Supporting FAIR data: categorization of research data as a tool in data management

Abstract

The demand for implementation of the FAIR data principles is in many cases difficult for a researcher to adhere to in efficient ways due to lacking tools. We suggest categorizing data in a more extensive and systematic way with focus on the inherent properties of the data as means to enhancing research data services. After discussing different approaches to categorizing data, we propose a tripartite research data categorization based around the inherent aspect of stability. The three research data types are operational data, generic research data and research data publications. Generic research data is validated data and can be cumulative, i.e. data can be added without versioning, however if it is dynamic it should be versioned. Generic research data should be separated from immutable dataset publications that are published for reasons of reproducibility of specific research results.

Keywords

• research data management
• data management
• FAIR principles
• persistent identifiers [http://www.yso.fi/onto/yso/p29018]
• data citation
• open science