Google will improve its featured snippets thanks to its AI Multitask Unified Model (MUM)

Featured snippets are small summaries offered by Google in response to a question frequently asked by Internet users. On August 11, Google presented an innovation to improve their quality: the company will use MUM, (Multitask Unified Model) its AI-based algorithm built on the Transformer architecture, to bring the notion of consensus to its models.

Google’s ranking systems aim to bring out relevant information from the most reliable sources available. Google is constantly looking to improve them and has announced a significant innovation to improve the quality of featured snippets, featured snippets of text providing information following a query.

The company states:

“Using our latest AI model, Multitasking Unified Model (MUM), our systems can now understand the notion of consensus, which is when multiple high-quality sources on the web all agree on the same fact. Our systems can check snippet captions (the word or words called above the featured snippet in a larger font) against other high-quality sources on the web, to see if there is a general consensus for this legend, even though the sources use different words or concepts to describe the same thing. »

She adds :

“We have found that this consensus technique has significantly improved the quality and usefulness of featured snippet captions. »

Also, Google will highlight the caption, to provide a better answer to the user’s query, like in the screenshot below.

Detect false information

Another enhancement to featured snippets targets questions for which there are no answers, which Google calls “false premises.” The company cites recent research as an example: “When did Snoopy murder Abraham Lincoln? “. The snippet provided gives a specific date and information about Lincoln’s assassination, but for Google “this is clearly not the most useful way to display this result.” »

Google said it has improved the snippets it displays for queries containing information about events that didn’t happen:

“We’ve trained our systems to better detect these kinds of false premises, which aren’t very common, but are instances where it’s not useful to show a featured snippet. We’ve reduced the triggering of featured snippets in these cases by 40% with this update. »

Information literacy

Google is also developing information literacy features in its search to allow people to evaluate information, whether it comes from social media or conversations. The company has thus invested in the creation of the fact-checking explorer, “Fact Check Explorer”, the reverse image search “Reverse Image Search” and “About this result” and announces that it has updated these functionalities.

Thus, About this result, which has been used more than 2.4 billion times since its launch last year, will soon be available in eight other languages: German, Spanish, French, Indonesian, Italian, Japanese, Dutch and Portuguese. Google also added more context to it.

Leave a Comment