With this system, in fact, Google is starting to understand how combinations of words can express different meanings and intentions. Rather than focusing on the individual words that make up the query , BERT aims to analyze the entire sequence to better extrapolate the search intent . This feature is useful for providing relevant results especially with complex and long queries , in which greater weight is also given to prepositions and conjunctions. The image above helps to understand the evolution of SERPs with Google BERT.
What happened? Before BERT, Google gave more importance morocco number data to the search for a “math book”, ignoring the “adult” specification. The suggested result was therefore a popular workbook. After the introduction of BERT, Google is evidently able to understand search intent better. ( adults/grownups ) but responds more precisely to the actual needs of users. In short, BERT puts the last nail in the coffin of the keyword , that magic little word that we SEO consultants loved so much.
.. Today BERT comes into play in most searches performed on Google . The advanced language understanding guaranteed by BERT helps to quickly classify web pages and sort them based on their relevance to the specific query . Extraordinary, isn't it? Yes, but only if you haven't heard of Google MUM yet… Read also: Google BERT (easy explained) Google MUM, from language understanding to information processing In May 2021 Google presented its latest creation to the world: MUM (Multitask Unified Model).