Big Data, data jour­nal­ism and the future of media

The nature of jour­nal­ism is under­go­ing a trans­for­ma­tion. New dig­i­tal tools, such as AI, Big Data, as well as the dis­tri­b­u­tion chan­nels of social media are chang­ing what is required, what is pos­si­ble in terms of career and offers a new set of chal­lenges. What are these exact­ly? How can jour­nal­ism retain its cred­i­bil­i­ty in the dig­i­tal age? How will it ensure its inde­pen­dence in this era of vast­ly dis­parate opin­ions in dig­i­tal space? How can the pop­u­la­tion digest the cur­rent infor­ma­tion flood con­sist­ing of real and fake reports? How does jour­nal­ism cre­ate room for thinkers to have dis­cus­sions on what we want to achieve using all that is pos­si­ble in the dig­i­tal world?

In her lat­est Duet inter­view, Dr Cal­daro­la, author of Big Data and Law, and Prof. Christi­na Elmer, the first Pro­fes­sor for data jour­nal­ism and dig­i­tal jour­nal­ism in Ger­many, dis­cuss the many impor­tant changes in journalism.

You have received the first pro­fes­sor­ship for data jour­nal­ism and dig­i­tal jour­nal­ism in Ger­many. Can you please pro­vide our read­ers with a short intro­duc­tion on what data jour­nal­ism is, how the future in media might look like and what devel­op­ments, ini­tia­tives and break­throughs exist in oth­er countries?

Prof. Christi­na Elmer: Data jour­nal­ism is essen­tial­ly about unlock­ing struc­tured data as sources of infor­ma­tion in jour­nal­ism. These can be exten­sive sta­tis­tics in clear for­mats or even larg­er data leaks that we have to eval­u­ate in an inves­tiga­tive way. For this type of analy­sis, we need new skills, meth­ods and process­es in jour­nal­ism – and teams that are organ­ised in a much more inter­dis­ci­pli­nary way than before and are strong­ly con­nect­ed with oth­er dis­ci­plines and depart­ments. My goal is to ground these nec­es­sary skills more com­pre­hen­sive­ly and, at the same time, train spe­cial­ists who can focus on this excit­ing and increas­ing­ly rel­e­vant area. The field of data jour­nal­ism has been devel­op­ing in Ger­many for about 15 years; in the US, for instance, it has a much longer tra­di­tion. We can, how­ev­er, dis­cern a clear trend in this direc­tion glob­al­ly, which is not sur­pris­ing: With dig­i­tal­i­sa­tion and the grow­ing avail­abil­i­ty of open data, jour­nal­ists can increas­ing­ly access rel­e­vant data sources in their inves­ti­ga­tions – and should do so in order to gain new insights and check oth­er sources of infor­ma­tion. What emerges can, at best, be called evi­dence-based jour­nal­ism. And in my view, this is real­ly an impor­tant devel­op­ment. Because, after all, we have to do some­thing to coun­ter­act the flood of uncer­tain, biased or even manip­u­lat­ed infor­ma­tion in the dig­i­tal pub­lic sphere.

Dig­i­tal media cur­rent­ly offers a huge spec­trum for all sort of pub­li­ca­tions. Once there were news­pa­pers and mag­a­zines, then came the advent of tele­vi­sion while today we use social media and tools, such as AI, Big Data and alike. Infor­ma­tion is not only cre­at­ed by jour­nal­ists but also by peo­ple using their mobile devices that record occur­rences and pro­vide oppor­tu­ni­ties to post them on dif­fer­ent dis­tri­b­u­tion chan­nels. How do jour­nal­ists han­dle the amount of dis­tri­b­u­tion chan­nels and the quan­ti­ty of infor­ma­tion? How are they digest­ed, edit­ed, and val­i­dat­ed con­sid­er­ing the fact that media com­pa­nies have few­er employ­ees and small­er budgets?

