Spam Filtering
Choosing the Right Anti-Spam software
Spam is easily one of the biggest annoyances for those who use the internet as their form of communication. The numbers on these bothersome messages is unsettling when considering that it’s believed that more than 80% of all email is spam. Most of it is unsolicited junk mail while a portion contains the latest scam or viral strain. When viewing the situation from the latter, this is certainly the type of email you want to avoid at all costs.
There are various techniques and mechanisms used to aid in the fight against spam. One of the most popular methods involves software solutions. In this article we will discuss spam filtering software and offer a few tips to keep in mind when searching for a program.
Spam Filtering Methods
First and foremost, it is important to understand that no single spam filtering method is 100% effective, mainly because spammers get more sophisticated each year. You can however, get quality results out of a software product that combines a variety of anti-spam techniques. Some of the most common filtering methods include the following:
• Bayesian filtering
• Heuristic analysis
• Signature matching
• Traffic-pattern analysis
Bayesian Filtering
In regard to anti-spam products, a Bayesian filter works by comparing words in a spam email to those found in a legitimate message. This type of filter uses a scoring sample to determine spam. For instance, words that appear in a spam sample are given a high score while those in a legitimate message get a low score. The compilation of these numbers is used to evaluate incoming messages, so if the score is above the threshold you specified, the email is then marked as spam.
Heuristic Analysis
Heuristic analysis is a method that uses various rules to detect spam. If a message violates the rules, it is then deemed as spam. Some of the common rule violations are the excessive use of dollar signs, phrases like “make money” or “click here” and messages that have all capital letters in the subject field. This method is similar to Bayesian filtering in the way that it allows you to set values and classify spam as any email that exceeds the predefined threshold.
Signature Matching
Signature matching is a unique method that uses email accounts as decoys to catch spam. The accounts are generally created and monitored by the spam filtering software vendor. When messages reach the decoy account, those fingerprinted as spam are stored into a database. Any subsequent emails with fingerprints matching the message in the database are classified as spam.
Traffic-pattern Analysis
A spam filter equipped with traffic-pattern analysis examines the characteristics of email transmission. For example, if someone sends out a lot of messages in a short amount of time, then the software will likely conclude that the person is distributing spam. Other characteristics include sending messages to a large number of invalid email addresses and sending mail where it looks as if attempts were made to hide the sender’s identity.
What’s Available
There are several anti-spam programs on the market from commercial products such as SpamFighter and Spam Buster to open-source solutions like Spam Assassin and MailWasher. While they all serve the same general purpose, each program comes with different features and capabilities. Regardless of the software, make sure you find something that allows you to inspect the messages yourself before discarding them entirely. Some products are very aggressive and may cause you to lose out on important email. We recommend that you read a few reviews to get some user feedback and understand what to expect from a particular spam filter.
Leave A Comment