We recently did a blogpost aboutGoogle’s new Algorithm – Rankbrain – which detailed, to understand user queries and also the intent behind such queries to produce featured snippets. how Google is now using machine learning

In another recent article published on Wired, Cade Metz explains the way ‘Sentence Compression Algorithm’ is being used by Google as of this week, although the use of the algorithm to power featured snippets is limited to desktop search results only for now.

Therefore, Google’s Hand-fed AI is now providing featured snippets, and not just search results.

This algorithm is Google’s way of extracting the best answer that fit well with your query to be displayed as a featured snippet

(h1) Here’s how it works

(H2) Integer imperdiet, lacus eu blandit congue.

Ask the Google Search app: “Which is the fastest bird on Earth?” “Peregrine falcon”, Google will tell you. “According to YouTube, the peregrine falcon has a maximum recorded airspeed of 389 kilometers per hour. That is indeed the correct answer to that question, and you would expect it to be from some online database.

You’d be surprised to know that Google didn’t fetch that data from some pre-fed database, but rather from a ‘Video’ published on YouTube. Google sourced this information from a Video on YouTube that lists the five fastest birds on Earth and extracted just the information you were looking for, without displaying the other four birds.

Google answers these questions with the help of deep neural networks, a form of artificial intelligence that is rapidly transforming not just the search engine, but the whole of Google and along with it, the other giants of the internet like Facebook and Microsoft.

And of course, its use is not limited to featured snippets, but extends to Google Home, Google Assistant and more.

(H1) Here is a snippet to get you to the core of the matter, from the article on Wired

(H2) Pellentesque sit amet facilisis sem. Vestibulum sagittis tempus velit sit amet consequat.

Google answers these questions with the help from deep neural networks, a form of artificial intelligence rapidly remaking not just Google’s search engine but the entire company and, well, the other giants of the internet, from Facebook to Microsoft. Deep neutral nets are pattern recognition systems that can learn to perform specific tasks by analyzing vast amounts of data. In this case, they’ve learned to take a long sentence or paragraph from a relevant page on the web and extract the upshot—the information you’re looking for.

These “sentence compression algorithms” just went live on the desktop incarnation of the search engine. They handle a task that’s pretty simple for humans but has traditionally been quite difficult for machines. They show how deep learning is advancing the art of natural language understanding, the ability to understand and respond to natural human speech. “You need to use neural networks—or at least that is the only way we have found to do it,” Google research product manager David Orr says of the company’s sentence compression work. “We have to use all of the most advanced technology we have.”

(H1) Lorem ipsum dolor sit amet, consectetur adipiscing

• (H3) Ut bibendum : Aliquam molestie, nulla sit amet imperdiet rhoncus, purus turpis venenatis nulla, sit amet lacinia dui ipsum non mauris. Sed diam lacus, commodo vel laoreet a, lacinia nec eros. Proin tempus porta nisl, sit amet interdum orci consectetur nec.
• (H3) Morbi ultricies dignissim : Donec accumsan augue turpis, nec finibus risus tempor eget. Sed viverra luctus metus id maximus. Cras elementum eleifend tempus.
• (H3) Sed dictum cursus purus quis : Proin posuere eros sit amet est venenatis tempor. Curabitur venenatis enim venenatis feugiat porttitor. Vivamus convallis quam diam. Integer ut diam orci. Nulla facilisi. Proin non lectus a dolor cursus convallis accumsan in justo. Mauris fringilla nisi magna. Nam rutrum, magna vulputate faucibus blandit, sem urna malesuada augue, eget imperdiet nibh diam eu lectus. In erat quam, volutpat sed dolor et, molestie dictum leo. Maecenas feugiat orci eget rutrum ullamcorper.

LocalBusiness schema looks like this:

				
					// Logs "Hello World!"
					// in the browser's developer console
					console.log("Hello World!");
				
			

1. <!DOCTYPE html>  
2. <html lang="en">  
3.     <head>  
4.         <meta charset="utf-8" />  
5.         <title>Title</title>  
6.         <!-- styles -->  
7.     </head>  
8.     <body>  
9.         <div id="container">  
10.             <header role="banner">  
11.                 <h1>  
12.                     <a href="/">joel sutherland</a>  
13.                 </h1>  
14.             </header>  
15.             <nav>  
16.                 <ul>  
17.                     <!-- nav items -->  
18.                 </ul>  
19.             </nav>  
20.             <div id="blog" class="hidden"><-- blog dropdown -->  
21.             </div>  
22.             <!-- blog -->  
23.             <article>  
24.                 <header>  
25.                     <h1>  
26.                         <a href="">Post Title</a>  
27.                     </h1>  
28.                     <div class="postmeta">  
29.                         <time datetime="">December 31st, 1969</time>  
30.                         <a href="y#comments" class="comments">0 Comments</a>  
31.                     </div>  
32.                 </header>  
33.                 <!-- content -->  
34.                 <aside class="comments" id="comments">  
35.                     <!-- comments -->  
36.                 </aside>  
37.                 <!-- comments -->  
38.             </article>  
39.         </div>  
40.         <!-- container -->  
41.         <!-- scripts -->  
42.     </body>  
43. </html>  

These “sentence compression algorithms” just went live on the desktop incarnation of the search engine. They handle a task that’s pretty simple for humans but has traditionally been quite difficult for machines. They show how deep learning is advancing the art of natural language understanding, the ability to understand and respond to natural human speech. “You need to use neural networks—or at least that is the only way we have found to do it,” Google research product manager David Orr says of the company’s sentence compression work. “We have to use all of the most advanced technology we have.”

(H1) Mauris fermentum placerat efficitur

1. (H2) Donec et suscipit nisi

   Sed vitae ipsum in est tempus euismod. Nunc fermentum lacus quis purus vestibulum, eget suscipit dui vehicula. Integer risus enim, gravida at velit eget, euismod tempor velit. Pellentesque et libero ac ante lacinia convallis. Duis placerat lacinia ex, et ultricies ante blandit nec. Ut quis risus massa.

2. (H2) Phasellus tincidunt ipsum

   Integer lobortis vel metus maximus tristique. Etiam mauris mi, porta sit amet maximus ultrices, facilisis vel ipsum. Sed nec felis sem. Vestibulum at lorem in turpis lacinia ornare at sed magna. Ut sit amet metus sit amet neque tincidunt vulputate sit amet eu ante. Donec sagittis ultricies enim eu venenatis. Vestibulum bibendum aliquet dignissim.

3. (H2) Nunc vitae ligula ultricies, pulvinar leo

   Pellentesque posuere tristique lacus at consequat. Phasellus massa lacus, maximus at aliquam in, consectetur sit amet lacus. Duis ac porta ipsum. Mauris ullamcorper ligula eleifend orci viverra, pulvinar rutrum leo convallis. Aenean tincidunt metus in finibus pretium. Cras varius, est sed eleifend tempus, justo ipsum cursus erat, id consectetur leo ligula at purus.

(H1) Don't forget to share this post!