Of course, media hous­es try to allo­cate their resources wise­ly – even if their bud­get is tight. In rela­tion to the chal­lenges you men­tioned, this means, for exam­ple, that the area of devel­op­ing and refin­ing new prod­ucts and offers has grown in many out­lets, espe­cial­ly regard­ing dig­i­tal chan­nels and plat­forms. These teams take a close look at what media users need and what for­mats can be used to address them clev­er­ly. In this con­text, new prod­ucts with a strong tech­no­log­i­cal com­po­nent are often cre­at­ed, which is why new con­stel­la­tions and com­pe­tences are also need­ed in this area – from data analy­sis and pro­gram­ming to dig­i­tal dis­tri­b­u­tion and search engine opti­mi­sa­tion. As far as research in the dig­i­tal space is con­cerned, jour­nal­ists are indeed chal­lenged. Check­ing the authen­tic­i­ty of dig­i­tal sources has become an impor­tant skill – some news­rooms have now set up their own units for this type of work. In research as well as in dis­tri­b­u­tion, we also need new ways of struc­tur­ing infor­ma­tion via meta­da­ta and link­ing it in dif­fer­ent con­texts. This could also be a way to pre­pare infor­ma­tion in a man­age­able way so that we can use it appro­pri­ate­ly in our process­es and for publication.

A pic­ture used to be viewed as evi­dence for an occur­rence. Today every pic­ture can eas­i­ly and rapid­ly be manip­u­lat­ed by every­body by means of dig­i­tal tools. How much effort is required and what tech­niques do jour­nal­ists use to ver­i­fy sources? How do they main­tain the cred­i­bil­i­ty of their report­ing? How effec­tive are they and how do they function?

One cen­tral chal­lenge has always been to trace the source back to its ori­gin, if pos­si­ble, and to cross-check it with a sec­ond source: Who took a pho­to and how can we assess its cred­i­bil­i­ty as con­fi­dent­ly as pos­si­ble? Also, who was on the scene and could they pro­vide us with a sec­ond per­spec­tive on it? In addi­tion, in many cas­es one can indeed find traces of the pho­to-fak­ing process in the pic­ture itself. These strate­gies are not new and have always been used in edi­to­r­i­al offices ‑and not only for visu­al sources. Com­par­a­tive­ly new, how­ev­er, are tech­nolo­gies that enable entire video or audio sequences to be fab­ri­cat­ed with remark­ably lit­tle effort. How­ev­er, there also exist tech­no­log­i­cal tools to iden­ti­fy such fakes, which are already being used in news­rooms. Nev­er­the­less, from my point of view, it would be fatal to rely on these tools, since the fakes are nat­u­ral­ly also improv­ing all the time. Con­verse­ly, to me it seems rea­son­able to empha­sise the val­ue of cred­i­ble and ver­i­fied sources and to make jour­nal­is­tic research more trans­par­ent – and, in case of doubt, to always decide against pub­lish­ing rather than putting faith in jour­nal­ism at risk.

Peo­ple are there when some­thing hap­pens and with­in a few split sec­onds, post videos, their opin­ions, reports and sim­i­lar. How do jour­nal­ists han­dle this type of speed, quan­ti­ty, sen­sa­tion­al­ism as well as the lack of research, bal­ance, objec­tiv­i­ty and prop­er use of lan­guage? Or expressed dif­fer­ent­ly: How do jour­nal­ists ver­i­fy sources and con­tent and how does well researched con­tent reach the respec­tive media in time?

There isn’t real­ly a sin­gle answer that applies to jour­nal­ism as a whole – every news­room has its own process­es, dead­lines and time­lines, depend­ing on the pub­li­ca­tion chan­nel, the strat­e­gy and the busi­ness mod­el. How­ev­er, the fol­low­ing applies to all dig­i­tal media offer­ings: In order to stand out from the flood of infor­ma­tion, they have to con­vey more strong­ly than before what dis­tin­guish­es them from oth­er pub­lish­ers. This can mean dis­clos­ing the sources and sta­tions of the research, address­ing pos­si­ble con­flicts of inter­est and con­sis­tent­ly dif­fer­en­ti­at­ing between neu­tral infor­ma­tion and opin­ion, between ver­i­fied knowl­edge and the grey areas of report­ing. The cen­tral issues of our time, such as pan­demics, cli­mate cri­sis and soci­etal trans­for­ma­tion, are very com­plex and can­not be solved with sim­ple recipes. Media can gain trust if they cre­ate spaces in which such com­plex issues can be nego­ti­at­ed. How­ev­er, this also means resist­ing the temp­ta­tion to react quick­ly and thus work against the reward sys­tem of our real-time inter­net is no easy task in a news world that has steadi­ly accelerated.

