A new study (doi: 10.1162/dint_a_00034) by Larry Lannom (CNRI), Dimitris Koureas (Naturalis Biodiversity Center) and Alex Hardisty (Cardiff University) examines how Digital Object Architecture, a logical extension of the globally adopted Internet Architecture addresses the challenges of actively stewarding the aggregation and virtualization of widely distributed data about biological and geological specimens in natural science collections.
More than 3 billion specimens housed, organized and catalogued in the world’s natural science collections represent an unparalleled source of information for assisting to answer scientific questions related to solving global challenges in climate and the environment. However, exploiting this material today is slow, inefficient and expensive. By creating ‘digital specimens’ through digitization initiatives such as DiSSCo in Europe and iDigBio in North America it’s possible to open and expand possibilities for wider, more flexible and meaningful access for a varied range of science and policy uses.
Digital Object Architecture (DOA) is a particularly appropriate approach for this because it has several features that help to address the key challenges around making such data ‘findable, accessible, interoperable, and reusable’ – the so called ‘FAIR’ Principles. First, being able to reference digital specimens reliably and persistently over time and beyond changes in technology is crucial to making things findable. Second, once found, a single communications protocol, called ‘Digital Object Interface Protocol’ (DOIP) in conjunction with data about the digital specimen to be accessed – so called ‘metadata’ – can be used to access, retrieve and update/operate on the data. A notion of defined ‘data types’ within DOA directly addresses the interoperability requirement. Data of different types cannot easily be made to interoperate together so when data types are known, transformations can be applied automatically to deliver coherent seamless data from multiple sources. Lastly, all the principles of DOA – persistent identifiers, digital objects, metadata, communication protocol, data types – come together to make data managed through the DOA approach reusable.
The Digital Object Architecture approach has been adopted as the technical basis of Distributed System of Scientific Collections (DiSSCo), a priority pan-European research infrastructure aiming to unify more than 110 individual natural science collections across 21 countries into a single open and FAIR scientific information resource.
This work has been partly funded by the EU Horizon 2020-funded ICEDIG project.
The study article: L. Lannom, D. Koureas & A.R. Hardisty. FAIR data and services in biodiversity science and geoscience. Data Intelligence 2(2020), 122–130. doi: 10.1162/dint_a_00034.
The FAIR Data Principles: https://www.force11.org/group/fairgroup/fairprinciples
Digital Object Architecture (DOA) and Digital Object Interface Protocol (DOIP): https://www.dona.net