Google search can prove to be a frustrating situation when it focuses more on certain keywords in the search query and not on the relevant details in it. This deficiency brings out results that are of no use. Thanks to the magic of machine learning and natural language processing, Google is now working to improve its eponymous search engine to better understand user queries. As a result of the advancement, Google search can now better understand linguistic nuances and search elements such as propositions and combinations.
Pandu Nayak, vice president of search at Google, wrote in a blog post that the company is using a technique called bidirectional encoder representation from Transformers (BERT) to improve search. BERT is essentially a neural network-based technique for pre-training natural language processing (NLP) that helps build custom answering systems. The main advantage of BERT is that it analyzes search queries, focusing on the context of a sentence, how humans are able to communicate and understand language naturally, rather than just doing a word-by-word analysis.
“With the latest advancements of our research team in the science of language comprehension – made possible by machine learning – we are making a significant improvement in how we understand questions, represent the biggest jump in the last five years, and a search Biggest leap forward in the history of India, ”Nayak wrote. He says that improvements in Google search are more pronounced in the case of English, but results in languages such as Hindi, Korean and Portuguese have also been encouraging. The essence is that BERT will allow search engines to better understand linguistic elements such as “among others”, “to” and “for” and bring search results accordingly.
The focus here is on enabling users to see their query on Google search in the same way they put it in a conversation with someone else, rather than typing a keyword-heavy gibberish in the search field that would make no sense in the natural . chit chat. Google is currently testing the BERT-supported search model in two dozen countries and aims to improve Google search to the extent that it responds to queries of an interactive nature rather than being limited to a collection of hit-and-go Bring relevant results. Recall keyword-rich questions.