Google has begun working on natural language understanding and Now on Tap, Inbox for Gmail and Google Photos are the few consisting of it.
A very few people might argue that this year’s keynote by I/O was the flashiest by the company. If any of the readers were expecting it to make them jump out of the blimps to unveil the new Google Glass version, the event might have been a disappointed one for them then. Instead, the latest Google news reports that the company used the low-key keynote conference for announcing the evolutionary update for the mobile OS by the company, considered to be the new effort for bringing Android itself to Internet of Things and also couple of new tools for developers had been introduced so they can monetize, analyze and also advertise their apps in better ways.
There are three new products that had been the attraction, as far as the consumer side is considered, Google Photos, Android M’s Now On Tap and also the Inbox for Gmail (not announced in the keynote). To some extent, all these products are inclined towards understanding of natural language expertise for deep machine learning and also huge database from Knowledge Graph that the search engine giant has focused on in the recent years.
For instance, Google Photos has the search engine for photos which is probably the best feature for searching through the pictures and is the best which has been introduced up till now. Search for ‘flowers’, ‘trees’, ‘mountains’ or ‘birthday’ and only experience what it might find for you in all the collections it has for your photos. Photos had been the standalone version of the feature in Google+ but now it has gained some additional features along with the improvisation in search engine.
As far as the Google Now on Tap is concerned, it can understand the happenings in the app and give the user contextual information and answer the users’ questions regarding it. Google news presented the demo of this in the keynote where it expressed that it played the Skrillex song in Spotify and asked the Now on Tap, “What’s his real name?” In order to answer this, it had to understand what was happening in the app, what was meant by “his” and also go deep in the knowledge base in order to answer it correctly.
Inbox got left out at the keynote, but its update now brings the natural language understanding in action and now is relying on it for generating the automatic reminders for the user when it notices someone is asking them to do something. It is also capable of organizing the emails regarding the forthcoming trips and also compiles them together in a single group.