On the heels of major revisions to Microsoft's Bing search engine, Google is revamping its own. On Wednesday, the tech giant announced the launch of its Knowledge Graph, which is intended to help users quickly and easily discover new information.
In a posting on the Google Official Blog, Senior Vice President of Engineering Amit Singhal wrote that, instead of primarily focusing on matching keywords to queries, the enhancement enables the search engine to use an intelligent model that "understands real-world entities and their relationships to one another: things, not strings."
'Critical First Step'
Singhal said the Knowledge Graph "knows about" a variety of things, people, and places, such as landmarks, celebrities, cities, sports teams, buildings, geographical features, movies, celestial objects, works of art, and other subjects. The Graph's current inventory of knowledge, he said, is only the "critical first step" toward creating the next generation of search, which understands the world in ways closer to how people do.
The Graph is more than just calling up data in Wikipedia, the CIA World Factbook, and other supplies of knowledge. It's been populated with more than 500 million knowledge objects, with more than 3.5 billion facts about the relationships between those objects.
The first step in this new kind of search, Google said, is understanding the differences in meaning for a given query. For instance, is the search for "Taj Mahal" about the monument or the musician? The Graph will give choices.
Next, the Graph provides summaries containing key facts that a user might want about a particular subject. The example given by Singhal is Marie Curie. The Graph will deliver birth and death dates, as well as information on her education and scientific discoveries. There's also knowledge about her relationship with other entities, such as her Nobel-prize-winning relatives.
'People Also Search for'
The Graph's ability to determine what is the most useful information for a given subject is based in part on Google's learning from other searches. As an example, the volume of searches about books by Charles Dickens indicates that books are an important part of his info.
With this cloud of useful information about a topic, Google said that users will make serendipitous discoveries, such as through a "People also search for" feature.
There's also a version of the Knowledge Graph specifically for smartphones and tablets. It distills the same kind of information that the desktop/laptop version does, with drill-downs to additional relevant information through tapping or swiping.
In a separate posting on Google's Inside Search blog, Engineering Manager Junyoung Lee said the Graph provides an "interactive ribbon" at the top of the mobile device, with the alternative choices for different meanings of a search term, such as Andromeda. Is it a galaxy, a TV series, or a Swedish band? The user can tap on one of those choices, thus reducing the number of keystrokes to get at the desired results.
Posted: 2012-05-17 @ 10:04am PT
I always knew that Google would eventually become a knowledge engine. Nevertheless, my mouth has been open during this video introduction:
Posted: 2012-05-16 @ 10:57pm PT
stop talking about it....lets see it.