JSON is a useful data serialization and messaging format. This specification defines JSON-LD 1.1, a JSON-based format to serialize Linked Data. The syntax is designed to easily integrate into deployed systems that already use JSON, and provides a smooth upgrade path from JSON to JSON-LD. It is primarily intended to be a way to use Linked Data in Web-based programming environments, to build interoperable Web services, and to store Linked Data in JSON-based storage engines.

This specification describes a superset of the features defined in [[[JSON-LD10]]] [[JSON-LD10]] and, except where noted, documents created using the 1.0 version of this specification remain compatible with JSON-LD 1.1.

This document has been developed by the JSON-LD Working Group and was derived from the JSON-LD Community Group's Final Report.

There is a live JSON-LD playground that is capable of demonstrating the features described in this document.

This specification is intended to supersede the [[[JSON-LD10]]] [[JSON-LD10]] specification.

Set of Documents

This document is one of three JSON-LD 1.1 Recommendations produced by the JSON-LD Working Group:

Introduction

Linked Data [[LINKED-DATA]] is a way to create a network of standards-based machine interpretable data across different documents and Web sites. It allows an application to start at one piece of Linked Data, and follow embedded links to other pieces of Linked Data that are hosted on different sites across the Web.

JSON-LD is a lightweight syntax to serialize Linked Data in JSON [[RFC8259]]. Its design allows existing JSON to be interpreted as Linked Data with minimal changes. JSON-LD is primarily intended to be a way to use Linked Data in Web-based programming environments, to build interoperable Web services, and to store Linked Data in JSON-based storage engines. Since JSON-LD is 100% compatible with JSON, the large number of JSON parsers and libraries available today can be reused. In addition to all the features JSON provides, JSON-LD introduces:

JSON-LD is designed to be usable directly as JSON, with no knowledge of RDF [[RDF11-CONCEPTS]]. It is also designed to be usable as RDF in conjunction with other Linked Data technologies like SPARQL [[SPARQL11-OVERVIEW]]. Developers who require any of the facilities listed above or need to serialize an RDF graph or Dataset in a JSON-based syntax will find JSON-LD of interest. People intending to use JSON-LD with RDF tools will find it can be used as another RDF syntax, as with [[Turtle]] and [[TriG]]. Complete details of how JSON-LD relates to RDF are in section .

The syntax is designed to not disturb already deployed systems running on JSON, but provide a smooth upgrade path from JSON to JSON-LD. Since the shape of such data varies wildly, JSON-LD features mechanisms to reshape documents into a deterministic structure which simplifies their processing.

How to Read this Document

This document is a detailed specification for a serialization of Linked Data in JSON. The document is primarily intended for the following audiences:

A companion document, the JSON-LD 1.1 Processing Algorithms and API specification [[JSON-LD11-API]], specifies how to work with JSON-LD at a higher level by providing a standard library interface for common JSON-LD operations.

To understand the basics in this specification you must first be familiar with JSON, which is detailed in [[RFC8259]].

This document almost exclusively uses the term IRI (Internationalized Resource Indicator) when discussing hyperlinks. Many Web developers are more familiar with the URL (Uniform Resource Locator) terminology. The document also uses, albeit rarely, the URI (Uniform Resource Indicator) terminology. While these terms are often used interchangeably among technical communities, they do have important distinctions from one another and the specification goes to great lengths to try and use the proper terminology at all times.

This document can highlight changes since the [[[JSON-LD10]]] version. Select to changes.

Contributing

There are a number of ways that one may participate in the development of this specification:

Typographical conventions

Terminology

This document uses the following terms as defined in external specifications and defines terms specific to JSON-LD.

Design Goals and Rationale

JSON-LD satisfies the following design goals:

Simplicity
No extra processors or software libraries are necessary to use JSON-LD in its most basic form. The language provides developers with a very easy learning curve. Developers not concerned with Linked Data only need to understand JSON, and know to include but ignore the @context property, to use the basic functionality in JSON-LD.
Compatibility
A JSON-LD document is always a valid JSON document. This ensures that all of the standard JSON libraries work seamlessly with JSON-LD documents.
Expressiveness
The syntax serializes labeled directed graphs. This ensures that almost every real world data model can be expressed.
Terseness
The JSON-LD syntax is very terse and human readable, requiring as little effort as possible from the developer.
Zero Edits, most of the time
JSON-LD ensures a smooth and simple transition from existing JSON-based systems. In many cases, zero edits to the JSON document and the addition of one line to the HTTP response should suffice (see ). This allows organizations that have already deployed large JSON-based infrastructure to use JSON-LD's features in a way that is not disruptive to their day-to-day operations and is transparent to their current customers. However, there are times where mapping JSON to a graph representation is a complex undertaking. In these instances, rather than extending JSON-LD to support esoteric use cases, we chose not to support the use case. While Zero Edits is a design goal, it is not always possible without adding great complexity to the language. JSON-LD focuses on simplicity when possible.
Usable as RDF
JSON-LD is usable by developers as idiomatic JSON, with no need to understand RDF [[RDF11-CONCEPTS]]. JSON-LD is also usable as RDF, so people intending to use JSON-LD with RDF tools will find it can be used like any other RDF syntax. Complete details of how JSON-LD relates to RDF are in section .

