TDWG Standards Metadata

Title: Biodiversity Information Standards (TDWG) Standards Metadata

Date Version Issued: 2026-11-08

Date Created: 2017-10-10

Part of TDWG Standard: Not part of any standard

This version: http://rs.tdwg.org/index/2026-11-08

Latest version: http://rs.tdwg.org/index

Previous version: http://rs.tdwg.org/index/2021-02-01

Abstract: The https://github.com/tdwg/rs.tdwg.org repository contains the source CSV files containing the metadata necessary to generate human- and machine-readable representations of the various components of Biodiversity Information Standards (TDWG) standards in accordance with the TDWG Standards Documentation Specification). This document is the human-readable representation of the TDWG Standards data catalog. Dereferencing the data catalog as RDF is described in a section of this document.

Contributors: Steve Baskauf (TDWG Technical Architecture Group, TDWG Audiovisual Core Maintenance Group, TDWG Darwin Core Maintenance Group)

Bibliographic citation: Biodiversity Information Standards. 2026. Biodiversity Information Standards (TDWG) Standards Metadata http://rs.tdwg.org/index

1 Metadata model for TDWG standards

Fig. 1.A. TDWG Standards hierarchy model (image updated from original diagram in the Standards Documentation Standard GitHub repository)

Fig. 1.B. TDWG Standards hierarchy model represented as an ER diagram.

Section 2.2 of the TDWG Standards Documentation Specification defines a hierarchical model for components of TDWG Standards (Figs. 1.A. and B.). At the highest level (1st level) are Standards. All TDWG Standards are linked by virtue of a common publisher relationship with TDWG itself. Documents and Vocabularies (2nd level) are the parts of Standards. Vocabularies are composed of Term Lists (3rd level), which correspond roughly to namespaces that define sets of terms. The Term Lists are composed of Terms (4th level), which may be Properties, Classes, or Concepts.

Each of the levels in the standards hierarchy are linked to the level above by a isPartOf relationsip, and to the level below by a hasPart relationship.

Fig. 2. Version model used to describe standards-related resources (image from the Section 2.3 of the TDWG Standards Documentation Specification)

Any resource in the hierarchy above can be considered to consist of an abstract resource that persists over time, referred to as a “current resource”. In Fig. 2, the arrow represents a current resource named ex:resource. A current resource has a creation date, and a last modified date.

The state of current resources over time is documented by versions of that resource. The versions are “snapshots” of the resource at a particular instant in time. In Fig. 2, the vertical lines represent versions - for example the version named ex:resource-2009-07-18.

The state of most properties of a current resource at the present time are shared by the most recent version of that resource. The properties of a current resource may change when a new version is issued. The issued date of the new version becomes the modified date of the current resource.

A current resource is related to each of its versions by a hasVersion relationship. A version of a resource is related to its current resource by an isVersionOf relationship. A version is related to a previous version by a replaces relationship and to a subsequent version by an isReplacedBy relationship.

2 IRI patterns

2.1 Patterns for current resources in the standards hierarchy

The IRI denoting Biodiversity Information Standards (TDWG) as an organization is:

https://ror.org/03k171441

1st level: IRIs denoting TDWG standards

are in the form:

http://www.tdwg.org/standards/nnn

where “nnn” is a number unique to a particular standard. Example for the Darwin Core Standard:

http://www.tdwg.org/standards/450

2nd level: IRIs denoting vocabularies

are in the form:

http://rs.tdwg.org/vvv/

where “vvv” is a string unique to the vocabulary. For example, “dwc” is the string for the basic Darwin Core vocabulary and “ac” is the string for the Audiovisual Core vocabulary. Example for the Darwin Core main vocabulary:

http://rs.tdwg.org/dwc/

2nd level: IRIs denoting standards documents

are in the form:

http://rs.tdwg.org/sss/doc/docname/

where “sss” is a sequence of alphabetic characters representing the standard and “docname” is a short series of alphabetic characters representing the doucment. IRI examples are:

http://rs.tdwg.org/ac/doc/structure/

for the Audiovisual Core Structure document and

http://rs.tdwg.org/pnbd/doc/specification/

for the specification document of the Plant Names in Botanical Databases standard.

