Florian Gaffron, Friederike Klan, Alsayed Algergawy
In contrast to other data infrastructures, the BExIS 2 platform pursues a more structured approach to store and describe research data. This ensures the interpretability of the data and thus enables re-usability and automatic integration of the research data within a repository. The flipside of this strategy is that it imposes higher requirements to the researchers who provide data and also leads to more effort at the upload time. In the current version of the BExIS 2 platform, researchers first have to create a data structure in order to upload a dataset. After that, an Excel template can be downloaded, filled with data and finally uploaded to the system. In particular, the creation of the data structure seems to be challenging for data providers, that typically do not have a computer science background. The process as it is implemented at the moment is also time-consuming, since the data have to be reorganized and copied in order to make them available to the system. Hence, we strive for a more user-centered upload approach that supports the automatic extraction and annotation of primary- and metadata from user-designed spreadsheets. In our presentation, we will introduce the envisioned upload mechanism and present a basic implementation of the underlying workflow. In its current version, the process requires a great deal of user involvement. This will be gradually reduced by automatizing parts of the workflow in upcoming efforts.