Human beings are social beings who most like­ly share a major­i­ty opin­ion because it helps them feel secure. Are opin­ions being cre­at­ed with the aid of clicks, tweets and likes? Is an illu­sion of a mass of human beings being cre­at­ed via a sup­posed grass roots aris­ing as a move­ment of cit­i­zens which is then also sub­ject to manipulation? -

If soci­ety were to inform itself sole­ly through media and if media were to align their report­ing exclu­sive­ly to the met­rics you men­tioned, there could of course be such an effect. But from my point of view, nei­ther sit­u­a­tion is true at the moment. In seri­ous edi­to­r­i­al offices, even today, the most impor­tant news items are select­ed on the basis of their rel­e­vance – based not only on the response, but also on find­ings that go beyond that, for exam­ple, from sci­en­tif­ic stud­ies or statistics.

Math­e­mati­cians and data experts from the French Insti­tute for Com­plex Sys­tems have been inves­ti­gat­ing the bat­tle for con­trol of vir­tu­al space and have visu­al­ized it, using the exam­ple of the com­bat between cli­mate sup­port­ers and cli­mate oppo­nents. They showed how news were being shared, how accounts were inter­act­ing and how com­mu­ni­ties were grow­ing. This visu­al­i­sa­tion illus­trat­ed that cli­mate oppo­nents are, numer­i­cal­ly speak­ing, con­sid­er­ably few­er in num­ber than cli­mate sup­port­ers. By con­trast, cli­mate oppo­nents are clear­ly more active and can com­pen­sate for the enor­mous amount of evi­dence against them. They could also show that the Heart­land Insti­tute is behind the group of cli­mate oppo­nents – a think tank of around 40 employ­ees. This Insti­tute which is locat­ed in Chica­go pro­duces tweets, con­fer­ences and books. Are they the opin­ion mak­ers and jour­nal­ists of today or tomorrow?

You have brought up an excit­ing study that real­ly makes you think! But I would not call such actors jour­nal­ists –they are sim­ply not inde­pen­dent and impar­tial enough to deserve that title.

My favourite quote – almost three decades old, but still true:
 

“Com­put­ers don’t make a bad reporter into a good reporter. What they do is make a good reporter better”.

Elliot Jaspin

What they do is fol­low their own agen­da, which does not nec­es­sar­i­ly oblige them to com­mu­ni­cate in the inter­est of a well-informed pub­lic. But, of course, espe­cial­ly when it comes to top­ics with a sci­en­tif­ic basis, the media must pay even more atten­tion to ade­quate­ly cov­er­ing such dis­cours­es. Unlike in socio-polit­i­cal dis­cus­sions, it is sim­ply not enough to set the two sides against each oth­er when sci­en­tif­ic debates are involved – that would pro­duce an effect known as false bal­ance. If there is a posi­tion that the over­whelm­ing major­i­ty of sci­en­tif­ic experts in this field can agree on, then it should log­i­cal­ly also be giv­en more space in report­ing. Espe­cial­ly with regard to top­ics which are both sci­en­tif­i­cal­ly researched and social­ly nego­ti­at­ed, jour­nal­ists have to be care­ful. In my view, this is an impor­tant les­son from the pan­dem­ic report­ing. But it also means that we need new com­pe­ten­cies in some parts of the news­rooms, for exam­ple, to eval­u­ate sci­en­tif­ic experts. 

Will Bild TV become the “Heart­land” of tomorrow?

That depends on whether the edi­to­r­i­al team places jour­nal­is­tic prin­ci­ples and rel­e­vance cri­te­ria at the cen­tre of its work – or whether it fol­lows oth­er inter­ests and cam­paigns. That, of course, is some­thing to watch. I find it excit­ing that cur­rent or even live report­ing is to be giv­en such a promi­nent place in the new pro­gramme. Pro­vid­ing the nec­es­sary con­text which makes it pos­si­ble to cor­rect­ly clas­si­fy the top­ics of the day is cer­tain­ly a major challenge.