Some Darwin Core documents have legacy IRIs that do not follow this IRI pattern.

3rd level IRIs denoting term lists

are in the form:

http://rs.tdwg.org/vvv/sss/

where “sss” is a string representing the term list that is unique within vocabulary “vvv”. IRI examples are:

http://rs.tdwg.org/dwc/iri/

for Darwin Core IRI-value Terms and

http://rs.tdwg.org/ac/terms/

for terms defined by Audiovisual Core.

Note: In cases where term lists have been moved to a different vocabulary, this pattern may not be followed. For example, the eco: namespace terms have been moved from the Humboldt Vocabulary to the main Darwin Core vocabulary. But for stability the term list IRI (http://rs.tdwg.org/eco/terms/) has not been changed.

4th level IRIs denoting vocabulary terms

are in the form:

http://rs.tdwg.org/vvv/sss/ttt

where “ttt” is the local name of a term defined as part of the term list http://rs.tdwg.org/vvv/sss/. IRI examples are:

http://rs.tdwg.org/dwc/iri/recordedBy
http://rs.tdwg.org/ac/terms/caption

Term IRIs are not followed by a trailing forward slash. The “namespace” corresponding to the defining term list can be abbreviated to form a compact IRI (CURIE). Since dwciri: abbreviates http://rs.tdwg.org/dwc/iri/ and ac: abbreviates http://rs.tdwg.org/ac/terms/, the two term IRIs given as examples above can be expressed as the CURIEs dwciri:recordedBy and ac:caption.

2.2 Patterns for versions of resources in the standards hierarchy

1st level: Standards

A standard with IRI in the form

http://www.tdwg.org/standards/nnn

has versions with IRIs in the form

http://www.tdwg.org/standards/nnn/version/yyyy-mm-dd

where yyyy-mm-dd represents the issued date of the version, using the lexical space of an XML schema date datatype (without timezone indicator). Example for the Darwin Core Standard:

http://www.tdwg.org/standards/450/version/2015-03-27

2nd level: Vocabularies

A vocabulary with IRI in the form

http://rs.tdwg.org/vvv/

has versions with IRIs in the form

http://rs.tdwg.org/version/vvv/yyyy-mm-dd

with yyyy-mm-dd as described for standards. Example for the Darwin Core basic vocabulary:

http://rs.tdwg.org/version/dwc/2014-12-23

2nd level: Standards Documents

A document in the form:

http://rs.tdwg.org/sss/doc/docname/

has version IRIs in the form

http://rs.tdwg.org/sss/doc/docname/yyyy-mm-dd

with yyyy-mm-dd as described for standards. Example for the Standards Documentation Specification:

http://rs.tdwg.org/sds/doc/specification/2017-04-25

3rd level: Term Lists

A term list with IRI in the form

http://rs.tdwg.org/vvv/sss/

has versions with IRIs in the form

http://rs.tdwg.org/vvv/version/sss/yyyy-mm-dd

Example for the Darwin Core IRI-value term list:

http://rs.tdwg.org/dwc/version/iri/2015-03-27

4th level: Terms

A term with IRI in the form

http://rs.tdwg.org/vvv/sss/ttt

has versions with IRIs in the form

http://rs.tdwg.org/vvv/sss/version/ttt-yyyy-mm-dd

Example for a version of dwciri:recordedBy:

http://rs.tdwg.org/dwc/iri/version/recordedBy-2015-03-27

3 About directories in this repository

3.1 Relationship of parts of this repository to the metadata model

Each directory in this repository contain metadata for a group of resources at some level in the hierarchy. The data described in each directory can be considered a dcat:dataset as described by the W3C Data Catalog Vocabulary (DCAT).

Directories whose names do not end in “-versions” contain metadata about current resources. Directory names ending in “-versions” contain metadata for versions of the current resources described in the corresponding folders that do not end in “-versions”.

3.2 Files contained in current resources directories

Within each current resource directory there is a file whose name corresponds to the containing directory, plus the “.csv” file extension. That file contains the primary metadata about the resources described in that folder. For example, in the terms directory, the file terms.csv is the primary metadata file. The corresponding file with name ending in “-column-mappings.csv” maps the columns in the primary metadata file to standard metadata properties (similar in function to the meta.xml file of a Darwin Core Archive). Example: terms-column-mappings.csv.

