Activities

We develop ontologies to formalise and capture knowledge, and we support other researchers in creating their own ontologies. Currently we are co-developing the Helmholtz Digitization Ontology (HDO) in HMC. HDO produces harmonised descriptions for concepts related to digital data and their handling. We also develop the Dislocation Ontology Suite (DISOS) which describes materials science concepts and the representation of dislocations in materials.

We explore quantitative metrics to measure the FAIRness (Wilkinson et al. 2016) of published data sets and repositories. Currently, we are using the F-UJI tool by FAIRisFAIR and applied it to Helmholtz repositories. We contribute to the development of this tool and plan to capture metadata to analyze the evolution of FAIRness over time.

Together with Hub E&E & Hub Matter we run HMC’s unHIDE initiative that aims to improve metadata quality across Helmholtz and consolidate metadata into the Helmholtz Knowledge Graph (KG). Currently we lead the technical development of a prototype for the Helmholtz KG. This includes developing of tooling such as harvesters and data integration tools. We also develop and customize the unHIDE user interface that allows exploration of the data in the Helmholtz KG.

We developed the course “Fundamentals of scientific metadata: why context matters”, a hands-on training course that teaches early-career researchers the basics of structured data documentation. We teach the course regularly and are currently preparing the publication of the course material on The Carpentries Incubator.

Jointly with Hub Matter and a group of domain experts, we develop the EM Glossary. Our goal is to agree on consistent definitions for the most relevant concepts in electron microscopy. These can be use by other initiatives to align their schemas and ontologies with. Interested? Come and join our regular meetings!

To demonstrate the potential of metadata to be used during the research process, we develop Beaverdam. This tool allows the exploration of large data sets to identify specific sub-sets. To do this, it combines metadata from many experiments and provides and provides an interactive user interface to explore the entire collection. We develop Beaverdam with our colleagues from the INM-6 at Forschungszentrum Jülich.

Many instruments come with proprietary software that stores important metadata within binary files. To allow extraction of this inaccessible information, we develop the tool MARBLE together with our colleagues from IEK-2 at Forschungzentrum Jülich. This tool supports resarchers to decipher data and metadata from binary files. We organise events to bring together the potential user community and teach them how to use the Marble interface.