Data Model Overview

Generally speaking, the data model described by a JSON-LD document is a labeled, directed graph. The graph contains nodes, which are connected by directed-arcs. A node is either a resource with properties, or the data values of those properties including strings, numbers, typed values (like dates and times) and IRIs.

Within a directed graph, nodes are resources, and may be unnamed, i.e., not identified by an IRI; which are called blank nodes, and may be identified using a blank node identifier. These identifiers may be required to represent a fully connected graph using a tree structure, such as JSON, but otherwise have no intrinsic meaning. Literal values, such as strings and numbers, are also considered resources, and JSON-LD distinguishes between node objects and value objects to distinguish between the different kinds of resources.

This simple data model is incredibly flexible and powerful, capable of modeling almost any kind of data. For a deeper explanation of the data model, see section .

Developers who are familiar with Linked Data technologies will recognize the data model as the RDF Data Model. To dive deeper into how JSON-LD and RDF are related, see section .

At the surface level, a JSON-LD document is simply JSON, detailed in [[RFC8259]]. For the purpose of describing the core data structures, this is limited to arrays, maps (the parsed version of a JSON Object), strings, numbers, booleans, and null, called the JSON-LD internal representation. This allows surface syntaxes other than JSON to be manipulated using the same algorithms, when the syntax maps to equivalent core data structures.

Although not discussed in this specification, parallel work using [[[YAML]]] [[YAML]] and binary representations such as [[[RFC7049]]] [[RFC7049]] could be used to map into the internal representation, allowing the JSON-LD 1.1 API [[JSON-LD11-API]] to operate as if the source was a JSON document.

Syntax Tokens and Keywords

JSON-LD specifies a number of syntax tokens and keywords that are a core part of the language. A normative description of the keywords is given in .

@context
Used to define the short-hand names that are used throughout a JSON-LD document. These short-hand names are called terms and help developers to express specific identifiers in a compact manner. The @context keyword is described in detail in .
@direction
Used to set the base direction of a JSON-LD value, which are not typed values (e.g. strings, or language-tagged strings). This keyword is described in .
@id
Used to uniquely identify node objects that are being described in the document with IRIs or blank node identifiers. This keyword is described in . A node reference is a node object containing only the `@id` property, which may represent a reference to a node object found elsewhere in the document.
@value
Used to specify the data that is associated with a particular property in the graph. This keyword is described in and .
@language
Used to specify the language for a particular string value or the default language of a JSON-LD document. This keyword is described in .
@type
Used to set the type of a node or the datatype of a typed value. This keyword is described further in and .
The use of @type to define a type for both node objects and value objects addresses the basic need to type data, be it a literal value or a more complicated resource. Experts may find the overloaded use of the @type keyword for both purposes concerning, but should note that Web developer usage of this feature over multiple years has not resulted in its misuse due to the far less frequent use of @type to express typed literal values.
@container
Used to set the default container type for a term. This keyword is described in the following sections:
  • ,
  • ,
  • ,
  • ,
  • ,
  • , and
@list
Used to express an ordered set of data. This keyword is described in .
@set
Used to express an unordered set of data and to ensure that values are always represented as arrays. This keyword is described in .
@reverse
Used to express reverse properties. This keyword is described in .
@index
Used to specify that a container is used to index information and that processing should continue deeper into a JSON data structure. This keyword is described in .
@base
Used to set the base IRI against which to resolve those relative IRI references which are otherwise interpreted relative to the document. This keyword is described in .
@vocab
Used to expand properties and values in @type with a common prefix IRI. This keyword is described in .
@graph
Used to express a graph. This keyword is described in .
@nest
Used to define a property of a node object that groups together properties of that node, but is not an edge in the graph.
@none
Used as an index value in an index map, id map, language map, type map, or elsewhere where a map is used to index into other values, when the indexed node does not have the feature being indexed.
@prefix
With the value true, allows this term to be used to construct a compact IRI when compacting. With the value `false` prevents the term from being used to construct a compact IRI. Also determines if the term will be considered when expanding compact IRIs.
@version
Used in a context definition to set the processing mode. New features since [[[JSON-LD10]]] [[JSON-LD10]] described in this specification are not available when processing mode has been explicitly set to `json-ld-1.0`.
Within a context definition `@version` takes the specific value `1.1`, not `"json-ld-1.1"`, as a JSON-LD 1.0 processor may accept a string value for `@version`, but will reject a numeric value.
The use of `1.1` for the value of `@version` is intended to cause a JSON-LD 1.0 processor to stop processing. Although it is clearly meant to be related to JSON-LD 1.1, it does not otherwise adhere to the requirements for Semantic Versioning.
@json
Used as the @type value of a JSON literal. This keyword is described in .
:
The separator for JSON keys and values that use compact IRIs.
`@propagate`
Used in a context definition to change the scope of that context. By default, it is `true`, meaning that contexts propagate across node objects (other than for type-scoped contexts, which default to `false`). Setting this to `false` causes term definitions created within that context to be removed when entering a new node object.
`@protected`
Used to prevent term definitions of a context to be overridden by other contexts. This keyword is described in .
`@import`
Used in a context definition to load an external context within which the containing context definition is merged. This can be useful to add JSON-LD 1.1 features to JSON-LD 1.0 contexts.
`@included`
Used in a top-level node object to define an included block, for including secondary node objects within another node object.