The current resource directory also contains two files that contain link tables that describe one-to-many relationships between the current resource and related resources. The file with name ending in “-replacements.csv” (example: terms-replacements.csv) links current resources to other current resources that replace them. The file with name ending in “-versions.csv” (example: terms-versions.csv) links versions to their corresponding current term.

There are other files in each directory that contain configuration information or other information necessary to generate the machine readable metadata for the resources described in that directory.

3.3 Files contained in versions directories

Within each folder describing versions, the primary metadata are in the file whose name corresponds to the containing directory, plus the “.csv” file extension. For example, in the directory terms-versions, the file terms-versions.csv is the primary metadata file. The file with name ending in “-column-mappings.csv” (example: terms-versions-column-mappings.csv) maps the columns of the primary version metadata file to standard properties. The file with name ending in “-replacements.csv” (example: terms-versions-replacements.csv) is a link table that links versions to versions that replace them. Other files in the directory contain configuration or other information needed to generate machine readable metadata for the versions described in the directory.

3.4 Retrieving machine readable metadata from the datasets

3.4.1 Retrieving machine readable metadata as RDF

The IRI http://rs.tdwg.org/index denotes the TDWG Data catalog. When dereferenced with an Accept: header of text/html, this document should be served. When dereferenced requesting text/turtle, application/rdf+xml, or application/ld+json, the data catalog should be served in the RDF/Turtle, RDF/XML, and JSON-LD machine-readable serializations respectively. (The machine-readable data catalogs can be retrieved using the explicit IRIs http://rs-test.tdwg.org/index.ttl, http://rs-test.tdwg.org/index.rdf, and http://rs-test.tdwg.org/index.json.) Note that as of 2020-02-03 the JSON-LD representation does not conform completely to the JSON-LD specification and should not be relied upon as a source of machine-readable metadata. It can, however, be consumed as generic JSON.

The links in the machine-readable metadata can be followed to determine a dcat:downloadURL to retrieve a dump of each of the datasets included in the data catalog. The set download URLs can be dereferenced to obtain the complete set of TDWG standards metadata. The general form of a dataset dump is

http://rs.tdwg.org/dump/datasetname

where “datasetname” is the name of the directory containing the dataset.

For more information, see this blog post.

To access machine-readable translations metadata, see the Vocabulary Translations page. Note: As of 2026-11-08, translations for Darwin Core and Audiovisual Core are automated using the CrowdIn system. The files linked here are maintained for historical purposes, but are not being updated. For more information, contact a member of the Darwin Core or Audiovisual Core Maintenance Groups.

3.4.2 Retrieving machine readable metadata as CSV

Because all of the kinds of RDF serializations of TDWG standards metadata are generated from the source CSV files in this repository, and because the CSV files can be accessed programatically from GitHub as raw files, the CSV files themselves represent a type of machine-readable serialization of the those metadata. However, unlike RDF, the CSV files cannot be fully understood in isolation.

For most purposes, the primary metadata CSV file within a current resources directory (see Section 3.2) will contain the core metadata about that resource category. However, the column headers for that CSV cannot be understood without also referring to the corresponding “-column-mappings.csv” file within that directory, since the “-column-mappings.csv” table maps column headers to property CURIEs, and establishes the value type and language tag or datatype for the column. Additionally, the relationship of the current resource to its versions requires an additional tables as described in Section 3.2.

4 Relationships of directories to resources in the TDWG Standards model

4.1 Metadata about ratified TDWG Standards

Note: The authoritative and up-to-date data about existing term lists, their IRI patterns, and directories are at https://github.com/tdwg/rs.tdwg.org/blob/master/term-lists/term-lists.csv.

4.2 Utility metadata not governed by TDWG Standards

4.3 Metadata about obsolete pre-standard versions of Darwin Core

* The decisions directory also contains a CSV file that links vocabulary terms with the decisions that affected them.

** It does not seem necessary at the present time to maintain versions of the decisions themselves, although there are versions of the decisions list. This could be changed in the future if necessary.

4.4 Other directories in the rs.tdwg.org repository not directly associated with the standards hierarchy