Robot News: a new Polis report on data driven news production
This report is the outcome of a workshop held on 24/2/16 at LSE with participants from a range of organisations including Financial Times, The Times, Reuters Institute, BBC, Storyful, The Guardian, Global Editors Network, CUNY and Paris Est University.
The workshop was hosted by Dr Alison Powell, Assistant Professor and Programme Director of the MSc in Media and Communication (Data & Society) and Prof Charlie Beckett, Director of Polis, the LSE’s journalism think-tank.
Journalism has gone through cycles of change, from offline to online, to web 2.0 and now ‘big data’.
In some ways the new uses of data as content, revelation and as information on audience behaviour have deepened journalism and made news media more diverse, rich and effective. But there are also dangers such as using data created by organisations that have their own agendas.
The move towards a journalism of ‘facts’ which is driven by a belief in data as fact is happening at the same time as a shift towards personalisation, subjectivity and relativism. This tension between the factual and the emotional is not entirely new for journalism, but it is does seem to be particularly complex and radical at the moment, driven as it is by the shift to social, mobile and algorithmic news. As our report shows, it offers challenges but also great opportunities for journalism.
Data on audience behaviour can inform journalism
- There is a tension between the democratic/ public function of news and the business imperative of reaching a large audience.
- When using audience metrics, journalists should be mindful of the ‘news gap’: the preference gap between the supply (journalism) and demand (audience).
- While data can enable new stories to be found, journalists still have to make an editorial judgement on what should be made public.
- There is a difference between openness, (the presence of accessible data), and transparency, where journalists deem it in the public interest to make information accessible.
- There are two types of data-driven journalism: one, exploratory which sees data as the heart of the story, and another, explanatory, which sees data as supporting a story.
- When sampling from social media datasets and online communities to discover stories, journalists need to be aware of the bias inherent to the platform itself.
- To build a positive culture around data in the newsroom, there needs to be a willingness to imagine the future of journalism in collaboration with algorithms.
Book Review of ‘The News Gap’ (LSE Media Policy Project Blog)
Legacy Newsrooms Embrace Innovation but not Cultural Change (Media Shift)
The Journalist-Engineer (Matthew Daniels)
Data Visualisation: What’s next? (Signal Noise)