It’s a mystery how many searches Google gets every day, but in a 2015 story by Steven Levy on Backchannel, he estimates that it’s at least 3 billion (that’s at least a trillion in a year!) and about 15% of them are completely new to Google.
In order to handle these queries, Google needs to keep improving its capabilities. In a report published by Bloomberg on October 26, the tech giant has been testing an artificial intelligence (AI) system called RankBrain to handle new queries it has never encountered before. In the last few months, Google engineers said that the new technology is responsible for handling a “very large fraction” of searches, 15% to be exact.
RankBrain works this way: when someone is searching for information that the search engine has never seen before, it makes educated guesses as to what words or phrases could have similar meaning and presents relevant results to the user. So far, the AI was able to achieve 80% query relevancy. Every time RankBrain presents the results to the user, it monitors how the user responds to the results and adjusts the filtering process accordingly. Because of this, it has become the third most important signal in Google search algorithms.
Senior research scientist Greg Corrado notes that “the other signals, they’re all based on discoveries and insights that people in information retrieval have had, but there’s no learning.” However, he was surprised with RankBrain, adding “I would describe this as having gone better than we would have expected.”
RankBrain was implemented autonomously early this year, but it is continually monitored and updated with new data.
Tell us, have you ever tried searching Google and failed in getting answers to your questions? Let us know more about it, comment below!