Google Launches Technologies to Contextualize Web Searches


Google announced, during the Search On event this Wednesday (29), new features to make the experience of its users richer. The search platform is expanding the “About this Results” feature, launched in early 2021, to present context information such as site description and comments from other users.

The tool now presents information about the source itself, such as its description from Wikipedia and what the site says about itself, as well as news, analysis and other details that can help to better understand the reputation of new or unknown sources.

Google will also display when the site first came online, which should help users identify pages that may have appeared more recently as part of a disinformation or advertising campaign.

Another feature, called “About This Topic” will also point users to other information related to the user’s original query. This can help prevent people from isolating themselves into a so-called “content bubble,” something that has become commonplace in the social media era and restricts exposure to users’ preferred sources of information to generate more engagement.

The new features are not available yet, but will be released soon in the United States. The company says it is working to take the tool to more countries in the future.

Machine Learning

By providing more insight into the surveys, the company plans to leverage its new Multitask Unified Model (MUM) machine learning technology. Google wants to encourage users to ask specific questions in order to deepen the experience.

According to Google, artificial intelligence is expected to change the way people use the search engine in complex tasks. MUM is being trained in 75 different languages ​​and in many different tasks at the same time, with text and image information. Soon, the technology will be able to understand other modalities, such as video and audio.

According to the company, the algorithm could develop a more comprehensive understanding of the information and knowledge of the world than previous models. With that, the technology will be able to curate more deeply what users are looking for and deliver more relevant and useful results.


Please enter your comment!
Please enter your name here