Google Search is now better at showing news articles thanks to machine learning

Google Search is now better at showing news articles thanks to machine learning

We may earn a commission for purchases made using our links.

Even though Google News is a thing, a lot of us still rely on Google Search to stay on top of current events. A simple search on any given topic yields thousands of relevant results and it’s because of this convenience that many users, including me, haven’t transitioned to Google’s dedicated platform for news. In a bid to improve news-related search results and help users develop a deeper understanding of any topic of their choice, Google is now making some changes to how Google Search displays news articles.

According to a recent blog post from the company, Google Search will now make use of machine learning to better organize top news stories. Once the change is implemented, Google Search will now showcase a carousel of relevant articles at the top of your search results highlighting all the top stories related to the search keyword. In case there are multiple stories related to the search, Google will also organize the results by story to make it easier to understand what’s most relevant. Thanks to this approach, users will be able to make a more informed decision on which articles to explore.

Additionally, within each story, the structure will now make room for high-quality content beyond the most recent coverage. It will also include more diverse sources to give users more context and perspective of the topic at hand. For instance, searching for “NASA news” will now show results grouped under distinct news stories ‘NASA adds five companies to moon bid’ and ‘NASA detects water vapor on Jupiter’s moon Europa’, along with additional results under “Also in the news.”

Powering this new approach is a story-understanding technology that maps the people, places, and things involved in any given story. It then draws connections between them and creates distinct groups of news articles. Google uses a variety of different machine learning techniques to generate these groups, including BERT models that examine the related articles and determine where one story ends and the other begins. Google’s research suggests that combining results in clearly defined stories helps users easily navigate the results and identify the content that best suits their needs.

Google Search

Along with the new machine learning-based organization, Google Search will also showcase key information like notable quotes and related opinion pieces in the top stories carousel. These articles are expected to provide users with a more well-rounded of the news story and help them decide which angle to explore further. Google notes that this change is part of its long-term efforts to expand the types of journalistic contributions highlighted in Search. All the aforementioned changes in Google Search are now available on mobile devices for English speaking users in the US. Google plans to roll out these features with additional language support in other regions over the coming months.

Source: Google