 |
|
|
|
|
|
|
|
|
There are two major categories of email spam filter methods. One tries to identify the email content. The assumption behind this kind of email spam filter is that spam is trying to relay information, even if it’s concealed, and that message can be categorized. The other approach works with the message’s routing metrics, like the sender’s address, ISP, or domain, either to verify that the source is legitimate or as the criteria of the email spam filter.
|
|
|
|
More specifically, email spam filter software usually uses one of these types of filters:
- Spam fingerprinting looks both the header information and the message body, and stores it. Every incoming message is compared to that database. Spam fingerprinting is usually used in conjunction with honeypots, bait email addresses which are never given out, so only unsolicited messages ever arrive. Spam fingerprinting is effective, but impaired by time delays updating the database. Bait email accounts are extremely dangerous for networks to implement.
- A Bayesian email spam filter looks at the order of words to determine context and, therefore, the likelihood that it is spam. Bayesian algorithms can be very accurate at identifying spam with few false positives, but this kind of email spam filter is limited by images or HTML-based messages which may not use text messages.
- Blacklisting compiles lists of IP addresses, hostname, email addresses, and domains to block. Like spam fingerprinting, this is limited to only spammers which have been identified by the administrator.
- A Heuristics email spam filter verifies the sender’s address and host and domain names, as well as the server routing information in the header to ensure that senders are whom they claim to be, since spammers obfuscate their identity to avoid being caught.
LivePrism uses a combination of all these methods in a series of filters, each email spam filter focused on a different vulnerability:
- Reputation D email spam filter which uses IP address analysis to identify spam and slows spam delivery while speeding up “good” email delivery.
- A spam fingerprint database, which uses the email spam filter to immediately recognize known spam messages, including image-rich messages.
- Virus scanning email spam filter, which identifies messages with attached viruses or malware.
- User-defined content filters scan messages for specific words, phrases, or charactersets, attachment file types and sizes, and other preset policies.
- Blacklisting email spam filter, which is set by individual users as well as administrators, blocks emails from specified senders or domains.
A hosted service model makes implementing an effective email spam filter easier for your business, because the burden of configuring and maintaining high-quality filtering lies with LivePrism, not your IT team:
- Automatic updates to spam databases, filters, and software
- Millions of emails and users as a pool to identify more spam emails, more quickly
- Dedicated email spam filter resources, without additional software, hardware, or maintenance
- More filtering technologies integrated with more features, like reporting and monitoring, without additional software or setup
|
|
|
|
