RankBrain, the new Google algorithm

What is RankBrain?

To begin with, it is important to mention that Google implements its algorithms in its search engine with a primary goal: to offer the most relevant results, quickly and regardless of how the search was performed (what terms or the order in which they were entered).

And that’s how they created RankBrain, an artificial intelligence system that has been in operation since 2015 and is based on learning.

RankBrain is artificial intelligence

Google intends that its search engine does not focus only on the words entered for the search, because in many cases if the word or expression is not well written, the searches may not be understood by a machine. RankBrain converts the words into “vectors”, that is, mathematical entities intended to be understood by the search engine.

Google has been researching and working on artificial intelligence for years. It has also been working on speech, translation, and computational visual processing. With this, it can be clearly seen that Google intends to convert its search engine into ANI in the near future.

ANI (Artificial Narrow Intelligence): It is a limited intelligence, but not thereby having real human intelligence (strong AI), having a mind, mental states, or consciousness.

Google’s algorithms

Google SEO was mainly focused on the Hummingbird and Panda algorithms:

• Hummingbird: This algorithm is in charge of managing and relating the information in Google, with the objective of ordering it in the search results.
• Penguin and Panda: These algorithms value the quality and freshness of the content; and are in charge of penalizing for black hat SEO practices such as massive link building.

RankBrain has become the third algorithm that simultaneously influences search results, learning constantly and evolving with each new search so that the next one offers an even more relevant result in less time. That is why RankBrain is said to be a learning system.

This means that RankBrain is in charge of interpreting both ambiguous and new queries; and that it learns from the user’s response to the results offered, thus determining whether they have been relevant or not, in order to be able to take it into account for future searches.

The very success of RankBrain’s operation is therefore the user’s own response to the search results. Based on the user’s interaction, Google will be able to determine which page is the best match for the query.

What factors does RankBrain take into account?

When converting searches into mathematical entities (called vectors) for artificial intelligence understanding, it takes into account factors such as:

• A query that a user makes
• A web page responsive to the query
• The search session in which the user makes the query (when searching for something, several different queries can be made until something is found. This would be the search session)
• The time the query was made
  Ads displayed in the query result
• The anchor text of a link on a page
• The domain associated with that page

The relationships between these search entities can create a “probability score”, which could determine whether a page is displayed on the search results page or not.