Google AI has unveiled a new language model called BERT

Google Bert

What is BERT?

BERT is a deep learning algorithm that was developed by Google AI as a way to better understand the meaning behind complex natural language queries. It uses a technique called transformers, which are neural networks that can process sequences of data, such as words in a sentence.

Unlike previous language models, which processed text in a left-to-right or right-to-left sequence, BERT is bidirectional. This means that it looks at the context of a word in relation to both the words that come before and after it, allowing it to understand the full meaning of a sentence.

How does BERT improve search?

With BERT, Google is able to better understand the nuances of natural language queries and provide more accurate and relevant search results. For example, let’s say you’re searching for information about a specific type of bird, like a “roseate spoonbill.” In the past, Google might have returned results for spoonbills in general, or for roses. But with BERT, the search engine is able to understand that “roseate spoonbill” is a specific type of bird and provide results that are tailored to that query.

Another way that BERT is improving search is by helping Google better understand longer, more complex queries. These types of queries are becoming more common as people increasingly use voice assistants like Siri and Google Assistant to search the web. BERT allows Google to understand the context of these longer queries and provide more accurate results.

What are some other AI-powered search updates from Google?

In addition to BERT, Google AI has announced several other updates to its search algorithms. One of these updates is called Neural Matching, which uses AI to better understand the meaning behind queries that don’t necessarily include exact match keywords. This allows Google to provide more relevant results for searches like “how to make a cake without eggs” or “what’s the best pizza place near me.”

Another update is called Passage Ranking, which is designed to help Google better understand the relevance of specific passages within a webpage. This allows the search engine to surface results that contain the most relevant information for a particular query, even if that information is buried deep within a webpage.


Google AI’s BERT algorithm is a game-changer for search. By allowing the search engine to better understand the nuances of natural language queries, BERT is making search results more accurate and relevant than ever before. And with other AI-powered updates like Neural Matching and Passage Ranking, Google is continuing to improve its search algorithms and provide users with the best possible results.

Leave a Reply

Your email address will not be published. Required fields are marked *