Google’s Knowledge Graph: Smarter Search Results

Google has introduced a new, smarter way of handling search: easy access to answers through the Knowledge Graph. “We’ve always believed that the perfect search engine should understand exactly what you mean and give you back exactly what you want,” Google SVP Amit Singhal posted over at Google’s blog yesterday. “And we can now sometimes help answer your next best question before you’ve asked it.”

Examining the Knowledge Graph

The Knowledge Graph is essentially Google’s new way of showing search results. It’s a more intelligent manner of handling search, you can say, as it attempts to take in the context of keyword searches to produce quicker results that anticipate the next possible queries. Rather than simply showing a list of results, Google will now display graphical features too – maps, photos, and the like – through pop-ups that display related results. This means faster, broader but deeper searches that take you to the heart of the matter in almost no time at all.

Of course, this search-redefining move is nothing new. Microsoft already released the new Bing last week, which features web results, snapshots and a separate sidebar that accommodates social search, all in one page. And while Google does away with the 3-column design – the new web results page will mainly feature an additional panel next to the list results – the idea is essentially the same. This new way of handling online search aims to provide users with knowledge, and not just information. This means users can get the answers they need, obtain a summary of information and discover more in the process.

With 66% of the U.S market preferring Google over any other search engine, Google is still on the lead. The Knowledge Graph has been released to U.S. based users and a global roll out is expected to follow.

“The Knowledge Graph is about collecting information in the real world,” Google Technical Lead Shashi Thakur says, “As we grow the Knowledge Graph, and make it larger and richer, we are really excited at the opportunity we have to understand more of the user’s queries, to understand more about the information that’s out there on the web, and intelligently connect the two to each other.”

How it works : 3 main points

1. Understanding the meaning of a same word and showing the right result. Example : Taj Mahal, do you mean the monument or the singer ?

2. Displaying a Summary : a short biography and description in the top right column. Example with the biography of Marie Curie : mini-bio from wikipedia, birth date, husband, children, education, employment and inventions. Below the summary, links (photos) to related people that can be of interest (here some scientists).

3. Predictive Results. In an attempt to add an additional layer of information, Google can also predict answers to questions users might have. For example, where did Matt Groening got the idea of the names of Marge, Lisa and Omer Simpson ? The answer is given by the search engine: in his family.