Where are the var­i­ous dis­sent­ing opin­ions? How are these opin­ions being pro­tect­ed? Or are they get­ting lost in the flood of information?

In a plu­ral­is­tic soci­ety that pro­tects free­dom of the press and free­dom of opin­ion, it is fun­da­men­tal that minori­ties be heard. Accord­ing­ly, there are var­i­ous pre­cau­tions to ensure diver­si­ty of opin­ion in jour­nal­ism – for exam­ple, through con­cen­tra­tion con­trol or the appro­pri­ate require­ments for both pub­lic and pri­vate broad­cast­ing. In this respect, such pro­tec­tion is pro­vid­ed for – in the­o­ry – but we must indeed be care­ful to keep spaces avail­able for minor­i­ty per­spec­tives and posi­tions. This is espe­cial­ly true for media offer­ings that are sup­posed to be prof­itable and are there­fore par­tic­u­lar­ly prone to focussing on large or afflu­ent tar­get groups. 

Ama­zon has devel­oped in part due to the online sale of books. They sell dig­i­tal books and audio books for their Kin­dle. With the help of Big Data analy­sis, Ama­zon analy­ses what their clients pur­chase, read, which sen­tences and chap­ters they under­line, quote and which para­graphs they read again. Ama­zon has now pur­chased a pub­lish­ing house and wants to pro­duce tai­lor-made con­tent based on their clien­tele analy­ses to raise their sales. Is that a mod­el for journalism?

Cer­tain­ly not. Jour­nal­ism that com­plete­ly fol­lows what its read­ers want to hear and opti­mis­es its con­tent sole­ly for sales would make itself irrel­e­vant. On the con­trary, we should empha­sise what jour­nal­ists are still need­ed for today – for ques­tions that no one has asked before, for an empa­thet­ic view of the world and a cre­ative approach to forms of expres­sion that can evolve as a result. Of course, edi­to­r­i­al offices should use new tech­nolo­gies to stream­line their process­es in mean­ing­ful places, to enable research involv­ing large amounts of data and to gen­er­ate for­mats for new plat­forms. But we should not leave the con­tent cre­ation and the eval­u­a­tion of top­ics to algo­rithms. If jour­nal­ism is to be the place where social dis­course is mod­er­at­ed, then it should be shaped by soci­ety – and not exclu­sive­ly accord­ing to mea­sur­able cri­te­ria that can be processed by algorithms.

Wikipedia has replaced ency­clopae­dias. The pub­lish­ing hous­es of ency­clopae­dias used to have trained employ­ees who researched, eval­u­at­ed and processed facts, knowl­edge etc. This group of employ­ees dis­ap­peared and every lay­man can now add knowl­edge and pseu­do- sci­ence to the largest and best-known dig­i­tal ref­er­ence book. Wikipedia makes efforts to check entries, erase untruths or duly endorse such con­tri­bu­tions. Is this trans­for­ma­tion an improvement?

The idea behind Wikipedia is bril­liant in my opin­ion – a par­tic­i­pa­to­ry ency­clopae­dia that is close­ly inter­twined with today’s soci­ety, gives space to niche knowl­edge and thus reflects much more than a stan­dard ency­clopae­dia ever could. Of course, this is asso­ci­at­ed with a dif­fer­ent qual­i­ty stan­dard: A Wikipedia entry can at best serve as a sec­ondary source, that is, as a guide to cor­re­spond­ing pri­ma­ry sources. There­fore, in my view, the two for­mats are not tru­ly com­pa­ra­ble. Nev­er­the­less, it is of course excit­ing to observe how Wikipedia has devel­oped and what chal­lenges it is cur­rent­ly fac­ing – such as the calls for a com­pre­hen­sive reform, includ­ing more trans­paren­cy in fund­ing and in the author­ship of con­tri­bu­tions. That would cer­tain­ly be desir­able, as would efforts for more diver­si­ty and gen­der equal­i­ty with­in Wikipedia.

What is jour­nal­ism doing to com­bat fake news, agno­tol­ogy, manip­u­la­tion, and shrink­ing bud­gets to pre­serve its credibility?

