Stop spam!
Are you getting too much spam? (Here we're talking about the unwanted e-mail kind of course!)
There are plenty of spam-fighting products available. Some products work only with certain e-mail programs (for example Outlook, Outlook Express, or Thunderbird. Some come with a one-time cost, some require an ongoing subscription. They employ a variety of spam-fighting techniques. It's worth knowing how a product works before buying one, or even before installing the (usually free but time-limited) trial product, so you don't waste time on something that doesn't meet your needs.
Programs or plug-ins that work entirely on your computer. This capability may already be built into your e-mail program! For example Thunderbird has a junk mail filter built-in. You can also buy add-ins to do this kind of filtering; an example is ella for Spam Control. These spam filters use one or both of two techniques: Adaptive filters and sender lists.
Adaptive filters attempt to "learn" the difference between good and bad e-mail. Ads for these anti-spam products may include words like "heuristic" or "Bayesian". Usually you train the filters by using them to sort through your e-mail. As you mark items as spam or not spam, the filters analyze the messages to look for patterns that they will use to classify future incoming messages. The main shortcomings of this anti-spam techique include the risk of overtraining and the changing nature of spam.
Overtraining happens when you mark too many messages as spam: You've seen those messages that have little or no mention of any product or service, just lots of meaningless words or sentences. The spammer is using those words and sentences to get past the filters. As you mark those messages as spam, you are training the filter to reject messages that look like English text, which makes the filters less effective at recognizing messages you really do want to receive.
The other problem with "learning" filters is that they don't keep up with the changing nature of spam. On the simplest level, you try to train the filters to reject misspelling of "viagra" or "pharmacy", but you've probably noticed there seem to be an infinite variety of misspellings, including the use of spaces and punctuation. You've also seen spam that contains one big picture of words, with noisy dots and slanting text to prevent software from detecting the words in the picture. Spammers know all the anti-spam tricks, and are continually working to get around anti-spam systems.
Sender lists are used in many anti-spam products. This involves sorting your messages into approved senders and blocked senders. The anti-spam software might add a "block this sender" button to the mail program. However, you've probably noticed that spam often comes from a different sender every time, so the blocked sender list is of little use.
Some anti-spam products work by implementing an extreme form of sender lists, called challenge and response: Every incoming e-mail is assumed to be unwanted unless the sender is already in your "approved" list. Everyone else is sent a message that requires them to verify that they are a real person in order to have their message delivered. This method is highly effective at reducing unwanted e-mail because spammers don't respond to the challenge. Unfortunately, it's also annoying to legitimate senders, since it delays the arrival of their message to you and takes more action on their part. It's also more work for you: You'll probably end up looking through the blocked messages to see if "good" messages are stuck there. If you frequently receive mail from unexpected senders (for example if you hope new clients will contact your business via e-mail) then challenge and response is probably not for you. Spamarrest is an example of a challenge and response system.
Another anti-spam technique involves installing software that makes your computer part of an anti-spam network. Examples of this type of software is Cloudmark, formerly known as Spamnet, and Spamfighter. This software consults a central anti-spam server to learn what the newest spam looks like. As other users of this software mark messages as spam, that information gets reported to the central server. When your e-mail program downloads your mail, it compares it to the information on the server to decide if the message looks like spam. It's like other people are training your computer to recognize spam! This spam-fighting method keeps itself up-to-date with the latest spam techniques. Disadavantages of this technique are that it requires a fast internet connection to consult the anti-spam server and that a monthly subscription to continue getting the latest spam profiles.
In summary, there are a lot of anti-spam products which use a variety of strategies to cut down on spam. The best product for you depends on which e-mail client you use, how you use e-mail (e.g. do you have a close circle of friends or do you want to get mail from new senders), and whether you want to pay for ongoing services.
- JB
There are plenty of spam-fighting products available. Some products work only with certain e-mail programs (for example Outlook, Outlook Express, or Thunderbird. Some come with a one-time cost, some require an ongoing subscription. They employ a variety of spam-fighting techniques. It's worth knowing how a product works before buying one, or even before installing the (usually free but time-limited) trial product, so you don't waste time on something that doesn't meet your needs.
Programs or plug-ins that work entirely on your computer. This capability may already be built into your e-mail program! For example Thunderbird has a junk mail filter built-in. You can also buy add-ins to do this kind of filtering; an example is ella for Spam Control. These spam filters use one or both of two techniques: Adaptive filters and sender lists.
Adaptive filters attempt to "learn" the difference between good and bad e-mail. Ads for these anti-spam products may include words like "heuristic" or "Bayesian". Usually you train the filters by using them to sort through your e-mail. As you mark items as spam or not spam, the filters analyze the messages to look for patterns that they will use to classify future incoming messages. The main shortcomings of this anti-spam techique include the risk of overtraining and the changing nature of spam.
Overtraining happens when you mark too many messages as spam: You've seen those messages that have little or no mention of any product or service, just lots of meaningless words or sentences. The spammer is using those words and sentences to get past the filters. As you mark those messages as spam, you are training the filter to reject messages that look like English text, which makes the filters less effective at recognizing messages you really do want to receive.
The other problem with "learning" filters is that they don't keep up with the changing nature of spam. On the simplest level, you try to train the filters to reject misspelling of "viagra" or "pharmacy", but you've probably noticed there seem to be an infinite variety of misspellings, including the use of spaces and punctuation. You've also seen spam that contains one big picture of words, with noisy dots and slanting text to prevent software from detecting the words in the picture. Spammers know all the anti-spam tricks, and are continually working to get around anti-spam systems.
Sender lists are used in many anti-spam products. This involves sorting your messages into approved senders and blocked senders. The anti-spam software might add a "block this sender" button to the mail program. However, you've probably noticed that spam often comes from a different sender every time, so the blocked sender list is of little use.
Some anti-spam products work by implementing an extreme form of sender lists, called challenge and response: Every incoming e-mail is assumed to be unwanted unless the sender is already in your "approved" list. Everyone else is sent a message that requires them to verify that they are a real person in order to have their message delivered. This method is highly effective at reducing unwanted e-mail because spammers don't respond to the challenge. Unfortunately, it's also annoying to legitimate senders, since it delays the arrival of their message to you and takes more action on their part. It's also more work for you: You'll probably end up looking through the blocked messages to see if "good" messages are stuck there. If you frequently receive mail from unexpected senders (for example if you hope new clients will contact your business via e-mail) then challenge and response is probably not for you. Spamarrest is an example of a challenge and response system.
Another anti-spam technique involves installing software that makes your computer part of an anti-spam network. Examples of this type of software is Cloudmark, formerly known as Spamnet, and Spamfighter. This software consults a central anti-spam server to learn what the newest spam looks like. As other users of this software mark messages as spam, that information gets reported to the central server. When your e-mail program downloads your mail, it compares it to the information on the server to decide if the message looks like spam. It's like other people are training your computer to recognize spam! This spam-fighting method keeps itself up-to-date with the latest spam techniques. Disadavantages of this technique are that it requires a fast internet connection to consult the anti-spam server and that a monthly subscription to continue getting the latest spam profiles.
In summary, there are a lot of anti-spam products which use a variety of strategies to cut down on spam. The best product for you depends on which e-mail client you use, how you use e-mail (e.g. do you have a close circle of friends or do you want to get mail from new senders), and whether you want to pay for ongoing services.
- JB
Labels: anti-spam, antispam, block spam

0 Comments:
Post a Comment
Subscribe to Post Comments [Atom]
<< Home