1. Introduction

The National Library of Medicine’s (NLM) Dataset Metadata Model (DATMM) is a linked data scheme designed to support a national catalog of biomedical datasets. As a slim linked data model, DATMM provides sufficient descriptive metadata to find datasets of potential interest across biomedical disciplines and to access those datasets at their homesites. For complete information about the creation and application of DATMM please see: DATMM Application Profile.

Current Version: 2.1.0

Date: 2024-04-15

Previous Version: 2.0.0

Objective

Biomedical datasets are available on the Web from a multitude of disparate sites and use a multitude of local metadata schema. NLM seeks to make these datasets more easily findable by describing them with a single metadata scheme, so they are searchable and accessible from a single site. This is similar to the concept of PubMed where articles from many medical journals can be searched and accessed from a single site.

NLM found no existing metadata schema that put datasets themselves at the center of the design, providing a simple and concise description of each dataset and a home site link for more information and access. Therefore, NLM decided to create its own linked data scheme to describe and provide access to biomedical datasets in disparate sites across the Web. Rather than creating something entirely new, the goal was to design a linked data model that largely adapts and makes of use of portions of existing schema and could itself be either extended by or incorporated into other linked data schema. The resulting DATMM scheme provides a concise set of metadata for describing, finding, and accessing biomedical datasets in their home sites, as well as linking to publications about those datasets.

Please see the DATMM white paper for more information about the development of DATMM.

Scope

As a Resource Description Framework (RDF) model, DATMM includes only a few new Classes and relies heavily on re-use of Classes and properties from other RDF schema, such as Dublin Core, Schema.org, BIBFRAME, FOAF, et al. DATMM is not expected to remain entirely static; it may occasionally change to accommodate new or differing data needs. Indeed, while experimenting with the conversion of external site metadata to DATMM, a Collection Class for multiple datasets that exist as part of a single study was added and a Distribution Class for directly downloading datasets was deleted.

Download Schema

Download the current version of DATMM in Resource Description Framework (RDF) format. 

DATMM RDF