Data Management Ontology

Every Horizon 2020 project is generating large volumes of useful diverse data, which has to be managed properly in accordance with the guidelines given by the European Commission. As IRES, is highly involved in data management activities in various EC funded projects and in order to deal with the highly diversified data across different projects, an attempt was made during the past months to develop a Data Management Ontology. This ontology is expected to create a solid structure for the metadata schema that will be used to describe present and future data sets, while allowing for the execution of useful queries that will extract valuable insights regarding the projects. A first draft of the ontology was developed using the tool Protégé, while the WebVowl online tool is used for the visualization. The elementary entity of our ontology is the dataset. The term dataset refers to a volume of data that is sufficient to effectively describe something (a process, a system, etc.). The dataset may come from various diverse disciplines, such as physics, data science, economics etc. and is described by generic metadata that apply to all disciplines. This way, all different data is treated in a homogenous manner, compatible with the FAIR data principles. In order to include information regarding the different scientific fields, existing specialized knowledge representation systems will be considered, such as domain specific ontologies. The aim for the data management ontology is to be rather minimal and functional, and in agreement with existing data management and domain specific practices. Balancing and connecting the general and the specialized concepts is considered to be our main challenge, given the high diverse data we are expecting to be dealing with. The created ontology will be designed to most effectively capture the information regarding each data set of an EC funded project. Therefore, the application of the FAIR data principles (especially Interoperability and Reusability) is of utmost importance. This implies, that a lot of metadata regarding storage, access, format, project, partner etc. has to be generated. All these general concepts apply more or less to all kinds of data and will be captured under a top – level ontology. On the other hand, each data set is associated with some discipline or scientific field and is generated or collected using specific methods or processes. Additionally, it often involves specific measurement equipment or possible s