Persistent Data, Sustainable Information

In almost all academic fields, results are derived from found evidence such as digitized objects persisted at a repository. Deriving results from such repositories can be time and cost intensive, as data is often difficult to reuse. Guidelines such as FAIR (Findability, Accessibility, Interoperable, and Reuse) disseminate proper persisting of data so that it can be more easily reused. However, we argue that even if one follows FAIR guidelines, it is still challenging to derive results by reusing data. First of all, before reusing data from any repository, one needs to find the data. Search engines can index data stored at repositories, as data is associated with metadata as proposed by the FAIR guidelines. Deciding whether the data found is relevant usually requires downloading, extracting and visualizing the entire dataset which is time-consuming and costly. We propose that data need to be persisted by associating it with metadata that determines how data can be prepared to support decision making. Metadata links data with executable code that can create an information system from associated data on demand. With our solution, data is easier to reuse, as one can decide whether found data is relevant. In addition, if the data is found to be relevant, our information system allows researchers to clearly refer to specific regions in the data for data governance. Hence, if one persists her or his data, then one gets an sustainable information system almost for free.