Google is changing the way it pulls results from 'strings' into 'things," Mashable explains.
With Google's new Knowledge Graph, the company is reprogramming its search engine to understand that, say, the terms "New" and "York" aren't just two strings of information, but rather an actual "thing," as in a state in the U.S. surrounded by other states and the Atlantic Ocean, with a history, specific sports teams, places of interest, etc.
In simpler terms, Knowledge Graph works with the user to determine what he or she meant by a typed search entity. To use another example, instead of showing a clutter of search results for the word "kings," Google will now show a new window on the right side of the results, asking users which thing they meant to look for (i.e. the sports team, the TV show or actual royalty). Once determined, Google will filter its results for that specific thing.
In addition, Google will provide short blurbs about the search entity, while also offering suggestions for related things. This is similar to Amazon.com's feature of using a user's purchase to display other items bought by people who also showed interest in that item.
"It's a skeleton of knowledge that allows you to explore information on the web," Google fellow Ben Gomes told the news source. "The more you dig into things, the more things you learn about."
To power Knowledge Graph, Google will be pulling information from Wikipedia, Google Local, Google Maps, Google Shopping and Freebase, "a community-built knowledge base packed with some 12 million canonical entities," the media outlet notes in a separate article.
Knowledge Graph's effect on SEO is still unknown, but it's likely to create some change since the feature will push aside keyword results in favor of relationships.
Google's new feature is also intended to get smarter over time, depending on how a specific user searches. The media outlet predicts the same technology could be useful in robotics, enhancing robot-human interaction to the point where bots with access to a Knowledge Graph-type search engine will be able to understand that a baby is, say, small, fragile and hungry.
Knowledge Graph is scheduled to roll out within the next week in the U.S., but will eventually be available globally.