Google has been successful in lowering the number spam emails users receive, but in the process, every now and then bulk emailed messages like statements and ticket receipts get mis-classified. This is an immense problem for large emailers. So to deal with it, Google has come up with new tools to help users in identifying what’s happening to their emails.
Using artificial intelligence
Google has added an artificial neural network to Gmail’s spam filter to detect and block “the especially sneaky spam,” Google informs users. Artificial intelligence is in addition to the machine leaning the internet firm is already using for Gmail’s filters.
Below is our 13F roundup for some high profile hedge funds for the three months to the end of March 2021 (Q1). Q1 2021 hedge fund letters, conferences and more The statements only include equity positions as 13Fs do not include cash and debt holdings. They also only include US equity holdings. Funds may hold Read More
To fight spam, Google said it is using the same artificial intelligence it developed for search. The artificial neural network uses self-training algorithms that can identify tricky spam which looks similar to normal messages. It can also catch spoofed messages which seem to come from a genuine sender but actually are not.
Google informs users that the amount of required mail moving to the spam folder is lower at less than 0.05%, while less than 0.1% of email in the average Gmail inbox is spam. Still, Google wants to do better, and its Gmail team aims to give users “the mail you want, not the spam you don’t.” So instead of blocking emails from the companies, Google is working with senders of wanted email such as ticket receipts from airlines, account statements, etc., and helping them to analyze their emails.
Sri Harsha Somanchi, a Google product manager, wrote in a blog post that their postmaster tool is designed to give emailers details on delivery errors and spam reports.
Google making it more customizable
In addition, the internet firm is working to make its spam filter more customizable for consumer accounts. While users can already categorize messages by labeling them as spam, what is spam for one may be a relevant email for another.
“So while your neighbor may love weekly email newsletters, you may loathe them,” Somanchi said, adding that now with “with advances in machine learning,” the spam filter can now be customized for these individual choices.
Chinese search giant Baidu uses neural networking to offer lucrative ads, while Facebook uses it to guess what a user likes on their News Feed. Google is also an avid user of neural networking technology. In recent times, the internet firm has used the technology in many of its services, including recognizing commands given to an Android phone, recognizing photos posted on Google Photos, and more.