Spam/en

De FdIwiki ELP
Saltar a: navegación, buscar

Spam (emails)

Spam messages are those emails we recieve without our permission. In other words, unwanted messages from unknown. In general, the purpose of spam is advertising; it is sent in large quantities to different users, or specific user. There are also spam with massive purposes, such as harming companies' servers causing blockages with the vast number of e-waste, or embedding viruses in email content. Therefore, when the user opens the mail, the viruses attack his system.


Filtering spam

Spam cannot be completely eradicated, because some of these emails may be important to part of users and unwanted for others. However, there are mechanisms that help us filter spam classified in the following types:

According to the message content:

These filters analyze the content of the emails to determine whether the content is suspect or not. They look for specific words (spamwords) or words that hide spamwords. They also analyze the links pointing to other websites.

According to the message source:

The essence of this filter is the reputation of the issuer. This reputation is one or several scores associated with the identification of the sender, which is determined by the sending IP address and domain name used.

Based on SMTP

This filter uses the SMTP technique, which is a set of rules that describe how to send and receive an email, using this method we get correctly identify the sender and thus ensure verification.

Based on the user

This filter requires manual user intervention. Users add the IP addresses and domain names suspects to the blacklist. In this way the filter identifies them as spam.


Spam detection of with I.A. (machine learning)

E-mails are labeled as "spam" or "legitimate" by users. The prediction process begins with an analysis of what characteristics or patterns have the emails already marked with one or other label. It can be determine, for example, that a spam email is one that comes from certain IP addresses, has a certain relation text/images, contains certain words, or there is no one in the "To:" field, Etc.

This would be just one of the patterns. Once all the patterns are determined (this phase is called "learning"), new emails that have never been marked as spam or legitimate are compared with the patterns and classified (predicted) as "spam" or "legitimate" in function of its characteristics.

External Links