Global Investigative Journalism Network (https://archive.gijn.org/tag/algorithms/)
In a panel discussion at the IRE23 Conference, experts shared tips on how reporters can identify and investigate algorithmic harm and AI bias, and hold the human masters of these systems accountable.
Many in the journalistic profession fear Artificial Intelligence will leave them without a job. But AI could become the savior of the trade, making it possible to better cover the increasingly complex, globalized and information-rich world we live in. Open Society Foundations’ María Teresa Ronderos writes about the ways some of the world’s newsrooms are using it now.
Around the world, people are toning down their celebrations in a bid to mitigate spread of the coronavirus. The latest casualty: Thanksgiving. Our NodeXL #ddj mapping from November 16 to 22 found FiveThirtyEight’s timely piece explaining the risk of COVID-19 transmission from even small Thanksgiving dinner gatherings. Also popular: The New York Times tracking the status of all vaccine trials in progress, and The Pudding analysis of race and gender in crossword puzzles from five major US news publications.
How does climate change increase the likelihood of weather extremes? Our NodeXL #ddj mapping from October 12 to 18 finds ProPublica examining how the warming global climate exacerbates wildfire activity, and designer Will Chase visualizing some of the most destructive hurricanes with an interactive wheel graphic. The Financial Times is summarizing key trends of the pandemic using charts, maps, and graphics, while The Markup announced its latest project looking into the black box algorithms of social media platforms.
What was the magnitude of the Beirut port blast and how did it compare to other infamous explosions in history? Our NodeXL #ddj mapping from August 31 to September 6 finds Reuters illustrating just how powerful last month’s blast was in relation to the Chernobyl disaster and other explosions. The Tampa Bay Times highlighted how a county sheriff’s office is using an algorithm to supposedly predict and intercept the criminals of the future, while The New York Times used satellite maps to show how physical and political geographies interact across the United States. ESPN has been looking into the potential of people in college football crowds to become COVID-19 super-spreaders, and Slate analyzed the overuse of the word “murmur” in the popular Twilight novel series.
This week’s Friday 5, where we round up our favorite reads from around the online world in English, includes a recent algorithm change on Facebook’s News Feed that will boost original news stories, lessons learned on an academic and investigative journalism collaboration, and European media’s race problem.
What’s the global data journalism community tweeting about this week? Our NodeXL #ddj mapping from February 24 to March 1 finds The New York Times digging into some curious data from two Instagram accounts of the British royal family, Al Jazeera analyzing Trump’s plan to resolve the Israeli-Palestinian conflict, The Markup launching with an investigation into auto insurance algorithms, and Pew Research Center sharing some American election data snapshots.
What’s the global data journalism community tweeting about this week? Our NodeXL #ddj mapping from November 11 to 17 finds an investigation by The Financial Times uncovering how private health information is being traded online by health sites to big advertisers, The Economist looking at the correlation between Americans’ music preferences and political leanings, The Bureau of Investigative Journalism investigating the government’s algorithmic decision-making, and the data community coming together to highlight excellent female data journalists.
How often is social media used as a source in news stories? Can a decision tree algorithm generate tens of thousands of 250-word stories? And what is belief-driven data journalism? These questions were at the heart of some of the promising projects featured at the 2019 Computation + Journalism Symposium.
What’s the global data journalism community tweeting about this week? Our NodeXL #ddj mapping from December 31, 2018 to Jan 6, 2019 finds experts sharing their thoughts on machine learning in journalism with @storybench, @funkeinterativ and @webk1d’s useful tool to merge datasets, @pewresearch’s overview of female under-representation in online image searches, and an analysis by @EDNNews on corpses bodies sent back to El Salvador.