C+J Symposium 2014

Preamble: What's BBC News Labs?

Preamble: What's BBC News Labs?

Automation in the newsroom?

Idea: Leverage news nerds' skills to make journos' lives easier.

Postulates:

  • The machine is excellent at executing repetitive tasks
  • Coding is a super-power allowing to handle things a human couldn't
  • The "hacks and hackers" paradigm seems to generate great results

i.e. doing magic in news

Approaching computer-assisted journalism

  • Helping journalists to talk to their audiences:
    linked data, news recommendations and personalisation
  • Helping journalists to find stories:
    monitoring new sources of informations

BBC Linked data and semantic tagging

Linked data powered elections

Linked
data powered
Olympics

Biography, medals, records,
news about this athlete....
aggregated automatically!

Covering the European Elections with Linked data

Covering the European Elections with Linked data

News coverage and associated concepts
for each candidate

Finding stories with data

Computer-assisted journalism isn't requiring technical proficiency from journos

It's also providing tools for them to use, and assistance in their research

Datastringer's architecture

JavaScript and Node for (re-)usability

Parametrised queries and comparisons of data...
And notifications when defined events happen

BBC London example

Monitoring crime statistics for the
Greater London area
Alert when monthly stats show
a rise or a fall of 25%...
compared to yearly average

Prop results

So far, email is used for notifications

Conclusion


Merci


@basilesimon

basile.simon01@bbc.co.uk