2020-03-24: StoryGraph at Computation + Journalism Symposium 2020 Non-Trip Report

Click to expand: Overview of StoryGraph illustrating the process of generating a news similarity graph is four primary steps. The four steps are explained in the StoryGraph Tech Report
We never did give StoryGraph a proper introduction. Over three years, I have tweeted, created a Twitter account (@storygraphbot) for StoryGraph, and published two blogposts that utilized the StoryGraph service to determine the top news stories of 2018 and 2019. But I never really introduced and motivated the need for StoryGraph. I hoped that the Computation + Journalism Symposium would provide the opportunity for giving StoryGraph a proper introduction, but the COVID-19 pandemic interrupted it.

The Computation + Journalism Symposium (CJ) is a conference centered on promoting "new ways to find and tell news stories with, by, and about data and algorithms." StoryGraph was accepted as a poster at CJ 2020, and Dr. Weigle and I were scheduled to present StoryGraph at Northeastern University in Boston, the location of this year's conference. We very much looked forward to this year's conference with keynotes from Amanda Cox, Data Editor, New York Times, Deen Freelon, Associate Professor, UNC Hussman School of Journalism and Media, and David Rothschild, Economist, Microsoft Research. However, CJ 2020 was postponed because of the Coronavirus pandemic.

Instead of waiting until after we present StoryGraph at CJ 2020 (hopefully sometime during the fall), we thought it reasonable to announce the publication of the StoryGraph Technical Report which introduces and motivates StoryGraph, and presents two studies that applied StoryGraph.

StoryGraph started out as an experiment to study manufactured consensus in a media manipulation study I conducted during my internship at the the Berkman Klein Center for Internet & Society at Harvard University under the supervision of Dr. Rob Faris. It's been running since August 8, 2017, creating news similarity graphs at 10-minute intervals. We never intended to run it this long. And I'm glad we didn't switch it off.

-- Alexander Nwala (@acnwala)