An important mission of the KIM project is long-term reproducibility of atomistic simulations, i.e. the ability to replicate a computation using exactly the same code and interatomic model used by the original researcher. A prerequisite for this is the versioned archival storage provided by openkim.org for interatomic models and a persistent mechanism for citing this online content.
The Digital Object Identifier (DOI) system is the industry standard for citing online content. A DOI is a unique identifier defined by a standard of the International Organization for Standardization (ISO). The DOI is associated with metadata describing the item and a web address (URL) where the content is located.
DOIs are issued by a number of registration agencies associated with the International DOI Foundation including DataCite, Crossref, EIDR, and others. The registration agencies have member institutions that collect the necessary information and create DOIs for specific items. As of June 2018, KIM is a member of DataCite and can issue DOIs, see here.
In addition to reproducibility, there are other benefits to obtaining a DOI for an interatomic model (or other KIM content):
Increased visibility. DOIs can be searched through the DataCite Search Interface. In addition, DataCite works with several aggregators and catalogs worldwide including Google, Clarivate, BASE, OpenAIRE and others to improve search results for its DOIs and metadata.
Citation Metrics. DataCite collects citation metrics for its DOIs from its indexing partners as part of the Make Data Count Project, which helps gauge the impact of an item. This is information that developers can put on their website or CV.
We encourage developers of interatomic models to upload their potentials to openkim.org in order to obtain DOIs for their intellectual property. This service is provided at no charge. Further, we encourage researchers engaged in molecular simulation to make use of interatomic models with DOIs and to provide this information in their published work.
A DOI name consists of two parts, a prefix and a suffix separated by the "/" character. The prefix identifies the naming authority (OpenKIM in our case) and the suffix identifies the content (an interatomic model, model driver, test driver, or verification check, etc.). OpenKIM DOIs have the following form:
xxxxxxxx is an 8-character (case insensitive) alphanumeric string.
Each OpenKIM DOI is associated with an item's metadata
(title, description, contributor, source citations, etc.) and the permanent
URL on openkim.org where the item is located.
KIM content with a DOI can be cited in publications, for example:
Ackland, G. J. (2018). Finnis-Sinclair potential (LAMMPS cubic hermite tabulation) for Ti for the hcp-fcc transition developed by Ackland (1992) v005. OpenKIM. https://doi.org/10.25950/276be3c4
The permanent URL https://doi.org/10.25950/276be3c4 resolves to the content on openkim.org.