Skip to content

MIDAS Catalog Metadata Schema

LinkML representation of the MIDAS Coordination Center catalog's JSON-LD metadata model. Stage 1 covers the Thing identity/naming base and the shared CreativeWork slice inherited by every resource type.

URI: https://w3id.org/midas-catalog/schema

Name: midas-catalog

Classes

Class Description
ConformsTo A metadata standard/specification this resource's metadata conforms to (Schema.org, Bioschemas, Dublin Core, DCAT, ...). Standalone node typed as dcterms:Standard; deliberately NOT a Thing.
DataDownload A distribution of a dataset, modeled as an access point: the landing page is its dcat:accessURL (and schema:url). It does not link a direct file (no contentUrl/downloadURL), so it is deliberately not a schema:DataDownload. encodingFormat/mediaType describe the data available through it. Anonymous (no identifier) — projects as an inlined blank node (PropertyValue/Location precedent); the media-type IRI is a value (mediaType), not the node's identity.
Location A postal address for an Organization. Standalone node (not a Thing).
MediaType Reference to an IANA media type by its IRI — dcterms:MediaType is dcat:mediaType's declared range. Identifier-only and referenced with inlined:false, so it renders as a plain IRI string in JSON / Pydantic / JSON-Schema but as an @id resource node in JSON-LD. This stub exists because gen-jsonld-context only emits @type:@id for class-ranged slots (a uri/uriorcurie range stays an xsd:anyURI literal); it is the generator-native way to give RDF/DCAT consumers a resource without changing the string every JSON consumer sees. No local metadata — the media type is defined by IANA.
PropertyValue A structured characteristic (name/value) used in additionalProperty to carry dataset characteristics with no dedicated schema.org property, each dispatch-keyed by propertyID. Standalone node (not a Thing) — these are anonymous structured values with no IRI of their own, so they project as inlined blank nodes; the term IRI lives on valueReference. Mirrors the Location precedent.
ProvActivity The activity that generated a resource (PROV-O).
ProvAgent A PROV-O agent (creator/author); id/name/url only.
ProvEntity A PROV-O entity (this resource, or a source it derived from). license is a schema:CreativeWork node (range License: @id + name + url), the same representation as sourceLicense/DataDownload; rights is a plain usage-statement literal under the global rights slot (emitted key "rights"; predicate per the @context — confirm in the regenerated context).
Thing Root of the type hierarchy. Carries the identity and naming slots shared by every node that appears in a record's JSON-LD — the top-level resource and every nested object (Person, Organization, Place, Language, DefinedTerm, License, etc.). Modeled as its own class, not folded into CreativeWork, because those nested classes are Things but not CreativeWorks.
        Agent A person or organization — used for creator. Concrete subclass is selected per node by agentType (curated is_person flag: true -> Person, otherwise Organization).
                Organization An organization — used for publisher, funder, and the MCC fallback creator.
                Person A person — used for dataset/software creators, with optional identifier (e.g. ORCID) strings; id/name inherited from Thing.
        ComputerLanguage A programming language (software programmingLanguage / learning-resource programmingLanguage). name + sameAs from Thing; no additional slots.
