Web Log: Google attacks spam with machine learning

Gmail has rolled out some improvements to tackle the problem

Spam is a thoroughly modern, technological affliction that has yet to be treated successfully due to its increasing sophistication. More and more commonly, spam can appear to be from a legitimate contact, thus slipping though email filters. You may be savvy enough to avoid interacting with this kind of spam but it does clog up the inbox something awful.

Gmail has rolled out some improvements to tackle this using an artificial neural network. What this means is that machine learning can tailor spam filtering “to detect and block the especially sneaky spam”. Additionally, if you hate getting those daily newsletters, Gmail will learn from this and weed them out because what you consider spam may not be unwarranted, just unwanted.

http://iti.ms/1J8Kc1b