Don't change the container!

• Founded 2009

• Gerd Kamp + 4 employed + 5 free-lance

• Demos not Memos

• R & D and "first customer" for dpa

• part of dpa's digital subsidiary dpa-infocom

• ready-to-publish

• digital-only customers

• live content (liveblog, soccer/football, election results)

• apps, interactive infographics, etc.

• Founded 2014

• Santa Clara University (California)

• Directed by Sally Lehrman

• Funded by Craig Newmark*, Google, Democracy Fund and the Markkula** Foundation.

• Participants: Google, Bing & Facebook; The Associated Press, BBC, La Stampa, The Economist, The New York Times, The Washington Post,  ZEIT Online, ....

* Craig Newmark founded the classified website Craigslist that took a lot of income away from US newspapers.  
** Mike Markkula was employee No. 3 and long time board member at Apple.
Here’s an piece in The Atlantic where Sally Lehrman describes the project: https://www.theatlantic.com/technology/archive/2017/05/what-people-really-want-from-news-organizations/526902/

Basis for Trust: Transparency
 

Participants like dpa make information about Trust Indicators* visible to humans and machines.


 

* the Trust Indicators were developed after interviewing news consumers about what they think makes a news story trustworthy.

Source level information (constant for every news item), i.e.

• Best Practices
  (Link to ethics policy, corrections policy, masthead, ownership structure, point of contact

News item information, i.e.

• Type of Work
  (News, Opinion,  Analysis ...)

• Citations and References
  (aka public part of our Notizblock)

• Methodology
  (Corrections, location of Journalists)
   

• Author Information
  (Still in evaluation at dpa, The Economist / BBC have a "why we run no bylines" explainer)

Details:

https://thetrustproject.org/collaborators-2/

Trust Indicators - The Theory

Trust Indicators - The Practice (1): Humans

https://dpa-newslab.github.io/trust-project-landing-page/

Trust Indicators - The Practice (2): Machines

schema.org markup

Better technologies have been available for decades, but most news agencies' stories are processed in a capability-poor format conceived when teletype machines were the best means of getting stories to customers. A format hostile to hyperlinks, processed by poorly standardized, slow-release-cycle legacy systems, with a human copy-and-pasting plain text more often than not part of this workflow. (More generally, many editorial content management systems tend to have the same limitations, regardless of the use of news agency material)

This whole ecosystem is not build to easily accomodate new metadata needed for the kind of transparency that can help make news more trustworthy.

Don't change the container!

Details:

https://github.com/dpa-newslab/metadata-as-a-service/

• REST

• Browser-ready
  (-> customer content metrics, we have a GDPR-compliant solution :-)

• Cloud-based: AWS Lambda + Cloudfront + Robert Schmidt-Nias's AWS-backed content repository
   

• Future: "dpa-newslab native" with AWS-backed GraphQL -> "select your own data shape"

Thank you. Questions?

twitter.com/mvtango <mvirtel@dpa-newslab.com>