"You Are Here": Helping The Public Navigate Their Informational Choices

Over the years we've explored how the GKG's outlink graph can be used to construct "you are here" maps of the global news landscape. How might such concepts be more seamlessly integrated into public consumption of the news? We would love to see new kinds of analyses and visualizations in these areas. Here are a couple of ideas to get you started:

  • A network visualization of global media outlets with connections based on outlinks or similarity overlap or a combination therein. Users could see where their top news sources sit within this landscape or even enter a list of their top sources to see their overall isolation or connectivity to the central "news core" of their nation. Implemented as a browser plugin, this could even be displayed in realtime, perhaps with a bulleted list of the top news outlets most similar to the one they are current consuming and recommendations for outlets closer to the core of the media landscape for their nation on the given topic, potentially integrating the Global Similarity Graph (GSG) to compute document overlap.
  • An "isolate score" for each news outlet that ranks outlets by how close they are to the "central core" of the news landscape of a given country or topic. Outlets further from the core represent the more fringe areas of journalism where falsehoods and conspiracy theories are more free to proliferate. Such quantitative rankings could even be integrated into various efforts to combat information, such as flagging high isolate sites as riskier sources of information.
  • Narrative displays that visualize complex narrative landscapes by clustering the underlying stories, making it possible to rapidly survey the landscape of coverage on a topic and identify the core topical clusters. A particularly powerful application would be a realtime visualizer that could operate as a browser plugin, taking the current news story the user is reading and display a narrative map showing the full landscape of coverage of that story and where the current article sits within that landscape. Even more powerfully, such a system could score how far from the narrative center of the story each article is, allowing the browser plugin to provide a warning to the reader that the article appears to be offering a take on the story that deviates sharply from mainstream coverage.

We'd love to see what kinds of ideas you come up!