All keys, keywords, and values in JSON-LD are case-sensitive.

A JSON-LD document complies with this specification if it follows the normative statements in appendix . JSON documents can be interpreted as JSON-LD by following the normative statements in . For convenience, normative statements for documents are often phrased as statements on the properties of the document.

This specification makes use of the following namespace prefixes:

Prefix IRI
dc11 http://purl.org/dc/elements/1.1/
dcterms http://purl.org/dc/terms/
cred https://w3id.org/credentials#
foaf http://xmlns.com/foaf/0.1/
geojson https://purl.org/geojson/vocab#
prov http://www.w3.org/ns/prov#
i18n https://www.w3.org/ns/i18n#
rdf http://www.w3.org/1999/02/22-rdf-syntax-ns#
schema http://schema.org/
skos http://www.w3.org/2004/02/skos/core#
xsd http://www.w3.org/2001/XMLSchema#

These are used within this document as part of a compact IRI as a shorthand for the resulting IRI, such as dcterms:title used to represent http://purl.org/dc/terms/title.

Basic Concepts

JSON [[RFC8259]] is a lightweight, language-independent data interchange format. It is easy to parse and easy to generate. However, it is difficult to integrate JSON from different sources as the data may contain keys that conflict with other data sources. Furthermore, JSON has no built-in support for hyperlinks, which are a fundamental building block on the Web. Let's start by looking at an example that we will be using for the rest of this section:

  
  

It's obvious to humans that the data is about a person whose name is "Manu Sporny" and that the homepage property contains the URL of that person's homepage. A machine doesn't have such an intuitive understanding and sometimes, even for humans, it is difficult to resolve ambiguities in such representations. This problem can be solved by using unambiguous identifiers to denote the different concepts instead of tokens such as "name", "homepage", etc.

Linked Data, and the Web in general, uses IRIs (Internationalized Resource Identifiers as described in [[RFC3987]]) for unambiguous identification. The idea is to use IRIs to assign unambiguous identifiers to data that may be of use to other developers. It is useful for terms, like name and homepage, to expand to IRIs so that developers don't accidentally step on each other's terms. Furthermore, developers and machines are able to use this IRI (by using a web browser, for instance) to go to the term and get a definition of what the term means. This process is known as IRI dereferencing.

Leveraging the popular schema.org vocabulary, the example above could be unambiguously expressed as follows:

In the example above, every property is unambiguously identified by an IRI and all values representing IRIs are explicitly marked as such by the @id keyword. While this is a valid JSON-LD document that is very specific about its data, the document is also overly verbose and difficult to work with for human developers. To address this issue, JSON-LD introduces the notion of a context as described in the next section.

This section only covers the most basic features of JSON-LD. More advanced features, including typed values, indexed values, and named graphs, can be found in .

The Context

When two people communicate with one another, the conversation takes place in a shared environment, typically called "the context of the conversation". This shared context allows the individuals to use shortcut terms, like the first name of a mutual friend, to communicate more quickly but without losing accuracy. A context in JSON-LD works in the same way. It allows two applications to use shortcut terms to communicate with one another more efficiently, but without losing accuracy.

Simply speaking, a context is used to map terms to IRIs. Terms are case sensitive and most valid strings that are not reserved JSON-LD keywords can be used as a term. Exceptions are the empty string `""` and strings that have the form of a keyword (i.e., starting with `"@"` followed exclusively by one or more ALPHA characters (see [[RFC5234]])), which must not be used as terms. Strings that have the form of an IRI (e.g., containing a `":"`) should not be used as terms.

For the sample document in the previous section, a context would look something like this:

    
    

As the context above shows, the value of a term definition can either be a simple string, mapping the term to an IRI, or a map.

A context is introduced using an entry with the key @context and may appear within a node object or a value object.

