With the launch of Google Now, the company wanted to provide predictive pieces of information based on contextual alerts such as location, time of day, interests, and more. For a while, this was Google's pseudo digital personal assistant since Google Assistant wasn't around back then. This feature evolved with Google Now on Tap and some of these features have moved onto Google Assistant while others have been left behind in the Google application.
Google calls the remnants of Google Now the "Feed", and it's located in the Google application when you launch it. Sure, you can do a Google search at the top of the Google application, but it can also provide you with things you are interested in right below it. Today, the company has announced some new features to the Google application Feed feature that will begin rolling out to devices in the United States later today, with plans for other countries over the next couple of weeks.
As with most of Google's services lately, the feed is getting a hefty dose of machine learning assistance that is said to be more advanced than it was before. The goal is to still be as predictive as possible and anticipate what Google thinks you'll be interested in. Like before, these predictive cards will showcase content such as sports highlights, top news, engaging videos, new music, and stories to read, but it will also try to curate things from your area that you'll be interested in.
The new advancements are also said to gauge how interested you are in a certain topic. Using the example given, you'll see more content about photography if you're an enthusiast while seeing less content about fitness if you're just causally into it. Soon, you'll also be able to tell Google you want to follow a certain topic with a new Follow button that appears after you do a Google search. If you're interested in the news, some stories will have multiple viewpoints from a variety of sources while also giving you a way to fact check these reports.
Source: The Keyword
from xda-developers http://ift.tt/2vCvVe5
via IFTTT
Aucun commentaire:
Enregistrer un commentaire