For years now, Google has been implementing and developing new technologies related to artificial intelligence and machine learning in its Search, as this way they are improving the daily search experience for users. In this sense, since 2015 when RankBrain was implemented, Google has continued to implement AI systems that improve language understanding and therefore improve search results.
Google uses several artificial intelligence algorithms with different functions and that directly influence the positioning of content.
In this sense, Google uses several AI systems: RankBrain, neural matching, BERT and MUM. Each of them is aimed at a specific part of the searches, so it is important to know a little better how each of them works.
-RankBrain: RankBrain is the first AI system implemented by Google in 2015 and helps Google understand how words relate to concepts and can take a broad query and better define how that query relates to real-world concepts. While it was initially used in only 15% of searches, it is now implemented in all of them, in all languages and in all regions. RankBrain helps to rank search results and is part of Google’s ranking algorithm.
-Neural matching: “Neural Matching” or “neural matching” was the second AI system launched by Google in 2018 and expanded to local search the following year. Neural Matching helps Google understand how queries relate to pages by contrasting the full query and page content to provide context to the search.
-BERT: Biderctional Encoder Representations from Transformers or BERT is a neural network for natural language processing that was implemented in 2019. BERT allows Google to understand how combinations of words express different meanings and intentions, including looking at the sequence of words on a web page, so even seemingly unimportant words in each query count. BERT is also part of Google’s ranking algorithm.
-MUM: Multitask Unified Model or MUM is Google’s newest AI, having been introduced in 2021 and expanded later in the year to work in other branded applications. MUM helps Google understand different languages and also generate them, so it can be used to understand variations on new terms and languages. For now, MUM is not part of Google’s ranking algorithm.
-Broad algorithms: Google implements periodic changes and updates to its “broad algorithms” that often cause changes in Google rankings. Although these algorithms are also based on Artificial Intelligence, they only work in combination with the systems mentioned above.