Dear Visitor,

Our system has found that you are using an ad-blocking browser add-on.

We just wanted to let you know that our site content is, of course, available to you absolutely free of charge.

Our ads are the only way we have to be able to bring you the latest high-quality content, which is written by professional journalists, with the help of editors, graphic designers, and our site production and I.T. staff, as well as many other talented people who work around the clock for this site.

So, we ask you to add this site to your Ad Blocker’s "white list" or to simply disable your Ad Blocker while visiting this site.

Continue on this site freely
You are here: Home / Mobile Apps / Google Responds to E-Mails for You
Google's New Feature Will Respond to E-Mails for You
Google's New Feature Will Respond to E-Mails for You
By Jef Cozza / CRM Daily Like this on Facebook Tweet this Link thison Linkedin Link this on Google Plus
If you've ever wished your phone could respond to simple e-mails with automatic replies, Google might have just the thing for you. A new feature called Smart Reply is debuting on Gmail’s smartphone client Inbox that can read an e-mail and formulate a variety of automatic responses on your behalf.

The new functionality, debuting later this week, will suggest up to three different responses based on each particular e-mail. Google said the feature will allow users to issue simple responses with just one click. Smart Reply can help users get started writing longer, more complicated responses as well.

Yet Another Predictive App

The company is expecting that the new functionality will make responding to e-mail messages on your smartphone less of a pain. Like other predictive applications launched in recent years, such as Siri, Google Now, or Cortana, the new functionality seems geared to turning your smartphone into something closer to a full-fledged automatic assistant.

Smart Reply is built using the same deep neural network technology Google used to improve its voice search capabilities and automatically generate single frames for YouTube thumbnails to help people find videos more easily. Smart Reply’s deep neural network is also capable of machine learning, meaning it can improve its ability to write e-mails the more you use it.

“There’s a long history of research on both understanding and generating natural language for applications like machine translation,” said Greg Corrado, senior research scientist at Google, in a blog post about the technology. Last year Google engineers began working on combining the two tasks into a process known as sequence-to-sequence learning, he said. “This end-to-end approach has many possible applications, but one of the most unexpected that we’ve experimented with is conversational synthesis,” Corrado added.

Inbox will also use machine learning to recognize e-mails that need responses and generate natural language responses on the fly. It works much like the way the Inbox learns to recognize spam based on what e-mails a user tags over time. Similarly, Smart Reply will learn what responses are most appropriate based on which responses the user chooses.

Machine Learning for the Future

The Smart Reply debut may also determine whether Google can live up to the promise of providing a feature capable of reading natural language text and composing natural language responses.

The ability to recognize the semantic meaning embedded in human speech and respond convincingly is an ambitious goal that has challenged a number of developers, thanks in part to the infinite number of ways that people communicate with each other.

But if Google's deep neural network is able to make Smart Reply work as advertised, it could potentially open the door to broader technologies, such as better voice commands and natural language interfaces. Currently only available in English, Smart Reply will be rolling out on Google Play and the App Store.

Tell Us What You Think


Like Us on FacebookFollow Us on Twitter

© Copyright 2018 NewsFactor Network. All rights reserved. Member of Accuserve Ad Network.