Every academic discipline now works with digital research data. Yet questions invariably arise about how to ensure that the troves of generated data remain available for further research after the project ends. It is a topic that keeps Christian Bittner and Ben Stöver, members of an interdisciplinary nine-person team at the Research Data Service Center, perpetually busy: “Science is most trustworthy when it's transparent,” Bittner feels. Which is why researchers have a strong self-interest in making their data accessible. The service center is operated jointly by the Bonn University and State Library and University IT, with the mission of supporting researchers at the University of Bonn with research data management.
The center promotes the FAIR principle: Date should be easily Findable, and as Accessible as possible (i.e. no complicated hurdles to accessibility). The data should ideally be saved in open source or widely available data forms (Interoperable). And finally, it should be Reusable. The first step is clear: generous license terms and through documentation of how the data was generated are musts.
Some disciplines have more experience with data than others
“Researchers have been managing data since the dawn of research itself,” Stöver says. “We provide professional support from the start, even when the framework conditions shift.” It is an acknowledgment that requirements continue to get stricter. Alongside the guidelines for good research practices, many sources of funds often apply strong transparency conditions and open access stipulations. In other words, all paths lead through good data management. Researchers need to factor in extra time and effort in the early phases of a project. In particular, consideration needs to be given in advance to optimal structuring of work with the data. In the long term, however, this planning benefits both the project and its participants.
Research data management is still not universally taught as part of basic academic training. As a result, many researchers feel overwhelmed at the requirements. “Yet many seemingly complex processes can actually be resolved relatively easily,” Bittner assures. In the same way, fears can generally be assuaged quickly when it comes to protecting sensitive data even while making it available. The center conducts roughly 80 consultations per year, involving anything from doctoral projects to long-term projects that very much benefit from external impulses and group-think on how to improve research data management at a structural level.
Faith in data management
The pair of research data gurus have been working with data management since their own graduations: Stöver with a doctorate from University of Münster, where his research in bioinformatics led him to create software related to the fundamental processes of research. “Research always makes more sense to me when there are potential reuse scenarios for the data,” Stöver says.
Christian Bittner earned his doctorate in Geography at the University of Erlangen, with a concentration on social sciences issues. He experienced data management issues first hand: “I worked very intensively with data and stepped into pretty much every possible data management trap there is,” he remembers. On the side, he was also involved with the establishment of a geodata center.
Tools and data storage for better cooperation
The service center’s range of offerings is broad. In addition to its consulting work, the team also offers a regular schedule of training and is developing new IT services as well. Perhaps the most crucial of these is the new research data infrastructure that will supplement Sciebo (a cloud storage service) and the UniVM service (offering virtual machines). It allows teams or individual researchers to store and process large volumes of data securely—the system is currently working with 2.5 Petabytes (2500 Terabytes) of replicated network storage, and more will be added soon.
There is also a service for electronic lab journals as well as RADAR, a research data repository. This latter service accommodates data for publication and long-term archiving based on the FAIR principles. The data is assigned a Digital Object Identifier (DOI), which provides a permanent citation location for researchers to use in their publications lists. Other services are also under preparation, such as Gitlab and a JupyterHub to support code-based research work.
Note
Need a consultation on research data? forschungsdaten.uni-bonn.de