What is Google BERT?

[Ehsanuls55]

Google BERT ( Bidirectional Encoder Representations from Transformers ), as explained on Google's official website (English), is an algorithm responsible for understanding language in a more intuitive way .

More precisely, it is a natural language processing model , or NPL. NPL is a type of artificial intelligence that transforms human language into computer-understandable information and then extracts meaning from it.

The same thing happens with new Artificial Intelligence technologies for SEO optimization (AI SEO) .

Of course, this artificial intelligence has limitations. One of them is understanding the subtleties of what a person might say or write. Human language is extremely complex and elements such as context, word arrangement or intonation can completely change the meaning of a sentence.

Google BERT is a step in the direction of closing that gap. It's an algorithm that takes into account things like the use of prepositions to extract meaning from the entirety of a user's search.

Google BERT is the algorithm responsible for understanding language in a more intuitive way, using artificial intelligence that transforms human language.

Google's algorithm has always paid attention to the words in a search query alone, discarding vp maintenance email list prepositions and ignoring much of the context.

With BERT, those things now matter. The algorithm is now able to identify the correct meaning of words with multiple different uses and extract a complete meaning from a sentence. For example, whereas before Google's algorithm ignored prepositions, it now includes them in its process to arrive at a more correct interpretation.

Let’s look at the example Google used to demonstrate the effect of BERT. A user searches for “ 2019 brazil traveler to usa need a visa .” In English, that would be equivalent to “2019 viajero de Brasil a EEUU necesita visa.” Previously, Google’s algorithm would simply show the most relevant results for those words, without distinguishing the user’s intent. In the example, the first result the old algorithm returns is for information for travelers from the United States to Brazil.

Google Bert, what is it, how does it work and how to optimize your SEO?

Instead, the first result that BERT gives in the example is about information on obtaining a visa to travel to the United States from Brazil. This is because BERT takes into account the user's intent. Because it can interpret the preposition “to” (or “a” in Spanish), BERT understands that the user wants to travel from Brazil to the United States.