Slowly as a Euro­pean Open Sci­ence Cloud

Con­cerns over slow progress of ini­tia­tive to share sci­en­tific in­for­ma­tion. David Matthews re­ports

THE (Times Higher Education) - - CONTENTS - David.matthews@timeshigh­ere­d­u­ca­tion.com

It’s the € 6.7 bil­lion (£5.9 bil­lion) project that no one has heard of. The Euro­pean Open Sci­ence Cloud prom­ises to al­low re­searchers, busi­nesses and the pub­lic to share and re­use the ter­abytes of data gen­er­ated by modern sci­ence and the big data rev­o­lu­tion.

Rather than leav­ing their data stored on a mem­ory stick in a drawer, re­searchers will open it up to their peers across the con­ti­nent. This should im­prove the re­li­a­bil­ity of find­ings, al­low datasets to cross dis­ci­plinary bound­aries and free up en­trepreneurs to cre­ate the datadriven dig­i­tal prod­ucts of the fu­ture.

Or at least that’s the plan. Con­ceived by the Euro­pean Union in 2016, the cloud is set to be­come a re­al­ity by 2020, with pi­lots al­ready run­ning.

But there are fears that progress is too slow and that, with re­searchers lack­ing in­cen­tives to sign up, it may not have the trans­for­ma­tive ef­fect that Brussels hopes for.

“The ques­tion is: are we fast enough? I am afraid that we are not,” warned Ge­org Schütte, state sec­re­tary at the Ger­man Fed­eral Min­istry of Ed­u­ca­tion and Re­search at the Open Sci­ence Con­fer­ence 2018 in Ber­lin.

Ger­many has been build­ing na­tional data and re­search in­fra­struc­ture for more than three years, he said, and is only about half­way through.

At the Euro­pean level, it will per­haps take even longer, Dr Schütte pre­dicted. If only up and run­ning by 2024-25, “the world will have changed fun­da­men­tally” and Europe will have moved too slowly to cap­ture the sci­en­tific and eco­nomic ben­e­fits of open data, he told del­e­gates.

To try to un­der­stand the project is to strug­gle through a thicket of work­ing groups, acronyms and organograms. But one thing to grasp is that “cloud” is more a metaphor for a “seam­less” com­mons of data across the con­ti­nent, as the Euro­pean Com­mis­sion puts it – it does not pri­mar­ily mean the build­ing of new com­puter in­fra­struc­ture. In­stead, at least in the early stages of the project, the chal­lenge is to “fed­er­ate what ex­ists”, ac­cord­ing to Jean-Claude Burgel­man, head of the Brussels unit in charge of the cloud. “There are 2,000 repos­i­to­ries in Europe and they all want to be part of it,” he said.

Nor is it go­ing to be ex­clu­sively Euro­pean. The cloud is part of a global push to­wards mak­ing data “FAIR”: find­able, ac­ces­si­ble, in­ter­op­er­a­ble and re­us­able. This in­volves mak­ing sure data and meta­data have such things as unique iden­ti­fiers and clear reusage li­cences.

Karel Luy­ben, the na­tional co­or­di­na­tor for open sci­ence in the Nether­lands, and one of the driv­ing forces be­hind FAIR, said that he was talk­ing with re­search or­gan­i­sa­tions in the US, Aus­tralia and some Asian coun­tries. These prin­ci­ples of open data need to reach be­yond Europe, “oth­er­wise it’s not go­ing to work”, he warned del­e­gates.

Open-mind­ed­ness

The cloud is part of a wider push for “open sci­ence”, the younger cousin of the open-ac­cess move­ment. If open ac­cess is about mak­ing fin­ished ar­ti­cles pub­licly ac­ces­si­ble, open sci­ence fo­cuses on open­ing up un­der­ly­ing data and meth­ods.

One hope is that this could halt the so-called “re­pro­ducibil­ity cri­sis” in re­search: by mak­ing data ac­ces­si­ble, it should be­come more ob­vi­ous when sci­en­tists have spun their re­sults to get a big head­line find­ing. But why would re­searchers open them­selves up to more scru­tiny for no ap­par­ent gain?

Sci­en­tists are be­ing asked “to act in the in­ter­ests of sci­ence as a whole and the com­mu­nity in­ter­est”, said Sarah Jones, an as­so­ciate di­rec­tor at the Dig­i­tal Cu­ra­tion Cen­tre, which helps re­searchers take care of their data. “A lot of those re­wards for in­di­vid­u­als aren’t there yet.”

Uni­ver­si­ties need to re­ward their aca­demics ca­reer-wise for mak­ing data open, she ar­gued. Open sci­ence re­quire­ments are likely to be­come manda­tory in the EU’s next re­search frame­work from 2021, she added.

There is also a gen­eral lack of aware­ness. Ac­cord­ing to a sur­vey of re­searchers last year, less than 15 per cent had heard of the cloud, her pre­sen­ta­tion said. This is a “re­ally big con­cern for us”, she ex­plained.

One fi­nal hur­dle is money. Although costs will vary from dis­ci­pline to dis­ci­pline, Pro­fes­sor Luy­ben es­ti­mated that data open­ness will con­sume 5 per cent of to­tal re­search ex­penses.

The com­mis­sion has es­ti­mated that the cloud will cost € 6.7 bil­lion of pub­lic and pri­vate in­vest­ment, with € 2 bil­lion alone set to come out of the Hori­zon 2020 re­search pro­gramme bud­get.

So the bar­ri­ers are size­able but the costs of fail­ure are much big­ger, speak­ers warned. If the EU does not to build an all-en­com­pass­ing data cloud quickly, the fear is that pri­vate pub­lish­ers will get there first.

“There are lots of very strong play­ers in the mar­ket ready to take up the fail­ure of pub­lic pol­icy,” said Dr Burgel­man.

Only smoul­der­ing the Euro­pean Open Sci­ence Cloud might not have its de­sired im­pact if it is not up and run­ning soon

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.