Last week, search engine giant Google stated that it’s currently using a type of AI (artificial intelligence) system called RankBrain in order to offer its users the best possible search results based on their queries. The machine learning artificial intelligence system seems to be handling a significant portion of Google’s search queries.
Despite that though, it seems that it cannot be considered a full replacement of Google’s current search algorithm, Hummingbird. According to Bloomberg, the new AI system is part of the Hummingbird algorithm, but instead of handling all search queries, it interprets what users search for and then tries to understand how to submit their requests in different ways.
For instance, let’s say that someone searches for Angela Merkel. If we take a look at the way Google’s algorithm functioned in the past, we know that the results would consist of pages that contain that exact keyword. However, in the past 4 years or so, especially since Hummingbird’s launch, Google has greatly improved at understanding the relationships between words. Therefore, a search for Angela Merkel may bring back information and pages related to Germany’s chancellor, but also of her husband.
RankBrain represents a new age in the way queries are processed and as the previous example shows, it goes beyond what has been used. According to Google, it seems that this is their 3rd most essential factor for ranking websites. While Google doesn’t mention anything about their first and second factors, we have an idea about what they could be.
Therefore, it seems that RankBrain’s main use is to help users find pages that may not contain the exact key phrases/keywords they’re looking for. It’s a very effective way of helping better interpret queries and translate them effectively with the sole purpose of delivering the best search results for the searcher.