Image: © chones/Stock.adobe.com
With media organisations constricted more than ever by dwindling revenues and greater content demands, will data journalism be their greatest weapon?
The argument is often made that true investigative journalism – which leads to breakthroughs such as Woodward and Bernstein’s unearthing of the Watergate scandal – is so much more difficult to do in the current media climate.
The advent of the internet and 24-hour news has resulted in a rush for media outlets to be the first to report the news, or at least one of the first. This pushes newsrooms to their limit as they attempt to report the story, at the expense of a hidden story lying somewhere within a pile of dusty files – or in today’s case, a computer hard drive.
To get a sense of costs involved, it was recently revealed that the research and publishing of the ‘Panama papers’ – which revealed billions of dollars in offshore accounts – cost the International Consortium of Investigative Journalists $2m.
Yet in an era when the term ‘fake news’ is bandied about far too often, investigative journalism could be set for a resurgence, with one weapon now available that the greatest sleuths from decades ago could only have dreamed of.
That weapon is data journalism, and with its almost overwhelming possibilities comes a data journalist’s ability to sift through reams of information and be the watchdog over those in authority.
While many advocates see it as a means of helping struggling, small media outlets, some organisations have decided to do it full-time, most notably FiveThirtyEight, which dives deep into data to cover US politics and sports.
Not only #Erasmus. Almost 700,000 European students are enrolled in European universities outside their countries of origin https://t.co/5vHpzVBRuZ @BalkansCaucasus pic.twitter.com/Pvfk6kNF0J
— European Data Journalism Network (@EdjNet) October 15, 2018
Rapidly improving tools
In Ireland, meanwhile, one of its biggest advocates is Dr Bahareh Heravi, lecturer and assistant professor at University College Dublin’s (UCD) School of Information and Communication Studies. She is the founder and director of the college’s Data Journalism CPD programme and founding co-chair of the European Data and Computational Journalism Conference.
Speaking with Siliconrepublic.com, Heravi said the biggest breakthrough for data journalists has been access to tools, as well as their simplification for those less adept at data mining.
“The tools that are becoming available to journalists have changed significantly,” she said. “Between 10 and 15 years ago, a lot of the things that journalists can do now – such as using simple tools – you’d need to write very complicated code for. Only a programmer or someone who is an expert in it could do it, but now there’s some tools that only take minutes.”
In fact, some advocates have gone so far as to create a handy website listing some of the open source tools journalists can use, most of which require no experience to do so.
This week, #carto and #Tableau with fancy visualisations of crime in #hoxton #data #datajournalism #city pic.twitter.com/EMd3qr9zhv
— Connor J Ibbetson (@ConnorIbbetson) November 3, 2017
But what is the current state of data journalism today? Heravi was keen to have this question answered, resulting in a global data journalist survey published last year.
With many respondents being members of large media organisations, almost half (46pc) said they had at least a dedicated data desk or team within their company. Of these teams, 70pc work in teams of between one and five. Interestingly, only 18pc of respondents said they were experts in data journalism, with half admitting they had no prior training – but this is changing.
“A lot of [data journalists] are becoming more and more interested in data analytics and statistics. They all say they need data analysis, but they do it at a very basic level and, really, they don’t go into any sorts of statistical analysis.
“Now, suddenly, in the past six months or so, a number of different organisations have started creating courses online on statistics for journalists, and that’s one of the new trends.”
Role of AI
Another new trend – or at least one that has ramped up considerably in the past few years – is that of artificial intelligence (AI). In the majority of cases, a data journalist’s role is to collect enormous existing datasets and find what’s useful to make a story out of them.
But what if it were possible to use AI to link together datasets to find a story that doesn’t exist yet? Such technology is starting to pop up, including that created by researcher and Libre AI co-founder Dr Ernesto Diaz-Aviles.
His recent work with the Google News Initiative is helping to build a prototype based on AI and machine learning that mines the internet to predict interconnections of global risks that will be at the core of tomorrow’s news. While currently just an experiment, it could eventually guide future AI systems to give journalists a major helping hand, reducing investigative work dramatically.
Reuters, too, is attempting to use AI in a big way, using a specialised product to help journalists with story ideas, and in some cases write sentences.
So, does Heravi see a future where data journalism becomes a lot more AI-driven, both in terms of mining and predicting? Referencing a quote attributed to Danish physicist Niels Bohr, she said: “It’s difficult to make predictions, especially about the future.”
But she is still very hopeful of where this will go, having seen great interest in her own UCD course from professional journalists. “Yes, I’m very optimistic [about the future of data journalism] and I think it is definitely going towards becoming more mainstream,” she said.
“I would hope that every journalist isn’t a super-specialist data journalist, but at least does some basic stuff, and I think it’s already happening.”
