fbpx

How the search engine uses MUM and BERT for safer search results.

3 min read

How Google interprets your search queries

Have you ever wondered how Google interprets your search queries? It’s a complex process involving many factors, including language interpretation, machine learning, and artificial intelligence.

Google’s search systems have improved in understanding human language in recent years. This is thanks to AI and machine learning advances, which have allowed Google to develop better algorithms for understanding the meaning of search queries.

For example, Google’s MUM (Multitask Unified Model) is a new AI model that can understand the context of a query and give relevant results. BERT is another AI model that can help Google better understand the meaning of search queries, especially those related to explicit content.

Google is working to improve the way it interprets search queries. In the coming weeks, Google will begin rolling out changes using MUM to recognize better when a searcher is distressed. Google will also use BERT to identify better when a searcher is looking for explicit content.

These changes are designed to help Google provide more relevant and helpful search results while also protecting users from potentially harmful content.

Here are some key points to keep in mind:

  • Google uses AI to understand the meaning of search queries.
  • Google is working to improve the way it interprets search queries.
  • Google will begin rolling out changes in the coming weeks that will use MUM and BERT to improve the safety of search results.

 

 MUM is being used by Google to better serve people in personal crises.

MUM can transfer knowledge across languages

MUM, or Multitask Unified Model, is a large language model developed by Google AI. It is trained on a massive dataset of text and code, and it can perform a variety of tasks, including translation, summarization, and question answering.

One of the unique features of MUM is its ability to transfer knowledge across languages. This means that if MUM is trained on a dataset of text in English, it can also perform tasks in other languages that it has not been explicitly trained on.

This ability to transfer knowledge across languages is important for Google because it allows them to scale safety protections around the world more efficiently. For example, if Google wants to develop a new safety feature that works in English, they can train MUM on a dataset of English text and code. Once MUM is trained, it can be used to develop the same safety feature in other languages, without having to train a separate model for each language.

This is a significant improvement over previous approaches to safety, which required Google to train separate models for each language. This was a time-consuming and expensive process, and it made it difficult to keep up with the latest threats.

MUM’s ability to transfer knowledge across languages makes it a powerful tool for protecting users from harmful content. It is a step towards a future where safety protections are available to everyone, regardless of their language.

For example, Google uses AI to reduce unhelpful and sometimes dangerous spam pages in search results. In the coming months,  MUM will be used to improve the quality of spam protections and expand to languages where there is very little training data. With MUM google will be able to better detect personal crisis queries all over the world, working with trusted local partners to show actionable information in several more countries.

Like any improvement to Search, these changes have and will continue to go through rigorous evaluation. 

BERT Have a better understanding of what you’re looking for.

BERT helps Google reduce explicit content in search results

BERT is a machine learning model that helps Google understand the meaning of search queries. This includes understanding whether a search query is genuinely seeking out explicit content.

Google has been using BERT to improve the safety of its search results for some time. This improvement has reduced unexpected, shocking results by 30% in the past year alone.

This is a significant improvement and essential for searches related to ethnicity, sexual orientation, and gender. These searches can disproportionately impact women, especially women of color.

Google is always committed to providing everyone with a safe and inclusive search experience. BERT is one tool that Google uses to achieve this goal.

 

Source – The Keyword