examples of email messages by sender and by subject line from the one-hour test period are available at these links for your
reference
Each account holder can continue to further curb the number and type of messages being delivered to their email account using
the filtering capabilities of the Intelligent Inbox or their email client. It is believed that this additional filtering level will provide relief to the number of SPAM messages
being received by BGNet account holders.
|
|
