An article co-authored by one of the THOR partners was recently published by the Earth Science Informatics journal. Although written from a geo-domain perspective it contains much that can be applied to more general cases.
Beneath a review section on the introduction, history and subsequent use of DOIs in geosciences, the article also contains some interesting ideas on how to deal with persistent identifiers for data sets which may change over time (dynamic data sets, versioning) and for subsets of data sets.
Abstract: The first minting of Digital Object Identifiers (DOI) for research data happened in 2004 in the context of the project “Publication and citation of primary scientific data” (STD-DOI). Some of the concepts and perceptions about DOI for data today have their roots in the way this project implemented DOI for research data and the decisions made in those early days still shape the discussion about the use of persistent identifiers for research data today. This project also laid the foundation for a tighter integration of journal publications and data. Promoted by early adopters, such as PANGAEA, DOI registration for data has reached a high level of maturity and has become an integral part of scientific publishing. This paper discusses the fundamental concepts applied in the identification of DOI for research data and how these can be interpreted for alternative and future applications of persistent identifiers for research data.
Klump, J., R. Huber, and M. Diepenbroek (2015), DOI for geoscience data – how early practices shape present perceptions, Earth Sci. Inform., doi:10.1007/s12145-015-0231-5.
New DOI resolution methods are an important issue for THOR. In cooperation with the RDA working group on dynamic data citation we will prototype a service to resolve a DOI name directly to a stored data subset.