Source: Times of India
NEW DELHI: Tech giant Google is doing everything possible to get rid of spam mails. The company has now taken help of in-house machine learning system. Called TensorFlow, the tool will help filter additional spam for Gmail users. The company started using the new filters last month and claims that it has managed to block extra 100 million spam messages every day with its help.
TensorFlow was initially developed for Google’s internal use. The company released the machine learning framework under the Apache 2.0 license three years ago. The company claims since then the developers from all around the globe has produced 71,000 forks of public code and other open source contributions.
As reported by The Verge, the company says incorporating TensorFlow into Gmail will enable it to personalise spam filters in a better manner. This complete process has been taking place for years where Gmail looks for some particular signals from users on the basis of which it judges the spam. However, with TensorFlow the company is now able to turn these signals into better results.
The report adds, “TensorFlow makes managing this data at scale easier, while the open-source nature of framework means new research from the community can be quickly integrated.”
Along with this, Gmail has also been using AI with the rule-based filters. The rule-based filters are capable of blocking obvious spam, while machine learning look for new patterns which reveal whether an email is trusted of spam.
Recently, it was reported that Google is planning to introduce three new features to Gmail. These three features has been borrowed from Inbox by Gmail. As reported by XDA developers, these new features will be available on Android first. The three features which will soon come to Gmail are bundle which will keep all the news letters from publications clubbed at a single place. The second feature in pin, using which users can pin all the emails from a sender for easier communication and the third feature is of reminder.