FAIR Data (Day) — it takes a village!

Authors: Mar­ta Teperek, Kim­ber­ley Zwiers, Iulia Popes­cu, Saman­tha Willem­sen, Mar­jan Grootveld, Maria Cruz, Ruben Kok 

On the 29th of Novem­ber 2022 Research Data Nether­lands (RDNL) and the Nation­al Pro­gramme Open Sci­ence (NPOS) joint­ly organ­ised the ‘FAIR Data Day’ to award the bian­nu­al Dutch Data Prizes and to cel­e­brate the advance­ments of FAIR Data in the Nether­lands.

It was a fan­tas­tic event that was bustling with peo­ple and filled with inspir­ing keynote speak­ers, com­mu­ni­ty-led work­shops, thought-pro­vok­ing dis­cus­sions, over­joyed prize win­ners and an over­all vibrant atmos­phere. 

Dutch Data prizes

This year was the 7th anniver­sary of the bian­nu­al Dutch Data Prize com­pe­ti­tion. Peter Doorn, the for­mer direc­tor of DANS and one of the ini­tia­tors of the com­pe­ti­tion, gave us an inspir­ing overview of the impor­tance of the com­pe­ti­tion in reward­ing and cel­e­brat­ing FAIR datasets. 

The 2022 edi­tion of the Dutch Data prize com­pe­ti­tion received a record num­ber of 51 nom­i­na­tions that were even­ly spread between the three research domains: Social Sci­ences and Human­i­ties, Life Sci­ences and Health, and Nat­ur­al and Engi­neer­ing Sci­ences. The jury, chaired by Car­o­line Viss­er, vice chair of the Board of the Dutch Research Coun­cil (NWO), had the dif­fi­cult task of iden­ti­fy­ing one clear win­ner in each cat­e­go­ry. Car­o­line reflect­ed that the jury was impressed by the over­all qual­i­ty of the sub­mis­sions, the impact made by the datasets, and how Find­able, Acces­si­ble, Inter­op­er­a­ble and Re-usable (FAIR) they were. She called FAIR a shared respon­si­bil­i­ty of the entire sci­ence field, point­ing to one of the key nation­al ambi­tions in Open Sci­ence for the next decade.

The win­ner in each cat­e­go­ry received a large round of applause, a 3,500 EUR check and a tro­phy for their FAIR dataset. 

• In Life Sci­ences and Health, the prize went to Duong Vu and the ‘DNA bar­codes for fun­gal iden­ti­fi­ca­tion’ dataset, who togeth­er with her team devel­oped a data­base of over 24k fun­gal species.
• In Social Sci­ences and Human­i­ties, Coos­je Veld­kamp and the team behind the ‘YOUth cohort’ dataset were the win­ners with a study that fol­lowed the neu­rocog­ni­tive devel­op­ment of near­ly 4,000 Dutch chil­dren from preg­nan­cy until ear­ly adult­hood.
• In Nat­ur­al and Engi­neer­ing Sci­ences, Mitchell Van Zui­jlen and the team behind the “Mate­ri­als In Paint­ings (MIP)” dataset won with their anno­tat­ed dataset that uses machine learn­ing algo­rithms to clas­si­fy frag­ment­ed pieces of 19,000 paint­ings.

Unfor­tu­nate­ly, the win­ners in the Nat­ur­al and Engi­neer­ing Sci­ences domain could not be present to phys­i­cal­ly receive their prize in the after­noon. How­ev­er, 4TU.ResearchData data cura­tor Jan van der Heul was present to col­lect the award on their behalf. Inter­est­ing­ly, Jan van der Heul has not only curat­ed their dataset, but also played a key role in the FAIR­ness of the win­ning dataset with­in the NES domain in 2020.

Image on the left: Jan van der Heul, data cura­tor at 4TU.ResearchData, col­lect­ing the prize on behalf of the team behind “Mate­ri­als in Paint­ings (MIP)” dataset. Cred­it: Yan Wang.

Forget about egos: behind every achievement, there is a team

On the left: Shali­ni Kura­p­ati dur­ing her keynote ses­sion.

What stood out as the com­mon thread dur­ing the day and across all the win­ners is that behind every impres­sive dataset, every out­stand­ing achieve­ment, every new tool or intro­duced pol­i­cy, there was an entire team of ded­i­cat­ed peo­ple that had worked hard to achieve it. Shali­ni Kura­p­ati, the CEO of Clear­box-AI who opened the day with a cap­ti­vat­ing keynote speech about syn­thet­ic data, start­ed by say­ing: “it is not about me; it is about us”. We are a team! This thread con­tin­ued through­out the day and was beau­ti­ful­ly empha­sised by Car­o­line Viss­er dur­ing the Dutch Data Prize award cer­e­mo­ny. It is not about the egos and glo­ry. It is about col­lab­o­ra­tion, about knowl­edge cre­ation and the impact on soci­ety. And to be the most impact­ful and to effec­tive­ly tack­le chal­lenges, we need to work togeth­er. 

The work­shop led by Prof. Serkan Gir­gin and the team intro­duced the Jypter­FAIR tool that was devel­oped thanks to the NWO Open Sci­ence Fund. Jupyter­FAIR stream­lines the process of upload­ing and down­load­ing datasets to/from repos­i­to­ries. What was remark­able is that at the start of the pre­sen­ta­tion Serkan men­tioned and acknowl­edged all the col­leagues who were involved in the project. He con­clud­ed with an inspir­ing call to action: he invit­ed all the work­shop atten­dees, their com­mu­ni­ties and beyond to become mem­bers of the project and co-design, or sim­ply con­nect with oth­ers.

