Google has said that it needs to transition from a search engine to a “knowledge engine.” This is essentially the same thing Bing says it wants to do. I think Google is closer to the goal than Bing, but Google got a head start.
A knowledge engine implies a destination whereas a “search” engine implies a conduit, a place you pass through to get to your destination. That’s an important distinction.
So how does Google plan to transition itself into a knowledge engine? The key to that transition is what they are calling the Knowledge Graph.
The Knowledge Graph is essentially a set of semantic search protocols based on semantic markup. This semantic markup is code you can add to your website to communicate with Google and Bing on what specific types of information exists in each section of your website. But it goes beyond that even.
Where traditional search was based on keywords being matched to queries made by people searching for information, semantic search relies on matching synonyms and concepts with search queries.
Semantic search attempts to eliminate the guesswork by tossing out non-relevant queries with matching keywords. Instead, it tries to judge the searcher’s intent based on previous search data, current session clicks, and other information. Google, and Bing to some extent, have already started retrieving information on these bases. It’s just a matter of time before they perfect it.