Wednesday, February 15, 2017

Why data driven news generation has come of age

The electioneering period provides a strong

The electioneering period provides a strong case on how data journalism can be used to enrich the quality of news churned out on a daily basis. PHOTO|FILE 

By Timothy Oredo

Two recent events prompted me to write this article. First was the recent concern over growing fake news, and second was the controversy over differing reports on the attendance of the Donald Trump inauguration.

How can newsrooms address such controversies? The need for news verification is more pronounced during an electioneering period. Not to mention opinion polls that form a sizeable chunk of news content leading to an election.

One such growing intervention is Computational Journalism. This is a new way to find and tell news stories by using data and algorithms. Journalists routinely encounter social and political systems driven by new technologies.

To critique their operation, a journalist needs to understand computation. Data and computation exert real forces in the physical world, shaping and defining systems of power that play larger roles in people’s lives.

The electioneering period provides a strong case on how computational journalism can be used to enrich the quality of news that media houses churn out on a daily basis.

In November 2015, the more radical of the two Brexit campaigns, “Leave. EU,” supported by Nigel Farage, announced that it had commissioned a big data company to support its online campaign.

The company’s core strength: innovative political marketing — micro targeting — by measuring people’s personality from their digital footprints. The Trump team commissioned the same company weeks to the US election and it took credit for turning the odds by nuancing communication to a different profile of voters, propelling Trump to victory.

Instead of simply warning against the dangers of polls or urging journalists to change their model of political reporting, there is need to appreciate the reason behind the crisis of political polls. For instance, a proposed remedy might be non-probability polling, in combination with algorithms and big data approach, which does have advantages when it comes to cost, time and fine-grained estimates of public opinion.

On the controversy surrounding the Trump inauguration, where there were varied accounts of how many people attended the event, newsrooms could use analysis of photographs and video footage to give a more precise perspective of the situation.

Computational journalism lowers story discovery and production costs and makes news sources more efficient in their content creation.

The key question lies in how reporters can collaborate with researchers in other disciplines to discover stories. There is need to see new collaborations between journalists and computer and data scientists: create a bazaar for the exchange of ideas between industry (practice) and academia (research).

The writer is a certified big data scientist (MIT) and instructor at Strathmore Business School.