|
Aside from using keywords, what other evidence is taken into consideration in the algorithm that detects whether an email sent is a spam email and not an legit one? Like if I sent an email to my friend warning him not to say, open up any emails about viagra - how would the filter know that my email is not spam. |
||||
|
|
|
Because you would be in the list of frequent emailers probably. But it can also send it to the spam bucket. These algorithms to detect spam tend to use naive bayes algorithm, that builds a model with which they evaluate all the emails given their word content among other things |
|||
|
|
|
Spam can be detected by identifying the sender, using username, email address or IP address. Stop Forum Spam is one such service. Another common technique that is often used by blogs and forums is to take links into account. Wordpress has an option to automatically mark comments as spam if they contain more than one link. I am pretty sure there are more advanced strategies, but if you and your friend have been communicating via emails before, a decent spam filter should "know" that you two are friends, before attempting to filter the message based on keywords. |
|||
|
|
