Google’s search engine will be able to use the BERT algorithm to process complex user queries in more languages.
Google BERT (Bidirectional Encoder Representations from Transformers) is a search engine algorithm based on a neural network that is designed to improve the relevance of search results due to the ability to analyze sentences, not key phrases.
Thus, the system tries to grasp the context of the entire query, including taking into account prepositions and phrases, and select the appropriate result.
BERT was launched in October with only English support and applied to about 10% of requests. Now, Google announced that BERT will process requests in more than 70 languages, including Ukrainian and Russian.
Portuguese, Punjabi, Romanian, Russian, Serbian, Sinhalese, Slovak, Slovenian, Swahili, Swedish, Tagalog, Tajik, Tamil, Telugu, Thai, Turkish, Ukrainian, Urdu, Uzbek & Vietnamese.
– Google SearchLiaison (@searchliaison) December 9, 2019
Google says BERT is one of the biggest breakthroughs in the last 5 years. The use of technology for the company is to solve the problems of processing natural language (Natural Language Processing). For users, the algorithm will be useful for a more relevant output of results on their requests.