When an entry with a term key has a map value, the map is called an expanded term definition. The example above specifies that the values of image and homepage, if they are strings, are to be interpreted as IRIs. Expanded term definitions also allow terms to be used for index maps and to specify whether array values are to be interpreted as sets or lists. Expanded term definitions may be defined using IRIs or compact IRIs as keys, which is mainly used to associate type or language information with an IRIs or compact IRI.

Contexts can either be directly embedded into the document (an embedded context) or be referenced using a URL. Assuming the context document in the previous example can be retrieved at https://json-ld.org/contexts/person.jsonld, it can be referenced by adding a single line and allows a JSON-LD document to be expressed much more concisely as shown in the example below:

The referenced context not only specifies how the terms map to IRIs in the Schema.org vocabulary but also specifies that string values associated with the homepage and image property can be interpreted as an IRI ("@type": "@id", see for more details). This information allows developers to re-use each other's data without having to agree to how their data will interoperate on a site-by-site basis. External JSON-LD context documents may contain extra information located outside of the @context key, such as documentation about the terms declared in the document. Information contained outside of the @context value is ignored when the document is used as an external JSON-LD context document.

A remote context may also be referenced using a relative URL, which is resolved relative to the location of the document containing the reference. For example, if a document were located at http://example.org/document.jsonld and contained a relative reference to context.jsonld, the referenced context document would be found relative at http://example.org/context.jsonld.

    
    

Resolution of relative references to context URLs also applies to remote context documents, as they may themselves contain references to other contexts.

JSON documents can be interpreted as JSON-LD without having to be modified by referencing a context via an HTTP Link Header as described in . It is also possible to apply a custom context using the JSON-LD 1.1 API [[JSON-LD11-API]].

In JSON-LD documents, contexts may also be specified inline. This has the advantage that documents can be processed even in the absence of a connection to the Web. Ultimately, this is a modeling decision and different use cases may require different handling. See Security Considerations in for a discussion on using remote contexts.

This section only covers the most basic features of the JSON-LD Context. The Context can also be used to help interpret other more complex JSON data structures, such as indexed values, ordered values, and nested properties. More advanced features related to the JSON-LD Context are covered in .

IRIs

IRIs (Internationalized Resource Identifiers [[RFC3987]]) are fundamental to Linked Data as that is how most nodes and properties are identified. In JSON-LD, IRIs may be represented as an IRI reference. An IRI is defined in [[RFC3987]] as containing a scheme along with path and optional query and fragment segments. A relative IRI reference is an IRI that is relative to some other IRI. In JSON-LD, with exceptions that are as described below, all relative IRI references are resolved relative to the base IRI.

As noted in , IRIs can often be confused with URLs (Uniform Resource Locators), the primary distinction is that a URL locates a resource on the web, an IRI identifies a resource. While it is a good practice for resource identifiers to be dereferenceable, sometimes this is not practical. In particular, note the [[URN]] scheme for Uniform Resource Names, such as UUID. An example UUID is urn:uuid:f81d4fae-7dec-11d0-a765-00a0c91e6bf6.

Properties, values of @type, and values of properties with a term definition that defines them as being relative to the vocabulary mapping, may have the form of a relative IRI reference, but are resolved using the vocabulary mapping, and not the base IRI.

A string is interpreted as an IRI when it is the value of a map entry with the key @id:

  
  

Values that are interpreted as IRIs, can also be expressed as relative IRI references. For example, assuming that the following document is located at http://example.com/about/, the relative IRI reference ../ would expand to http://example.com/ (for more information on where relative IRI references can be used, please refer to section ).

  
  

IRIs can be expressed directly in the key position like so:

  
  

In the example above, the key http://schema.org/name is interpreted as an IRI.

Term-to-IRI expansion occurs if the key matches a term defined within the active context:

JSON keys that do not expand to an IRI, such as status in the example above, are not Linked Data and thus ignored when processed.

If type coercion rules are specified in the @context for a particular term or property IRI, an IRI is generated:

In the example above, since the value http://manu.sporny.org/ is expressed as a JSON string, the type coercion rules will transform the value into an IRI when processing the data. See for more details about this feature.

In summary, IRIs can be expressed in a variety of different ways in JSON-LD:

  1. Map entries that have a key mapping to a term in the active context expand to an IRI (only applies outside of the context definition).
  2. An IRI is generated for the string value specified using @id or @type.
  3. An IRI is generated for the string value of any key for which there are coercion rules that contain an @type key that is set to a value of @id or @vocab.

This section only covers the most basic features associated with IRIs in JSON-LD. More advanced features related to IRIs are covered in section .

Node Identifiers

To be able to externally reference nodes in an RDF graph, it is important that nodes have an identifier. IRIs are a fundamental concept of Linked Data, for nodes to be truly linked, dereferencing the identifier should result in a representation of that node. This may allow an application to retrieve further information about a node.

