So, let’s say you’re doing a Google search for “Kings.” Did you mean the L.A. hockey team or the Sacramento basketball team? Maybe a TV show? Or maybe you actually wanted to know something about monarchs.
Google on Wednesday announced Knowledge Graph, a significant change to how search results are delivered that the company believes will make their search engine think more like a human.
“The web pages we [currently] return for the search ‘kings,’ they’re all good,” Jack Menzel, director of product management at Google, told CNN in an interview. “You, as a human, associate those words with their real-world meaning but, for a computer, they’re just a random string of characters.”
With Knowledge Graph, which will begin rolling out to some users immediately, results will be arranged according to categories with which the search term has been associated. So, in the above example, boxes will appear with separate results for the hockey team, basketball team and TV show.
The user can then click on one of those boxes to only get results for the specific topic they were searching.
“It hones your search results right in on the task that you’re after,” Menzel said.
More specific searches, say for the name of a celebrity, will render boxes with basic information, as well as links to what Google believes are possibly related searches.
Menzel says the initial version of Knowledge Graph has information on 500 million people, places and things and uses 3.5 billion defining attributes and connections to create categories for them.
The feature will begin rolling out as early as Wednesday afternoon for some users in the United States and eventually be available on desktop, mobile and tablet searches. It will first become available in English, then in other languages, Menzel said.