Why are emails going into the Spam folder? How spam is detected?
To protect users against SPAM emailsystems use so called spam filters (anti-spam techniques embedded in the software). Inbox.lt spam filters work upon 2 basic filtration methods:
1. Analysis of the letter content
Detection whether the letter is spam or not is based on the content of an email. This method works very well if the advertising information is presented in the form of a plain text or HTML. After training on the big enough training set (so called machine learning technique), it is possible to filter up to 95—97 % of spam. However there also exist methods to avoid filtering by Spam filters. Spam letter usually contain any random text, but advertising information is presented as an image attached. Such random text prevents filtering and possibility to train filtering software. Therefore users are offered to mark their incoming emails as spam or non-spam. According to these judgments marked letters are analyzed and then used both for filtering and training.
2. Analysis of the sender
The so called lists of black IP addresses are used for spam filtering and blocking. The requests whether an IP address is in the black list are done via DNS. Therefore such lists are also known as DNS-based Blacklists, or DNSBLs. Unfortunately the given method is not always effective as spammers find new victims faster, than they are published in the lists. Thus more often the method of greylisting is used. Letters are temporarily rejected from unknown sender mail with the assumption that spam programs will not re-send letters to an address that are already exhibited a problem.
The main idea of grey lists is the considering of all unknown SMTP as „grey”. Letters form such mail servers are temporarily rejected with a SMPT error code. If the sender server retries to send an email after some period it became whitelisted. Therefore legitimate emails do not get lost but users will only experience a delay in delivery. For now this method allows to reject up to 90% of spam without the risk of losing important emails. However, also this method has disadvantages.
In the case of repeating illegal mailing of advertising to Inbox.lt users, the administration reserves the right to block domains, sender addresses, IP addresses and keywords without the prior notification.
Unfortunately one of the problems using automatic filtering methods is the detecting legitimate emails as spam and not rejecting all spam. Only the recipient of the letter can determine whether the email is needed or not with 100% accuracy.
Spam sending is profitable business with huge incomes. Persons engaged in spam sending constantly improve technologies of data gathering and delivery of spam emails. A large quantity of companies in the world is developing the software for decreasing the volumes of spam, but it’s only struggle against consequences. It means that users, despite hostility to spam, use services advertised via it.
As long as the returns from spam will exceed the cost of protection against it, spam will not disappear.
In the case of detecting the most amounts of letters from the given domain as Spam by Inbox.lt, it’s advised to send request to the abuse@Inbox.lt to find out the reason and necessary actions to solve the problem.
If you were unable to find answer for your question, please write to firstname.lastname@example.org