Google is setting the stage for its biggest renovation in its history, according to a report published in The Wall Street Journal. Instead of performing a matching of keywords typed into a search box, the search giant intends to incorporate semantic knowledge into its algorithms in order to better understand the context of a question--and deliver more relevant search results.
One particular weakness of traditional search engines is their inability to differentiate between disparate topics. For example, the word "jaguar" could mean an animal or a car, explained Danny Hillis to the WSJ. Hillis is an artificial intelligence pioneer who co-founded Metaweb, a semantic search company that Google acquired in 2010.
On this front, Google (NASDAQ: GOOG) has apparently been quietly amassing a database populated with hundreds of millions of "entities" which will be matched with search queries in order to generate better results.
These changes are expected to be rolled out over the next few months, though Google has no specific timeline to complete the full makeover to the "next generation of search," reportsCNET. Ten to 20 percent of all search queries could be impacted by this change.
On the other hand, Danny Sullivan from Search Engine Land is convinced that the reports are just so much spin for the purpose of public relations. Not that the technologies in question are false; rather, Sullivan argued in a detailed blog post that Google has never stopped tweaking its algorithms.
Moreover, he says Google has been working on semantic search since 2003, and that it has been implemented in various Google services such as Google Q&A and even the main Google search engine. "You're unlikely to see a massive change to how Google search looks and operates in the near term," maintains Sullivan.