In JSON-LD, a node is identified using the @id keyword:

The example above contains a node object identified by the IRI http://me.markus-lanthaler.com/.

This section only covers the most basic features associated with node identifiers in JSON-LD. More advanced features related to node identifiers are covered in section .

Uses of JSON Objects

As a syntax, JSON has only a limited number of syntactic elements:

The JSON-LD data model allows for a richer set of resources, based on the RDF data model. The data model is described more fully in . JSON-LD uses JSON objects to describe various resources, along with the relationships between these resources:

Node objects
Node objects are used to define nodes in the linked data graph which may have both incoming and outgoing edges. Node objects are principle structure for defining resources having properties. See for the normative definition.
Value objects
Value objects are used for describing literal nodes in a linked data graph which may have only incoming edges. In JSON, some literal nodes may be described without the use of a JSON object (e.g., numbers, strings, and boolean values), but in the expanded form, all literal nodes are described using value objects. See for more information, and for the normative definition.
List Objects and Set objects
List Objects are a special kind of JSON-LD maps, distinct from node objects and value objects, used to express ordered values by wrapping an array in a map under the key `@list`. Set Objects exist for uniformity, and are equivalent to the array value of the `@set` key. See and for more detail.
Map Objects
JSON-LD uses various forms of maps as ways to more easily access values of a property.
Language Maps
Allows multiple values differing in their associated language to be indexed by language tag. See for more information, and for the normative definition.
Index Maps
Allows multiple values (node objects or value objects) to be indexed by an associated @index. See for more information, and for the normative definition.
Id Maps
Allows multiple node objects to be indexed by an associated @id. See for more information, and for the normative definition.
Type Maps
Allows multiple node objects to be indexed by an associated @type. See for more information, and for the normative definition.
Named Graph Indexing
Allows multiple named graphs to be indexed by an associated graph name. See for more information.
Graph objects
A graph object is much like a node object, except that it defines a named graph. See for more information, and for the normative definition. A node object may also describe a named graph, in addition to other properties defined on the node. The notable difference is that a graph object only describes a named graph.
Context Definitions
A Context Definition uses the JSON object form, but is not itself data in a linked data graph. A Context Definition also may contain expanded term definitions, which are also represented using JSON objects. See , for more information, and for the normative definition.

Specifying the Type

In Linked Data, it is common to specify the type of a graph node; in many cases, this can be inferred based on the properties used within a given node object, or the property for which a node is a value. For example, in the schema.org vocabulary, the givenName property is associated with a Person. Therefore, one may reason that if a node object contains the property givenName, that the type is a Person; making this explicit with @type helps to clarify the association.

The type of a particular node can be specified using the @type keyword. In Linked Data, types are uniquely identified with an IRI.

A node can be assigned more than one type by using an array:

The value of a @type key may also be a term defined in the active context:

In addition to setting the type of nodes, @type can also be used to set the type of a value to create a typed value. This use of @type is similar to that used to define the type of a node object, but value objects are restricted to having just a single type. The use of @type to create typed values is discussed more fully in .

Typed values can also be defined implicitly, by specifying @type in an expanded term definition. This is covered more fully in .

Advanced Concepts

JSON-LD has a number of features that provide functionality above and beyond the core functionality described above. JSON can be used to express data using such structures, and the features described in this section can be used to interpret a variety of different JSON structures as Linked Data. A JSON-LD processor will make use of provided and embedded contexts to interpret property values in a number of different idiomatic ways.

Describing values

One pattern in JSON is for the value of a property to be a string. Often times, this string actually represents some other typed value, for example an IRI, a date, or a string in some specific language. See for details on how to describe such value typing.

Value ordering

In JSON, a property with an array value implies an implicit order; arrays in JSON-LD do not convey any ordering of the contained elements by default, unless defined using embedded structures or through a context definition. See for a further discussion.

Property nesting

Another JSON idiom often found in APIs is to use an intermediate object to group together related properties of an object; in JSON-LD these are referred to as nested properties and are described in .

Referencing objects

Linked Data is all about describing the relationships between different resources. Sometimes these relationships are between resources defined in different documents described on the web, sometimes the resources are described within the same document.

In this case, a document residing at http://manu.sporny.org/about may contain the example above, and reference another document at https://greggkellogg.net/foaf which could include a similar representation.

A common idiom found in JSON usage is objects being specified as the value of other objects, called object embedding in JSON-LD; for example, a friend specified as an object value of a Person:

See details these relationships.

Indexed values

Another common idiom in JSON is to use an intermediate object to represent property values via indexing. JSON-LD allows data to be indexed in a number of different ways, as detailed in .

Reverse Properties

JSON-LD serializes directed graphs. That means that every property points from a node to another node or value. However, in some cases, it is desirable to serialize in the reverse direction, as detailed in .