        CourseInstance An offering/instance of a Course. Extends Thing (consistent with Person/Place/Language). courseMode carries an IRI when the mode has one, else the mode name.
        CreativeWork Shared base for every MIDAS resource type (Dataset, SoftwareApplication, DataCatalog, LearningResource, and the Collection subtypes). Every slice below inherits these slots; type-specific slots are added by the subtype classes in later slices.
                DataCatalog A data catalog a resource is included in (the MIDAS Catalog or another). Common CreativeWork fields only; It is dual-typed (@type [schema:DataCatalog, dcat:Catalog]).
                        Repository A repository (type-or-record 7): a specialized DataCatalog. Emits as dual-typed via inheritance.
                Dataset A dataset (type-or-record 1).
                LearningResource An educational resource (type-or-record 16).
                        Course A course (educational resource with course=true); adds a course instance.
                Literature A literature collection (type-or-record 8). Emits as schema:Collection.
                        DiseaseSpecificCollection A disease-specific collection (type-or-record 10): Literature with genre="disease-specific collection"; no additional slots.
                        ModelingStudy A modeling-studies collection (type-or-record 9): Literature with genre="modeling studies"; no additional slots.
                RelatedData A related resource (publication, dataset, source page, related course), emitted as a CreativeWork-typed node. The kind of relationship is carried in its description; no slots beyond the inherited CreativeWork set.
                SoftwareApplication A software application (type-or-record 2). rights is inherited from CreativeWork.
                        Dashboard A dashboard (type-or-record 6). Emits as schema:SoftwareApplication with applicationCategory="Dashboard" set at runtime; no additional slots.
        DefinedTerm A controlled-vocabulary term (topic, biological scale, type-or-record, educational term, ...). It has a dual type set to the SKOS Concept type. The SKOS crosswalks below are INSTANCE-LEVEL — each term carries its own exactMatch/broadMatch IRI — which is exactly why they are slots on this class and NOT slot-level exact_/broad_mappings on the referring properties. (The source class carries a code comment stating the same.)
        Language A language of the resource content: an ISO 639-2 code as name plus, when available, a Lexvo ISO 639-3 sameAs cross-link — both inherited from Thing (no additional slots).
        License A license governing the catalog metadata or source data. Emitted as a CreativeWork-typed node (per its @JsonldType) carrying name/url from Thing; no additional slots.
        OperatingSystem An operating system the software runs on; id = the dbpedia resource IRI, name from Thing. (schema.org's OperatingSystem class — currently "new"/feedback area.)
        Place A geographic place identified by its GeoNames Concept URI, dual-typed as schema:Place and geonames:Feature. Carries the GeoNames feature code and feature class.
WebResource Reference to a web resource by its IRI — a landing page, access URL, or endpoint (dcat:accessURL / dcat:landingPage range over rdfs:Resource). Same identifier-only, inlined:false pattern and rationale as MediaType: a plain IRI string for JSON consumers, an @id resource node for RDF/DCAT consumers.

