Gmail is finding smarter ways to keep spam out of your inbox
Google announced a number of improvements to Gmail spam filtering in a blog post on Thursday. That includes applying machine learning technology, such as artificial neural networks to make filters smarter and new tools for businesses to ensure relevant mail always makes it into inboxes.
Artificial Neural Networks are essentially computers that mimic the connections of neurons in the brain, allowing the computers to “learn.” Google uses these to identify images, making tasks likereverse image search possible. These neural networks are also used to create those trippy images you may have seen online lately.
For Gmail, the networks will be used to filter out phishing scams that can easily be mistaken for a legitimate email. You know, the ones that ask you to log in to your Facebook account — but the email comes from a website designed to steal your login credentials.
The threshold for what is considered spam varies from person to person, so the networks will adjust over time, learning which emails you might want in your inbox versus an email someone else might regard as spam.
Google is also rolling out Gmail Postmaster Tools, which will allow “qualified high-volume” senders to get better analytics on the emails they send. With the new tools, users will be able to better avoid sending messages that end up in anti-spam filters; after all, no one wants to have to check a spam folder for bank statements.
Gmail Postmaster Tools launched on Thursday, and the usage of Artificial Neural Networks began earlier this year.
In 2012 Gmail missed 1% of spam messages, according to Google. Now, it only misses 0.1% of spam and only 0.05% of email ends up in the spam folder when it shouldn’t, but the company thinks it can improve on those figures.
Have something to add to this story? Share it in the comments.