The following sections describe such advanced functionality in more detail.

Advanced Context Usage

Section introduced the basics of what makes JSON-LD work. This section expands on the basic principles of the context and demonstrates how more advanced use cases can be achieved using JSON-LD.

In general, contexts may be used any time a map is defined. The only time that one cannot express a context is as a direct child of another context definition (other than as part of an expanded term definition). For example, a JSON-LD document may have the form of an array composed of one or more node objects, which use a context definition in each top-level node object:

The outer array is standard for a document in expanded document form and flattened document form, and may be necessary when describing a disconnected graph, where nodes may not reference each other. In such cases, using a top-level map with a @graph property can be useful for saving the repetition of @context. See for more.

Duplicate context terms are overridden using a most-recently-defined-wins mechanism.

In the example above, the name term is overridden in the more deeply nested details structure, which uses its own embedded context. Note that this is rarely a good authoring practice and is typically used when working with legacy applications that depend on a specific structure of the map. If a term is redefined within a context, all previous rules associated with the previous definition are removed. If a term is redefined to null, the term is effectively removed from the list of terms defined in the active context.

Multiple contexts may be combined using an array, which is processed in order. The set of contexts defined within a specific map are referred to as local contexts. The active context refers to the accumulation of local contexts that are in scope at a specific point within the document. Setting a local context to null effectively resets the active context to an empty context, without term definitions, default language, or other things defined within previous contexts. The following example specifies an external context and then layers an embedded context on top of the external context:

In JSON-LD 1.1, there are other mechanisms for introducing contexts, including scoped contexts and imported contexts, and there are new ways of protecting term definitions, so there are cases where the last defined inline context is not necessarily one which defines the scope of terms. See , , , and for further information.

When possible, the context definition should be put at the top of a JSON-LD document. This makes the document easier to read and might make streaming parsers more efficient. Documents that do not have the context at the top are still conformant JSON-LD.

To avoid forward-compatibility issues, terms starting with an @ character followed exclusively by one or more ALPHA characters (see [[RFC5234]]) are to be avoided as they might be used as keyword in future versions of JSON-LD. Terms starting with an @ character that are not JSON-LD 1.1 keywords are treated as any other term, i.e., they are ignored unless mapped to an IRI. Furthermore, the use of empty terms ("") is not allowed as not all programming languages are able to handle empty JSON keys.

JSON-LD 1.1 Processing Mode

New features defined in JSON-LD 1.1 are available unless the processing mode is set to `json-ld-1.0`. This may be set through an API option. The processing mode may be explicitly set to `json-ld-1.1` using the `@version` entry in a context set to the value `1.1` as a number, or through an API option. Explicitly setting the processing mode to `json-ld-1.1` will prohibit JSON-LD 1.0 processors from incorrectly processing a JSON-LD 1.1 document.

  
  

The first context encountered when processing a document which contains @version determines the processing mode, unless it is defined explicitly through an API option. This means that if "@version": 1.1 is encountered after processing a context without @version, the former will be interpreted as having had "@version": 1.1 defined within it.

Setting the processing mode explicitly to `json-ld-1.1` is RECOMMENDED to prevent a [[[JSON-LD10]]] processor from incorrectly processing a JSON-LD 1.1 document and producing different results.

Default Vocabulary

At times, all properties and types may come from the same vocabulary. JSON-LD's @vocab keyword allows an author to set a common prefix which is used as the vocabulary mapping and is used for all properties and types that do not match a term and are neither an IRI nor a compact IRI (i.e., they do not contain a colon).

If @vocab is used but certain keys in a map should not be expanded using the vocabulary IRI, a term can be explicitly set to null in the context. For instance, in the example below the databaseId entry would not expand to an IRI causing the property to be dropped when expanding.

Since JSON-LD 1.1, the vocabulary mapping in a local context can be set to a relative IRI reference, which is concatenated to any vocabulary mapping in the active context (see for how this applies if there is no vocabulary mapping in the active context).

The following example illustrates the affect of expanding a property using a relative IRI reference, which is shown in the Expanded (Result) tab below.

The grammar for `@vocab`, as defined in allows the value to be a term or compact IRI. Note that terms used in the value of `@vocab` must be in scope at the time the context is introduced, otherwise there would be a circular dependency between `@vocab` and other terms defined in the same context.

Base IRI

JSON-LD allows IRIs to be specified in a relative form which is resolved against the document base according section 5.1 Establishing a Base URI of [[RFC3986]]. The base IRI may be explicitly set with a context using the @base keyword.

For example, if a JSON-LD document was retrieved from http://example.com/document.jsonld, relative IRI references would resolve against that IRI:

    
  

This document uses an empty @id, which resolves to the document base. However, if the document is moved to a different location, the IRI would change. To prevent this without having to use an IRI, a context may define an @base mapping, to overwrite the base IRI for the document.