Cur­rent­ly, new mod­els and for­mats are being devel­oped and test­ed all over the media sys­tem to keep jour­nal­ism rel­e­vant and ensure its cred­i­bil­i­ty. There­fore, it is not at all easy to answer the ques­tion in an over­ar­ch­ing way. What I would like to address in this con­text is the user-based devel­op­ment of new prod­ucts and for­mats. This focus­es on the con­crete needs of the audi­ence so that edi­tors have to deal more inten­sive­ly with the real­i­ty of their audi­ence’s lives. I am con­vinced that this approach can help media to secure trust, stay rel­e­vant and build good rela­tion­ships with their read­ers. Espe­cial­ly in a dig­i­tal pub­lic sphere, where jour­nal­ism must always stand out from some­times aggres­sive com­peti­tors, these rela­tion­ships mat­ter. And of course, it needs to be com­pre­hen­sive­ly researched, crit­i­cal­ly exam­ined and qual­i­ta­tive­ly pre­pared con­tent – with­out it, even the smartest for­mat can­not be successful.

Are peo­ple, the read­ers of news, los­ing their abil­i­ty to dis­cern between actu­al and fake facts, rumours, opin­ions etc, giv­en the back­ground of con­stant infor­ma­tion flood?

We cer­tain­ly need new skills to fil­ter and clas­si­fy the infor­ma­tion we are con­front­ed with every day. Jour­nal­ists can help by not only pub­lish­ing their sto­ries, but also by mak­ing the process of their inves­ti­ga­tion trans­par­ent and explain­ing how they have come to their con­clu­sions. In my view, how­ev­er, we should also deep­en media edu­ca­tion, start­ing at schools, but even­tu­al­ly address­ing all age groups. This edu­ca­tion is impor­tant because many meth­ods and tools from the jour­nal­is­tic research process also help in every­day life, for exam­ple, when it comes to assess­ing alleged facts that are shared on social media plat­forms. If more peo­ple know how the work is being done in edi­to­r­i­al offices, this not only strength­ens trust in qual­i­ty jour­nal­ism, but also fos­ters their own judgement.

Is there a fight for opin­ions and truth tak­ing place on the of jour­nal­ism home front? How can jour­nal­ism safe­guard its inde­pen­dence in the dig­i­tal age?

I see jour­nal­ism, as I have said before, as an insti­tu­tion that gives space to rel­e­vant social dis­cours­es and mod­er­ates them. In this respect, a com­pe­ti­tion of opin­ions and a crit­i­cal search for truths are both essen­tial parts of this sys­tem. I would­n’t call it a bat­tle – but when I think of the dig­i­tal net­work pub­lic sphere as a whole, there are clear ten­den­cies towards manip­u­la­tion and polar­i­sa­tion, and per­son­al attacks have become stan­dard in many dig­i­tal dis­course spaces, espe­cial­ly against minori­ties. This ter­rain there­fore seems more prob­lem­at­ic to me, espe­cial­ly if it is more dif­fi­cult to influ­ence it in a con­struc­tive direc­tion. But since jour­nal­ism is a rel­e­vant actor in this sphere, we nat­u­ral­ly have a shared respon­si­bil­i­ty and must do our part. How­ev­er, this only works if we main­tain our inde­pen­dence, and this chal­lenge is indeed even more com­plex in the dig­i­tal sphere. It is not enough to sim­ply pub­lish arti­cles and sup­ply them to sub­scribers. Instead, we have to deal with a com­plex world of diverse plat­forms, actors and con­stant­ly chang­ing con­texts. In par­tic­u­lar, if we want to reach new, younger tar­get groups, we have to be present where they are active. It is thus cru­cial to be clear­ly recog­nis­able – and to empha­sise the val­ues of an inde­pen­dent, crit­i­cal, val­ue-ori­ent­ed journalism.

Prof. Elmer, thank you for shar­ing your reflec­tion on the changes in journalism.

Thank you, Dr Cal­daro­la, and I look for­ward to read­ing your upcom­ing inter­views with rec­og­nized experts, delv­ing even deep­er into this fas­ci­nat­ing topic.