In a Google Cloud blog post published on February 6, Google announced that it has been using an AI platform to further its spam-blocking endeavors on Gmail and has witnessed significant results.
The platform, dubbed TensorFlow, was developed by Google and is “an open-source machine learning (ML) framework.” TensorFlow was launched and open-sourced in 2015. According to Google, TensorFlow is allowing them to block 100 million more spam messages from reaching the primary inboxes of Gmail users every day. This is in addition to the 99.9% of spam messages Google already claims Gmail blocks.
Google is apparently able to do this identifying even the most complex forms of spam mail such as emails from newly created domains, image-based messages and even messages with hidden embedded content.
Google says integrating TensorFlow into Gmail will also allow it to better personalize spam filters. This process has been taking place for years, with Gmail looking for certain signals from users about what they judge to be spam, but TensorFlow is turning those signals into better results.
Since it is a machine learning based platform, TensorFlow continues to improve with time. Google also mentioned in the blog post that the platform is intended to also help Gmail customize its spam protections for each individual user’s needs.
“At the scale, we’re operating at, an additional 100 million is not easy to come by,” Neil Kumaran, Product Manager of Counter Abuse Technology at Google, told The Verge. “Getting the last bit of incremental spam is increasingly hard, TensorFlow has been great for closing that gap.”
Algorithms trained in this way balance a huge number of metrics, everything from the formatting of an email to the time of day it’s sent. “TensorFlow,” Kumaran added, “makes managing this data at scale easier, while the open-source nature of framework means new research from the community can be quickly integrated.”
“There’s no one definition of spam out there,” he adds. But AI could help work out the best definition for you.