Slots

Slot Description
about Subject matter of the content. Contains DefinedTerm objects (topics, biological scales, types).
accessibleForFree Whether the resource can be accessed at no cost to the user.
accessUrl Access point for the distribution: the landing page where the data can be browsed and obtained (DCAT allows a landing page as accessURL).
additionalProperty Aggregated dataset characteristics with no dedicated schema.org property, each a self-describing PropertyValue keyed by propertyID to its MIDAS category: geographic resolution (midas03), topic (midas2), digital object type (midas10).
additionalType Additional rdf:type-style classification (used by SoftwareApplication and Literature).
address Street address.
agentType Designates the concrete Agent subclass of this node (schema:Person or schema:Organization), set from the curated collection_creator.is_person flag.
applicationCategory High-level application category. Also the runtime discriminator for Dashboard (emits as schema:SoftwareApplication).
applicationSubCategory More specific application sub-category.
audience Intended audience. Default shape is the LearningResource form (DefinedTerm, multivalued); Literature overrides to a single string via slot_usage.
bibliographicCitation Recommended citation for the resource (authorship, title, date, URL) in standard academic citation format.
broadMatch SKOS broad-match crosswalk IRI. Instance-level (per term), not a slot-level mapping.
catalogLicense License governing the metadata catalog itself (the hardcoded default), separate from the license of the source data.
city City / locality.
codeRepository URL of the source-code repository.
competencyRequired Prerequisite competencies.
conformsTo Established standards or specifications that this metadata adheres to (e.g., Schema.org, Bioschemas, Dublin Core, DCAT).
country Country.
courseMode Delivery mode(s) of a course instance (e.g. online, in-person, MOOC). Emits the mode's IRI when available, else the mode name (iri-over-name) — so values are strings that MAY be IRIs; modeled as string (not uriorcurie) because the name fallback is not an IRI.
creator Agent(s) who created or compiled the data — each a Person or an Organization, selected per node by agentType from the curated is_person flag (may include ORCID identifiers). Always emitted as an array; a single-element set holding the MIDAS Coordination Center Organization when no creators are recorded. Range is Agent, the Person/Organization supertype, so the generated JSON Schema accepts either subclass.
dataRequirement Data prerequisite(s) for the resource (LRMI vocabulary).
dateModified Date the resource was last modified or updated. ISO 8601 (YYYY-MM-DD). Populated from the collection creation end date.
datePublished Last date the resource was made available (date of publication, or of the most recent update/modified version), from the curated Last Publication Date. ISO 8601 (YYYY-MM-DD). Absence is meaningful: a blank value denotes a regularly/automatically-updated resource, and MIDAS never defaults it.
description Brief description of the data. For datasets: what is included (variables, measurements), method/origin. For software: what it does, access requirements, supported languages.
displayName Human-readable display label for the term.
distributions Distributions of the dataset, one per available data format, each an access point (dcat:Distribution) whose accessURL is the landing page. (Field is plural; IRI is singular schema:distribution.)
educationalLevel Educational level(s) the resource targets.
educationalObjective Learning objectives (LRMI vocabulary).
educationalUse Intended educational use(s).
email Contact email for the organization.
encodingFormat Media/encoding format. Shared by DataDownload (single, per distribution) and SoftwareApplication (multivalued, formats the software reads/writes).
endpointUrl Web page where users can browse information about the dataset and access download links. Canonical landing-page slot (the authoritative landing-page predicate for external consumers).
exactMatch SKOS exact-match crosswalk IRI. Instance-level (per term), not a slot-level mapping.
externalIdentifiers External identifier strings (DOIs, URLs, etc.) that reference this resource in other systems. Shares schema:identifier with identifier.
featureClass GeoNames feature class (e.g. A, P). CURIE-compresses under the geonames prefix.
featureCode GeoNames feature code (e.g. PCLI). CURIE-compresses under the geonames prefix.
funder Organization that funded the creation or maintenance of this resource.
generated Entity generated by the activity.
genre Genre of the collection. Also the runtime discriminator for ModelingStudy ("modeling studies") and DiseaseSpecificCollection ("disease-specific collection").
hasCourseInstance An offering/instance of the course.
id Unique identifier (IRI/URL) for this resource. In JSON-LD, @id represents the globally unique identifier. (Maps to @id; the identifier slot is bound to @id by gen-jsonld-context.)
identifier Primary persistent identifier for this resource (e.g., DOI, Handle, ARK, or persistent URL). Can be a simple string or PropertyValue object.
identifiers External identifier strings for the person (e.g. ORCID). Shares schema:identifier with identifier/externalIdentifiers. Curator-recommended ("Creator Identifier — ID"); the backing registry entry carries the 6-arg recommended flag. Although this is a nested Person-level slot, recommending it is consistent with the model's identifiability stance: creator is recommended, and identifier/externalIdentifiers (resource-level) are recommended, so recommending the creator-level identifier extends the same principle rather than introducing a new one. The flag emits as flat property documentation (same as all recommended slots); it does not encode the "nested under a recommended creator" conditionality, which is accepted as an adequate representation. NOTE: there is no separate identifier-TYPE slot on Person, so the "Creator Identifier — Type" recommendation has no distinct emitted slot.