So we clear­ly need each other’s skills and exper­tise. We need researchers, the con­tent experts, we need data stew­ards, research soft­ware engi­neers, com­mu­ni­ty man­agers, project lead­ers, eth­i­cal experts: togeth­er we can do bet­ter and achieve more. The most beau­ti­ful illus­tra­tion of the pow­er of col­lab­o­ra­tion was when the win­ner of the Data Prize for the Social Sci­ences and Human­i­ties domain (the YOUth cohort study) was announced: instead of just one win­ner, the entire team rushed to the stage! 

Support staff also need to be recognised and rewarded

How­ev­er, to make teams work and ful­ly ben­e­fit from the diverse back­grounds or skills that dif­fer­ent con­trib­u­tors have to offer, every­one needs to feel val­ued and appre­ci­at­ed. And there is clear­ly work to be done in this area. Bar­ry Fitzger­ald, the chair­per­son of the day, asked a few ques­tions to the audi­ence via Men­time­ter. One of the ques­tions was whether data stew­ards felt appre­ci­at­ed at their organ­i­sa­tions. The results clear­ly indi­cat­ed that more can be done for data stew­ards to feel appre­ci­at­ed. 

Men­time­ter sur­vey results

Sad­ly, what is true for the data stew­ards, also holds true for oth­er pro­fes­sion­al sup­port staff. At many research organ­i­sa­tions, employ­ees are split into two cat­e­gories: ‘aca­d­e­m­ic staff’ and ‘sup­port staff’. How­ev­er, as has been argued else­where, “well-func­tion­ing teams rely on the shar­ing of respon­si­bil­i­ties and cred­it. For research to advance and progress, diverse per­son­nel must be able to con­tribute their tal­ent and skills with­out being too restrict­ed by con­ven­tion­al hier­ar­chies (…) Research insti­tu­tions need to fos­ter col­lab­o­ra­tive envi­ron­ments that empow­er prob­lem solv­ing and build mutu­al trust and respect for skills and exper­tise, regard­less of job titles and per­ceived rank­ing and sta­tus.” 

As beau­ti­ful­ly stat­ed by Gem­ma Der­rick and Simon Het­trick: “Nobel prize win­ners didn’t get there on their own”. There­fore, we need to stop think­ing in silos and start dis­cussing how to recog­nise and reward pro­fes­sion­al sup­port staff for their con­tri­bu­tions to the research process. To begin with, pro­fes­sion­al sup­port staff should be part of the dis­cus­sions on chang­ing recog­ni­tion and rewards sys­tems in acad­e­mia.

We don’t want an H‑index for data. Context matters.

The major­i­ty or our par­tic­i­pants seemed to agree that it is impor­tant to give cred­it where cred­it is due. How­ev­er, how should we reward data re-use? Can we objec­tive­ly mea­sure data re-use? Or should data re-use affect researchers’ H‑index/measurement of impact?

The reflec­tion shared by Maria Cruz from NWO seemed to res­onate well with the audi­ence: “We def­i­nite­ly don’t want to have an H‑index for data. We don’t want to end up with the same prob­lems as those cre­at­ed by the jour­nal pub­lish­ing sys­tem.” Maria also warned that down­loads, views and cita­tions can be mis­lead­ing. To deter­mine the real impact of research data, qual­i­ta­tive mea­sures are essen­tial. We need to under­stand the con­text in which datasets are gen­er­at­ed and shared in order to assess their val­ue or how they con­tribute to spe­cif­ic (research) com­mu­ni­ties, over­all knowl­edge cre­ation or to soci­ety at large. 

Our sec­ond keynote speak­er, Nadia Bloe­men­daal, one of the Dutch Data Prize win­ners of 2020, gave com­pelling exam­ples of how shar­ing of data on trop­i­cal cyclone risk made an impact on people’s liveli­hoods, risk pre­ven­tion and improv­ing the effec­tive­ness of human­i­tar­i­an response in affect­ed areas. Such sto­ries can­not be sim­ply expressed as the num­ber of down­loads or views.

The audi­ence was asked to reflect on ques­tions such as: “What does the reuse of data mean to you?”

FAIR Data Day — it takes a village

Ulti­mate­ly, we can con­clude that mak­ing data FAIR is a col­lab­o­ra­tive team effort and, also hijack­ing the title of a webi­nar organ­ised by NWO in Novem­ber 2021), that it takes a vil­lage to organ­ise the FAIR Data Day! 

With an inter­na­tion­al audi­ence of over 250 reg­is­tered par­tic­i­pants, more than 20 sub­mit­ted work­shop pro­pos­als and 50 nom­i­na­tions for the Dutch Data Prize, it was a tru­ly com­mu­ni­ty-led event. We are extreme­ly grate­ful to all the atten­dees, keynote speak­ers, work­shop providers, Dutch Data Prize jury mem­bers and nom­i­nees for a tru­ly splen­did event. This day would not have been pos­si­ble with­out all these con­tri­bu­tions. We also want to par­tic­u­lar­ly thank the pro­gramme com­mit­tee, organ­is­ing com­mit­tee and our col­leagues (Jan van der Heul, Mari­bel Bar­rera and Maaike Smit), who all helped and worked very hard to turn this day into a suc­cess.

It real­ly takes a vil­lage to make research data FAIR!

A frac­tion of the peo­ple involved in the organ­i­sa­tion of the FAIR Data Day.