Spam is bulk mailing of messages of advertizing character which are sent to users without their consent. Technologies of an antispam help to filter unnecessary messages and give the chance to avoid contamination of an e-mail useless letters.
Principle of work of an antispam
The antispam can be used on personal computers or remote servers. The procedure of filtering is implemented by means of the special software which is installed on the computer of the user or on the e-mail server. Spam filter analyzes each letter coming to an e-mail with use of technologies of the analysis of contents and check of reputation of the sender.
The system of an antispam helps to recognize the letter which has advertizing character by the used keywords. After that the system studies the e-mail address of the sender, information which is available in the e-mail profile of service. The filter defines a number of people which sent the same message. The fact of bulk mailing most often indicates spam, and therefore existence of the additional addresses of sending lower the status of the message for the system of an antispam at once.
After the program marks the letter with a flag "Spam", it will be sent to the corresponding folder on the server where will wait for further actions of the user. If the owner of a mailbox confirms that this letter really is unnecessary, the program will delete all excess data from a system at once.
If the user considers that the message comprises information necessary to it, the program of the e-mail server will move the file to the folder for letters "Entering", and the antispam the filter will create the rule under which messages from this sender will not be classified in the future as advertizing or harmful. The user often is in addition offered to configure manually filtering of messages with creation of the list of the entrusted e-mail addresses which automatically will not move to category "Spam".
Problems antispam systems
Each software maker of clients and the e-mail servers tries to implement own algorithm which will allow to define spam most effectively. However even the most known and effective solutions do not allow to exclude completely a filter error factor – any of modern services of filtering is not capable to define with an accuracy of 100% whether the letter is unnecessary for the user. The percent of successful scanning even at the best security systems reaches about 90%. Other 10% fall on false positives of a system.