English language newspapers are notoriously prone to losing their readership as soon as the story is published, and that’s certainly true for many of the world’s most popular news outlets.
But researchers from University of Toronto have come up with a new tool that they say could potentially help identify some of the most important stories that keep people reading.
They’ve developed a system that identifies news articles that are popular on Facebook and Twitter, and then compare the number of times each of those articles have been shared with the number that have been “liked” on social media.
And they found that it does not matter if the article is written in English or in another language, whether it is about a topic that has already been covered by news outlets, or even whether the subject of the article has any real connections to the subject.
This is because people share information about the same subjects on the internet with different frequency.
“It is possible to measure the frequency of people sharing news with friends and family and friends with friends, but these data do not necessarily reflect the extent of the news sharing that takes place over the internet,” they wrote in the journal Science.
The team used social media data from Facebook and other platforms, including news aggregator Reddit and news website The Verge, to compile a dataset of more than 14 million articles from all around the world, which included a mix of articles written in Chinese, English, French, German, Italian, Japanese, Korean, Portuguese, Russian and Spanish.
In total, the team collected more than 3.5 billion stories from the dataset.
“We then compared these stories to other published news stories published on social platforms,” they concluded.
“Using this information, we identified topics that were most likely to be shared on social networks, and we used these topics to categorise these stories into the top 10 news stories of the year.”
They also identified which news stories had the most shares on each of these platforms.
In the top 100, they found articles from China, which has one of the highest rates of internet addiction in the world.
The next five most popular articles were in Spanish, French and Korean, followed by articles from Italy and German.
“We observed that stories about the environment were most frequently shared by Chinese-speaking countries and by English-speaking ones,” they explained.
They found that the top story in English-language newspapers was a story on China, followed closely by the US, Germany, Japan and the UK.
But their findings were not conclusive: the authors note that, because they only counted the most popular stories, the analysis may not be as meaningful as the news sites themselves.
And although it might help people stop reading, the researchers concluded that “the data cannot be used to assess the extent to which a particular story is actually beneficial or harmful.”
“We do not know whether this phenomenon is unique to the internet, and whether it reflects a broader pattern of news consumption in which the news is most popular in countries with low levels of media literacy,” they added.
“Nor do we know whether the data are robust across news sites, and across countries, or across time.”
Follow James on Twitter: @jameschopin