Modern research is more and more data driven. In theory, the data could be reused. But often when reusing existing data the unclear legal aspects, such as missing or undefined licensing terms, hinders researchers and the progress of their research projects. When datasets are compiled out of multiple datasets, provenance and especially legal status of the resulting data can be hard to infer, more so for large datasets. Thus, the resulting complexity slows or even prevents research. (Research) Software faces the same problems. To make data more accessible and more reusable and thus FAIR, the complexity should be reduced as much as possible to allow researchers to focus on their actual research projects instead of just researching licensing.
The legal challenges researchers face in their work are manifold: Is the reuse of existing data allowed? Do the terms fit the intended purposes of the research project? When writing software, what libraries can they use and in which way? Publishing either data or software is even more complex due to the influence of dependencies on license choice.There are many follow-up questions depending on the specific use cases. There could be special requirements for a data use agreement which are not covered by standard licenses. There could be terms and conditions which exclude text and data mining or limit the use of data to a certain geographical area. All that while terms should be respected as much as possible and even European regulation partially requiring machine-interpretable terms in data spaces.
The Working Group on Licensing and Rule Automation (WG LARA) aims to address these problems with automating as much of the process as possible while building on existing solutions. Such solutions include the LicenseChecker for evaluating legal implications of software dependencies or the ODRL Builder to easily create machine-interpretable licensing terms along a human-readable version of the terms, but there are also many more. Yet those tools are neither widely adapted nor used. Being part of the NFDI Section ELSA we want to raise awareness and adoption of these and newer tools. Currently the Working Group focuses on checking the dependencies for software projects and creating custom data usage agreements for data in both a machine- and human-readable representation. The machine-readable version builds on the ODRL Standard. With using this standard we ensure that the legal information can easily be used by platforms and in automated or agentic tasks. We want to integrate our solutions with each other and also into existing data-management frameworks like FAIR Digital Objects (FDOs).
A few things are out of scope, however. This includes checking whether the claimed rights regarding the data or software really exist, as this task is nearly impossible to automate and usually needs input from a legal expert. We also do not aim to develop new standard licenses or propose a solution for enforcement.
