[H]ardOCP: Google Uses TensorFlow to Fight Spam Emails

Since then, "teams and researchers all over the world" have produced a whopping 71,000 forks of the public code and other open-source contributions, establishing a strong community that makes it easy to quickly apply "new research and ideas".

TensorFlow, which Google made open source back in 2015, has been used to improve spam detection for a while, but it's now received broader adoption and a larger set of spam classifications "at scale", said Neil Kumaran, product manager at Google's Counter Abuse Technology division, speaking to VentureBeat. With the new filters, Google claims Gmail is now blocking an extra 100 million spam messages every day.

TensorFlow has been added alongside existing Artificial Intelligence (AI) and rule-based filters that Gmail has utilised for years. Moreover, Google, Intel, SAP, Airbnb, and Qualcomm are the top users of this software.

Google LLC is beefing up Gmail's anti-spam capabilities with new protections, powered by its machine learning software framework TensorFlow, that are created to complement its existing algorithms.

100 MILLION spam messages blocked out every day sounds like a staggeringly high figure, but considering that there are more than 1.5 billion people using Gmail every month, it works out to only ONE extra blocked spam message per 10 users.

Before implementing TensorFlow-powered protection features, Google's effectiveness in blocking spam, phishing, and malware attempts was pretty great, but now it's essentially flawless. Thus, false positives are becoming rarer and rarer and spam categories that used to be particularly tricky to detect, like emails with hidden embedded content, are slowly vanishing.

This does not dismiss the achievement of TensorFlow, though, as the blocking of the additional nuisance emails suggests that Google's spam-blocking functionality has been enhanced through machine learning.

"Getting the last bit of incremental spam is increasingly hard", he points out. Hence, integrating TensorFlow with the existing Gmail framework has allowed the latter to personalize the spam filters more effectively.

What kind of additional spam messages did Google now get better at deflecting?



Other news