Setting @base to null will prevent relative IRI references from being expanded to IRIs.

Please note that the @base will be ignored if used in external contexts.

Using the Document Base for the Default Vocabulary

In some cases, vocabulary terms are defined directly within the document itself, rather than in an external vocabulary. Since JSON-LD 1.1, the vocabulary mapping in a local context can be set to a relative IRI reference, which is, if there is no vocabulary mapping in scope, resolved against the base IRI. This causes terms which are expanded relative to the vocabulary, such as the keys of node objects, to be based on the base IRI to create IRIs.


If this document were located at http://example/document, it would expand as follows:

Compact IRIs

A compact IRI is a way of expressing an IRI using a prefix and suffix separated by a colon (:). The prefix is a term taken from the active context and is a short string identifying a particular IRI in a JSON-LD document. For example, the prefix foaf may be used as a shorthand for the Friend-of-a-Friend vocabulary, which is identified using the IRI http://xmlns.com/foaf/0.1/. A developer may append any of the FOAF vocabulary terms to the end of the prefix to specify a short-hand version of the IRI for the vocabulary term. For example, foaf:name would be expanded to the IRI http://xmlns.com/foaf/0.1/name.

In the example above, foaf:name expands to the IRI http://xmlns.com/foaf/0.1/name and foaf:Person expands to http://xmlns.com/foaf/0.1/Person.

Prefixes are expanded when the form of the value is a compact IRI represented as a prefix:suffix combination, the prefix matches a term defined within the active context, and the suffix does not begin with two slashes (//). The compact IRI is expanded by concatenating the IRI mapped to the prefix to the (possibly empty) suffix. If the prefix is not defined in the active context, or the suffix begins with two slashes (such as in http://example.com), the value is interpreted as IRI instead. If the prefix is an underscore (_), the value is interpreted as blank node identifier instead.

It's also possible to use compact IRIs within the context as shown in the following example:

When operating explicitly with the processing mode for [[[JSON-LD10]]] compatibility, terms may be chosen as compact IRI prefixes when compacting only if a simple term definition is used where the value ends with a URI gen-delim character (e.g, /, # and others, see [[RFC3986]]).

In JSON-LD 1.1, terms may be chosen as compact IRI prefixes when expanding or compacting only if a simple term definition is used where the value ends with a URI gen-delim character, or if their expanded term definition contains a @prefix entry with the value true. If a simple term definition does not end with a URI gen-delim character, or an expanded term definition contains a @prefix entry with the value false, the term will not be used for either expanding compact IRIs or compacting IRIs to compact IRIs.

The term selection behavior for 1.0 processors was changed as a result of an errata against [[[JSON-LD10]]] reported here. This does not affect the behavior of processing existing JSON-LD documents, but creates a slight change when compacting documents using Compact IRIs.

The behavior when compacting can be illustrated by considering the following input document in expanded form:

  
  

Using the following context in the 1.0 processing mode will now select the term vocab rather than property, even though the IRI associated with property captures more of the original IRI.

    
  

Compacting using the previous context with the above expanded input document results in the following compacted result:

In the original [[JSON-LD10]], the term selection algorithm would have selected property, creating the Compact IRI property:One. The original behavior can be made explicit using @prefix:

    
  

In this case, the property term would not normally be usable as a prefix, both because it is defined with an expanded term definition, and because its @id does not end in a gen-delim character. Adding "@prefix": true allows it to be used as the prefix portion of the compact IRI property:One.

Aliasing Keywords

Each of the JSON-LD keywords, except for @context, may be aliased to application-specific keywords. This feature allows legacy JSON content to be utilized by JSON-LD by re-using JSON keys that already exist in legacy documents. This feature also allows developers to design domain-specific implementations using only the JSON-LD context.

In the example above, the @id and @type keywords have been given the aliases url and a, respectively.

Other than for @type, properties of expanded term definitions where the term is a keyword result in an error. Unless the processing mode is set to `json-ld-1.0`, there is also an exception for @type; see for further details and usage examples.

Unless the processing mode is set to `json-ld-1.0`, aliases of keywords are either simple term definitions, where the value is a keyword, or an expanded term definition with an `@id` entry and optionally an `@protected` entry; no other entries are allowed. There is also an exception for aliases of @type, as indicated above. See for further details of using `@protected`.

Since keywords cannot be redefined, they can also not be aliased to other keywords.

Aliased keywords may not be used within a context, itself.

See for a normative definition of all keywords.

IRI Expansion within a Context

In general, normal IRI expansion rules apply anywhere an IRI is expected (see ). Within a context definition, this can mean that terms defined within the context may also be used within that context as long as there are no circular dependencies. For example, it is common to use the xsd namespace when defining typed values:


In this example, the xsd term is defined and used as a prefix for the @type coercion of the age property.

