RankBrain is Google’s name for a machine-learning artificial intelligence system that’s used to help process its search results. It supports Google to understand the meaning behind queries and provide relevant search results to users.
RankBrain uses tensor processing unit (TPU) and application specific integrated circuits (ASICs) to process requests. It processes queries into word vectors or distributed representations and matches them. RankBrain attempts to predict what users mean and records the results and customizes the results to give satisfactory outcomes. If it encounters unfamiliar queries, it suggests words or phrases that may have the same meaning to the word or phrase the user is searching and filters the results.
RankBrain interprets the relationships between words, including words that were ignored previously by Google. Words such as the, and, without, etc are important to completely understand the meaning behind a search query. It can analyze patterns between searches that seem unconnected and understand how those searches are related to each other.
Google launched RankBrain and used it globally in 2015. It is a component of Google's search algorithm Hummingbird. RankBrain focuses on refining queries that Google processes by ranking web pages through identifying relevance features. RankBrain has become the third-most important signal contributing to the result of a search query.
Timeline
Further Resources
FAQ: All about the Google RankBrain algorithm
Danny Sullivan
Meet RankBrain: The Artificial Intelligence That’s Now Processing Google Search Results
Danny Sullivan
What Is Google RankBrain And Why Does It Matter?
Jayson DeMers