BeanSoftware Logo

ASP.NET Database Search Control

Anti Spam Filtering Methods

We all know that spams are the junk emails that you receive daily in your inbox and they are really a headache to most of the people. The precious time of the employee is wasted to browsing through the spam emails. Spam filtering is done to filter those junk emails from your inbox so that you save a lot of time. With spam filtering your receive only the genuine emails that are intended for your reading. Lots of products are available in the market to deal with spam.

Statistics reveal that lot of money is wasted by the way of lost productivity due to spam. Some of the anti-spam solution providers have calculators that can be used to calculate the amount lost in a year due to spam. They also provide stats that allow you to find out how much money a company can save by using anti-spam products that is in the market. These calculators are intended to promote their anti-spam solutions. Due to spam lots of memory is wasted in junk contents and the network traffic also is increased decreasing the useful bandwidth.

Using Artificial Intelligence in filtering spam

Now-a-days artificial intelligence is used in filtering spam from your inbox. This artificial intelligence filtering technology is used to adapt faster to the fast changing methods that are adopted by spammers. The software that is used to filter spam is made to learn the new methods that are adopted by the spammers so that it can filter out the new spam that come to your inbox. For this purpose the natural language processors are used. They are powerful algorithms that are used to scan your incoming messages and filter the spam. They come along with the mail filtering software.

These powerful algorithms in artificial intelligence read the messages and finds out the junk emails by analyzing the email content. Artificial intelligence techniques also look at the time at which the message was sent, the server address from where it was sent, and whether there is any other email like this that was sent earlier to the recipient. These factors are also considered before sending an email to trash. The steps that are involved in such a filtering are very simple.

The artificial intelligence filtering software is the first to receive all the emails that are sent to the corporate. It filters the emails, flags some email that are considered spam and sends it to a spam holding area. The genuine emails are sent to the email server through the firewall. An administrator to determine which emails are to be sent across to the email server can review the spam holding area. This update by the administrator is noted down by the software to process it in future. Thus artificial intelligence is used to filter spam from reaching even the firewall. The emails are scanned even before they reach the corporate firewall.

Different types of Anti-spam filters available

There are many types of anti spam filters available in the market. Based on how they work the filters can be categorized as Internet based spam filters, Statistical/Algorithmic spam filters, and General filters. In the Internet based filters the definitions on which the spam is filtered is available in the Internet on a remote server. Statistical / Algorithmic spam filters have the algorithms for filtering in the local machine itself. These filters are updated based on the training that your give to these software. The other type of filter called the general filters use the keywords that you feed and the email addresses that you have in your address book to filter the incoming emails. Based on the location of the filter we can categorize the anti spam filters in to Server side filters and Client side filters. As the name indicates, the server side filters reside in the mail server and the client side filters in the local machine. The client side filters come in to action only after the user downloads the emails.

Internet based spam filters

In these types of spam filters the definition for the spam are held in a remote server. When the user receives any email, that email is checked against the spam definition that is held in the remote server. If it is identified as spam then it is sent to the spam folder or trash. Otherwise it is sent to the Inbox. If the user gets any genuine mail as spam then he can mark that as 'Not Spam'. This information is updated in the definition. Any further email from that address is sent to the inbox. Similarly you can also mark some emails as 'spam'. This is also updated in the definition. Since many users are using the same spam definition the database is updated faster and newer spam are also updated faster.

There is also a negative side to this type of Internet based spam filters. Suppose a user who has subscribed to a service does not want to receive any emails from them further, there is every chance for him to mark it as spam. If he does so it will also affect others and they will also not receive email from that service provider. Spamnet and Matador are products that work on this model.

Statistical / Algorithmic spam filters

In these types of filters the software has to be trained by the user to identify the spam emails and the genuine emails. The software is trained to build a database of emails that are good and bad. Based on this database created the emails are filtered. The database is created faster and becomes larger as the incoming email traffic increases. Over time the database becomes large. The disadvantage of this type of anti spam filter is that at first the genuine emails also are marked as spam. Only as the user trains the software, the number of genuine emails that appear in the spam folder decreases. Bayesian sorting is one statistical method that is used in spam filtering. In this method a statistical analysis is done on the words that appear in a spam email based on which the emails are filtered. Thunderbird and Spammunition are examples of such anti spam filters.

Proxy Spam filtering

In this type of filtering the proxy sits between the email client and the mail server. All the emails to the email client are passed through the proxy. The spam definitions reside in the proxy. In this way you are not connected to the mail server directly, instead through the proxy. To setup a proxy the email accounts should be POP3 accounts. This is one of the main disadvantages of the proxy spam filtering. POPFile is one such system that is used for proxy spam filtering. There are many products available as proxy spam filtering. You can perform a search on the Internet to get this software.

Anti spam filtering in free email services

Anti spam filters are provided as a part of the free email services that are provided by Hotmail, Yahoo, Gmail, and a host of other free email service providers. All you have to do is to make the Spam filter ON in these accounts for the spam filter to be in action. You can also train the spam filter to identify the spam and genuine emails.


Copyright © 2002 - 2007 Bean Software. All rights reserved.