Gmail has announced that it is now able to detect 100 million more spam emails every day. This has been possible with the recent integration of TensorFlow that manages to catch spammers who slip through Gmail’s traditional scanning techniques. Implementation of TensorFlow has helped Gmail block image-based messages, emails with hidden embedded content, and messages from newly created domains that try to hide a low volume of spam messages within legitimate traffic.
TensorFlow is an open-source machine learning (ML) framework developed at Google, and was launched in 2015. It has, over the years, risen to become a popular Google service, and is used by many developers. Google says that for Gmail, recognising spam and blocking has been a challenging task given the varied forms and it’s constant evolving and incremental nature as well. The company says that it already blocks 99.9 percent of spam emails through its implemented machine learning and rule-based processes, and TensorFlow helps in filling that 0.1 percent gap.
Also, it’s open-source nature will allow easy implementation of any fresh community-based research for better fighting against spam. All in all, TensorFlow allows Gmail to scale its ML efforts, requiring fewer engineers to run more experiments and protect users more effectively. ‘Within Gmail, we’re currently experimenting TensorFlow in other security-related areas, such as phishing and malware, as part of our continuous efforts to keep users safe,’ Google explains on its blog.
“At the scale we’re operating at, an additional 100 million is not easy to come by. Getting the last bit of incremental spam is increasingly hard, [but] TensorFlow has been great for closing that gap,” Neil Kumaran, product manager of Counter Abuse Technology at Google, tells The Verge. Kumaran also says that TensorFlow will help Gmail personalise spam filters, learn from user patterns on what they judge as spam, and provide better customised results. This means that spam results will vary with each user, based on his browsing and interests.