According to Gregg, Rankbrain "uses artificial intelligence to embed vast amounts of written language into mathematical entities -- called vectors -- that the computer can understand. If RankBrain sees a word or phrase it isn’t familiar with, the machine can make a guess as to what words or phrases might have a similar meaning and filter the result accordingly, making it more effective at handling never-before-seen search queries." The system was implemented softly over the past few months and results show that a very large fraction of the millions of queries people type have been interpreted by the AI system highlighting for the first time the role of artificial intelligence in search engines.
The deployment shows that Google are beginning to embrace systems controlled in part by artificial intelligence to look after their most valuable businesses.
So how does RankBrain work?
RankBrain works by converting written language into vectors (mathematical quantity possessing both magnitude and direction, ref. http://dictionary.reference.com/browse/vector) that can be understood by computers. When the system scan and see word(s) shapes it doesn't recognise, it can make an educated guess to match words or phrases it resembles to in order to achieve a similar meaning. This process allows to handle search queries that haven't been conducted before. This is an important breakthroug. In 2013 Google stated that 15% of searches performed each day use queries that have never been seen before. The new system allows for advanced understanding of language semantics to gradually learn about natural searches and self store new searches. The system is based on a conversation model built by Oril Vinyals and Quoc Le. It uses what is known as a sequence to sequence framework in which the model brings up the next sentence by predicting it. Depending on the context, genre and type of the questions asked, it can collect, analyse and assemble various sets of data (for example from IT or technical manual, a movie and subtitles database, book quotes and so on), to produce an informed answer or reply. (re.: http://www.dailymail.co.uk/sciencetech/article-3290075/Google-uses-artificial-intelligence-handle-searches-RankBrain-makes-educated-guesses-new-queries.html).
The machine intelligent learning process performed by RankBrain is not done in real time, but rather occurs while offline. Google feeds the AI sytem large batches of historical searches, from which RankBrain then learns to make predictions. Those results are then analysed and tested; if they prove accurate, the new search replies go live. The semantic relationship between words is more efficient, as the system returns phrases with the same meaning as the keywords used in the search even though there isnt an existing exact match to the keywords themselves.
EXAMPLES OF Q&A POSED TO THE GOOGLE RANKBRAIN
Below is the result of the RankBrain Model Conversation that was conducted by research to analyse the system replies. The test include everyday converation questions but also more advanced, obscure and complexe ones (Quiz, Ethics, life etc.), once the system had been trained. The results are mixed
Philosophical Q&A Human: What is the purpose of life? Machine: To serve the greater good Human: What is the purpose of living? Machine: To live forever Human: What is the purpose of existence? Machine: To find out what happens when we get to the planet Earth Human: Where are you now? Machine: I'm in the middle of nowhere Human: What is the purpose of dying? Machine: To have a life Human: What is the purpose of being intelligent? Machine: To find out what it is Human: What is the purpose of emotions? Machine: I don't know General knowledge Q&A Human: Who is Skywalker? Machine: He is a hero Human: Who is Bill Clinton? Machine: He's a billionaire Human: Is sky blue or black? Machine: Blue Human: Does a cat have a tail? Machine: Yes Human: Does a cat have a wing? Machine: No Human: Can a cat fly? Machine: No Human: How many legs does a cat have? Machine: Four |
Human: How many legs does a centipede have?
Machine: Eight Human: What is the colour of water? Machine: Water Human: What is the color of blood? Machine: It is the same as a black eye Human: What is the usual colour of a leaf? Machine: It is a green one Human: What is the colour of a yellow car? Machine: Yellow Human: How much is two plus two? Machine: Four Human: How much is 10 minus two? Machine: 72 Morality Q&A Human: What is moral? Machine: What empowered humanity ; background: what intellectual the essence is Human: What is immoral? Machine: The fact that you have a child. Human: What is morality? Machine: What is altruism? Human: What is the definition of altruism? Machine: If you don't believe in God then you don't know. Human: Ok so what is the definition of morality? Machine: Well "> Human: How many legs does a spider have? Machine: Three |
Impact of RankBrain on Websites Page ranking
At this stage it is difficult to predict and to know what the impact of this new AI algorithm may have on page ranking since the system will learn and evolve behind the scenes to produce new sets of search results. However we know that Google are attaching a significant importance to this breakthrough and ambiguous system having ranked this new signal as 3rd amongst hundreds of others (see infographic pictures provided by http://www.audiencebloom.com/2015/10/google-ranking-factors-2015-infographic/). Corrado stated that RankBrain is now the third most important signal in matching results to a search query. He also added that "the system is carefully monitored" and "turning off this feature would be as damaging to users as forgetting to serve half the pages on Wikipedia". He also stated that Ranbkbrain quickly became the third most important signal our of "hundreds" (https://plus.google.com/+webpronews/posts/HjfXin1CEg8). This new update resembles that of the hummingbird update with more emphasis on semantic search and content. As a SEO expert, it is another important signal to consider, but how can this be analysed if indeed it is artificially and constantly evolving. Content is still and will remain key but the sophistication of such a search queries anylising system is now entering a new era and level.