Spam is one of those things we’ve all received it and tired of it. But now Google has annouced some improvements to the Gmail spam filter. The sevice implements advanced machine learning techniques to get a better handle on some of the trickler spam items and prevent putting relevant emails in the trash.
Google uses its machine learning techniques from Search and Google Now to Gmail. Some type of machine learning has been implemented in Gmail for while already, through the use of the “report spam” and “not spam” commands, which helped to build up a profile and learn from mistakes. The spam filter now will use some artificial neural network to detect spam that tries to pass itself off as an authentic email.
It will also detect your preferences for different tyoes of email. If you don’t like newsletters, the filter will block more of those types of messages. Apparently, now Gmail is better at figuring out if emails come from their supposed sources, which should help to clamp down scams.
For learning some more about what machine learning is, check out our overview of the subject.
Addicionally, Google has launched its ‘Postmaster Tools’. This tool, allows high volume senders to analyse their emails with detailed reports on any errors, spam reports from recipients and their overall reputation based on feedback.