includedInDataCatalog A collection of datasets in which this resource is included. Links to the MIDAS Catalog or other data catalogs.
inDefinedTermSet IRI of the controlled vocabulary / term set the term belongs to.
inLanguage Primary language(s) of the resource content. Language objects with ISO 639-2 codes; each also carries a Lexvo ISO 639-3 sameAs cross-link when available.
instructor Instructor(s) of a course, carried on the CourseInstance (schema:instructor is defined only on CourseInstance, not on Course/LearningResource). Populated from the collection's creator field when the educational resource is a course with delivery modes; for non-course educational resources that field maps to schema:creator instead. Range is Person per schema.org; MIDAS cannot guarantee the party is a person, so an organization named as instructor still emits as a Person node. No MCC fallback (unlike creator): absent when no instructor is recorded.
keywords Topic labels as plain text (schema:keywords), for keyword-based discovery (e.g. Google Dataset Search). Topic names only — the same subject fact also appears structured under about.
learningResourceType Type/format of the learning resource.
license A schema:CreativeWork license node (id/name/url), used by provenance entities (ProvEntity) and DataDownload. Distinct slot from catalogLicense/sourceLicense but the same range (License) and predicate (schema:license).
location Postal address of the organization.
mediaType IANA media type of this distribution as a resolvable IRI (dcat:mediaType). Projected as an IRI reference — mirrors accessUrl/endpointUrl.
name Title or name of the digital object, page, or resource.
operatingSystems Operating system(s) the software runs on, as OperatingSystem nodes (id = dbpedia resource IRI).
pavCreatedBy Agent responsible for creating the resource (authorship-specific counterpart to prov:wasGeneratedBy).
pavDerivedFrom Source resources from which this version was derived, one IRI per source (versioning emphasis). Multivalued: PAV permits repeated derivation, and a record built from several sources derives from all of them.
places Place objects with GeoNames references describing geographic locations covered by the resource. Shares schema:spatialCoverage with spatialCoverage.
programmingLanguage Programming language(s). Shared by SoftwareApplication and LearningResource.
propertyID IRI of the characteristic (category) a PropertyValue represents — the immutable dispatch key (midas10 digital object type, midas2 topic, midas03 geographic resolution).
provider Organization(s) providing the resource.
publisher Organization responsible for making the resource available. For MIDAS Catalog resources, the MIDAS Coordination Center.
relatedData RelatedData (CreativeWork) objects describing publications, datasets, or other resources related to this resource. The kind of relationship is carried per-object in each RelatedData's description.
rights Usage rights and conditions for accessing or using the resource (e.g., 'Registration required', 'Academic use only').
sameAs URL of a reference web page that unambiguously indicates the item's identity (e.g., ORCID profile, Wikidata entry). NOTE: deliberately NOT mapped to dcat:landingPage — although CreativeWork populates sameAs from the landing page, this is a generic Thing slot (a Person's sameAs is an ORCID, not a landing page), so the landing-page equivalence is builder behavior, not a slot-level fact.
sdDatePublished Date MIDAS published this structured-data record (when curation of the record was completed). A fact about the metadata, not the dataset: it is never used as the dataset's datePublished. ISO 8601 (YYYY-MM-DD).
sourceLicense License governing the original source data or related resources. Shares schema:license with catalogLicense (distinct slots, same predicate).
standardType Type/category of the standard.
standardVersion Version of the conformance standard this record was produced under. Set only on the MIDAS Catalog JSON-LD Schema standard entry, sourced from MidasCatalogJsonLdSchema.VERSION.
state State / region.
subjectAreas Subject areas the standard covers.
teaches Concepts/skills the resource teaches.
temporalCoverage Date range covered by the data. ISO 8601 interval (YYYY-MM-DD/YYYY-MM-DD, or YYYY-MM-DD/.. for ongoing).
timeRequired Approximate time to complete, as an ISO 8601 duration.
url URL to the website that contains and/or describes the digital object or person.
used Entities the activity used.
value The characteristic's value (e.g. the resolution level or topic label).
valueReference The value as a controlled-vocabulary DefinedTerm carrying its own IRI (when minted) and inDefinedTermSet membership.
version Version identifier of the resource (numeric or named, e.g., '1.0', 'v2.3').
wasAssociatedWith Agents associated with the activity.
wasAttributedTo Agent the entity was attributed to.
wasDerivedFrom Source entity from which this resource was derived. Multivalued on CreativeWork (see its slot_usage) where a record may derive from several sources; single-valued on ProvEntity.
wasGeneratedBy Activity that generated this resource. Links to a prov:Activity describing the creation process.
zipcode Postal / ZIP code.

Enumerations

Enumeration Description

Types

Type Description
Boolean A binary (true or false) value
String A character string
Uri a complete URI
Uriorcurie a URI or a CURIE

Subsets

Subset Description