This change to search is among the biggest in the company's history and could affect millions of websites that rely on Google's current page-ranking results. At the same time, they could give Google more ways to serve up advertisements.
Semantic search refers to the process of understanding the actual meaning of words that people type into a Web-search box. That would enable the search engine to differentiate between the car brand "Jaguar" and the animal "jaguar" on a Web page that it may want to provide a link to in search results.
In an interview, Google search executive Amit Singhal implied the semantic-search effort will help Google compete with Siri. "When we can deliver small nuggets of information, that system is far more suited to mobile phones and searching with voice," he said. Google search will look more like "how humans understand the world".
Google also hopes the change to semantic search will entice some people to stay longer on the search site, said people briefed on the plans, amid competition with social networking sites like Facebook and Twitter Inc. that are claiming more Internet users' time.
For instance, people who search for a particular novelist like Ernest Hemingway could, under the new system, find a list of the author's books they could browse through and information pages about other related authors or books, according to people familiar with the company's plans. Presumably Google could suggest books to buy, too.
Google also currently has some other semantic-search elements, such as the ability to assess what the web collectively thinks are the most significant items associated with certain keywords. For example, a search for "30 Rock," the name of a popular TV series, will bring up a section called "Actor searches for 30 Rock" at the bottom of the search-results page. There, people can find a photo of each actor and a link to execute a new Google search for that name.
But the newest change is expected to go much further, coming as a result of Google's acquisition in 2010 start-up Metaweb Technologies, which had an index of 12 million entities, such as movies, books, companies and celebrities. By comparison, online encyclopedia Wikipedia has 3.5 million English entries, though they include more detailed information.
Some semantic-search experts said Google also may be eyeing rival Facebook, which has created a database of information about hundreds of millions of entities, including people, places, countless articles, music and things that its users share with each other.
"Google is probably more concerned about Facebook than other search engines," said Danny Hillis, who co-founded semantic-search Company Metaweb Technologies. Facebook has created a database of information about hundreds of millions of entities, including people, places, countless articles, music and things that its users share with each other. Mr. Hillis said Facebook knows the difference between many people who have the same name, like "John Smith," while Google historically has not.
Facebook in the future could become a powerful tool for people searching for all types of information, he said.
Image Credit:Techspot
