A FAIR Digital Object Lab Software Stack

Preprocessing data for research, like finding, accessing, unifying or converting, takes up to large parts of research time spans. The FAIR (Findability, Accessibility, Interoperability, Reusability) principles aim to support and facilitate the (re)use of data, and will contribute to alleviating this problem. A FAIR Digital Object (FAIR DO) captures research data resources of all kinds (raw data, metadata, software, …) in order to align them with the FAIR principles.

FAIR Digital Objects are expressive, machine-actionable pointers to research data. As such, each FAIR DO points to one research data object. Additionally, they may link to other FAIR DOs, explaining their relations. The FAIR Digital Object Lab is an extendable and adjustable architecture (a software stack) for generic FAIR Digital Object tasks. It consists of a set of interacting components with services and tools for creation, validation, discovery, curation, and more. In this talk, we will present our plans for the FAIR DO Lab and explain our decisions, which are mostly based on the experience gained in previous developments.

The creation and maintenance of FAIR DOs is not trivial, as their persistent identifiers (PIDs) contain typed record information. ... mehrWhen creating or maintaining PID records of FAIR DOs, the required information has to be validated, involving calls to a public Data Type Registry (DTR). After a successful validation, the information has to be transformed to a representation of a PID service. After a FAIR DO has been registered successfully, the PID should be documented locally and disseminated. Using these PIDs as a starting point, tools may use the machine-actionability of FAIR DOs to maintain search indexes or to create collections. This enables researchers to look up PIDs by searching for record information or timestamp.

We are developing a set of services, offering a solution to support these use-cases, which we call the FAIR DO Lab. Its goal is to have a production-ready and configurable software stack, easing the development of FAIR-DO-aware tools and services by offering at least the described use-cases. We have already gained some experience by its predecessor, the FAIR DO Testbed, which was introduced at the Research Data Alliance (RDA) Virtual Plenary 17 Poster Session. The Lab will be configurable similar to the Testbed, as each service can be omitted or replaced to satisfy specific needs while integrating the Lab on top of existing research infrastructures.