The Google search engine is one of the most popular and widely used means of finding the right information and answers to various questions. The search engine is used all over the world, giving users the ability to quickly find the data they need. That is why the Google development team is constantly working to improve their own search engine.
Currently, each query is perceived by the Google search engine as a set of words for which relevant results are selected. The system copes worse with conversational and complex queries, and understanding the language remains an urgent problem for a long time.
In the near future, the company intends to introduce a new natural language query processing algorithm based on the BERT (Bidirectional Encoder Representations from Transformers) neural network introduced last year. The algorithm is able to analyze the request completely, without breaking it into words and taking into account prepositions and conjunctions. This approach will allow you to get the full context of the request, finding more appropriate answers.
Google developers say that the creation of an algorithm based on the BERT neural network is βthe most important achievement in the last 5 years and one of the most significant events in the history of the search engine.β The new algorithm is currently being used to process some of the queries to the Google search engine in English. In the future, the action of the algorithm will spread to all supported languages, but it is still difficult to say when this will happen.
Source: 3dnews.ru