Google calls Knowledge Graph first baby step towards ‘Star Trek computer’
Google has just launched Knowledge Graph, the latest refinement to its search engine product that seeks to provide users with more relevant and in-depth responses to search queries. The company actually started testing this new interface last week, but now its ready to take the wraps off its new method for connecting search queries to information-rich topics on people, places, or things. Along with the standard search results you're used to seeing, Google's search results page now displays instant results related to your queries — a search for Taj Mahal immediately brings up a list of facts, photos, and a map of the famous landmark, as well as quick links to other popular uses of the search term (like the musician or the casino in New Jersey). There are a multitude of sources behind this data — Google cites Freebase, Wikipedia, and the
CIA World Factbook, but also notes that "it's augmented at a much larger scale" and tuned based on what the average user searches for.
CIA World Factbook, but also notes that "it's augmented at a much larger scale" and tuned based on what the average user searches for.
Google's goal is to get you to the information you're looking for in fewer clicks, while also increasing the relevancy of what you see when searching, and there are three main innovations the company is highlighting. Knowledge Graph results seek to remove the ambiguity from typical search results by presenting different segments of results with one click — so if you're looking for Taj Mahal, the musician (and not the landmark), you can just click over instantly to tell Google which segment of search results you're looking for.
There's also new summary info, which might keep a lot of people from having to click through to Wikipedia. Google gave the example of Marie Curie — when doing a search for the scientist, Knowledge Graph brings up a photo, birth and death dates, and a list of her major discoveries and education. It doesn't have the in-depth information WIkipedia has, but it may save a few clicks when you're just searching for a quick bit of information.
Lastly, Knowledge Graph seeks to present information that users commonly look forafter making a search, letting users easily drill down into a subject to find deeper information. For example, Google notes that the information Knowledge Graph shows when searching for Tom Cruise answers 37 percent of the next search queries made about him. Google also shows other common searches that other people made when searching for your term — kind of like Amazon's "other people who shopped for "X" also liked..." features.
Google's calling this the first "baby step" towards a "Star Trek Computer" future, but it sounds like one of the bigger innovations the company has rolled out to its core product in some time (though it reminds us a lot of what Wolfram Alpha is trying to do, as well). For those who search on-the-go, Google's also gradually rolling this out to Android 2.2 and higher and iOS 4, both in the browser and in dedicated search apps. Unfortunately, we're not yet seeing the Knowledge Engine live — Google says this new feature is gradually rolling out to US English users, with no word on when exactly it'll make its way to other countries.
No comments:
Post a Comment