Terms may also be used when defining the IRI of another term:


Compact IRIs and IRIs may be used on the left-hand side of a term definition.


In this example, the compact IRI form is used in two different ways. In the first approach, foaf:age declares both the IRI for the term (using short-form) as well as the @type associated with the term. In the second approach, only the @type associated with the term is specified. The full IRI for foaf:homepage is determined by looking up the foaf prefix in the context.

If a compact IRI is used as a term, it must expand to the value that compact IRI would have on its own when expanded. This represents a change to the original 1.0 algorithm to prevent terms from expanding to a different IRI, which could lead to undesired results.


IRIs may also be used in the key position in a context:


In order for the IRI to match above, the IRI needs to be used in the JSON-LD document. Also note that foaf:homepage will not use the { "@type": "@id" } declaration because foaf:homepage is not the same as http://xmlns.com/foaf/0.1/homepage. That is, terms are looked up in a context using direct string comparison before the prefix lookup mechanism is applied.

Neither an IRI reference nor a compact IRI may expand to some other unrelated IRI. This represents a change to the original 1.0 algorithm which allowed this behavior but discouraged it.

The only other exception for using terms in the context is that circular definitions are not allowed. That is, a definition of term1 cannot depend on the definition of term2 if term2 also depends on term1. For example, the following context definition is illegal:


Scoped Contexts

An expanded term definition can include a @context property, which defines a context (a scoped context) for values of properties defined using that term. When used for a property, this is called a property-scoped context. This allows values to use term definitions, the base IRI, vocabulary mappings or the default language which are different from the node object they are contained in, as if the context was specified within the value itself.

In this case, the social profile is defined using the schema.org vocabulary, but interest is imported from FOAF, and is used to define a node describing one of Manu's interests where those properties now come from the FOAF vocabulary.

Expanding this document, uses a combination of terms defined in the outer context, and those defined specifically for that term in a property-scoped context.

Scoping can also be performed using a term used as a value of @type:

Scoping on @type is useful when common properties are used to relate things of different types, where the vocabularies in use within different entities call for different context scoping. For example, hasPart/partOf may be common terms used in a document, but mean different things depending on the context. A type-scoped context is only in effect for the node object on which the type is used; the previous in-scope contexts are placed back into effect when traversing into another node object. As described further in , this may be controlled using the `@propagate` keyword.

Any property-scoped or local contexts that were introduced in the node object would still be in effect when traversing into another node object.

When expanding, each value of @type is considered (in code point order) where that value is also a term in the active context having its own type-scoped context. If so, that the scoped context is applied to the active context.

The values of @type are unordered, so if multiple types are listed, the order that type-scoped contexts are applied is based on code point ordering.

For example, consider the following semantically equivalent examples. The first example, shows how properties and types can define their own scoped contexts, which are included when expanding.

  
  

Contexts are processed depending on how they are defined. A property-scoped context is processed first, followed by any embedded context, followed lastly by the type-scoped contexts, in the appropriate order. The previous example is logically equivalent to the following:

  
  

If a term defines a scoped context, and then that term is later redefined, the association of the context defined in the earlier expanded term definition is lost within the scope of that redefinition. This is consistent with term definitions of a term overriding previous term definitions from earlier less deeply nested definitions, as discussed in .

Scoped Contexts are a new feature in JSON-LD 1.1.

Context Propagation

Once introduced, contexts remain in effect until a subsequent context removes it by setting `@context` to `null`, or by redefining terms, with the exception of type-scoped contexts, which limit the effect of that context until the next node object is entered. This behavior can be changed using the `@propagate` keyword.

The following example illustrates how terms defined in a context with `@propagate` set to `false` are effectively removed when descending into new node object.

Contexts included within an array must all have the same value for `@propagate` due to the way that rollback is defined in [[[JSON-LD11-API]]].

Imported Contexts

JSON-LD 1.0 included mechanisms for modifying the context that is in effect. This included the capability to load and process a remote context and then apply further changes to it via new contexts.

However, with the introduction of JSON-LD 1.1, it is also desirable to be able to load a remote context, in particular an existing JSON-LD 1.0 context, and apply JSON-LD 1.1 features to it prior to processing.

By using the `@import` keyword in a context, another remote context, referred to as an imported context, can be loaded and modified prior to processing. The modifications are expressed in the context that includes the `@import` keyword, referred to as the wrapping context. Once an imported context is loaded, the contents of the wrapping context are merged into it prior to processing. The merge operation will cause each key-value pair in the wrapping context to be added to the loaded imported context, with the wrapping context key-value pairs taking precedence.

By enabling existing contexts to be reused and edited inline prior to processing, context-wide keywords can be applied to adjust all term definitions in the imported context. Similarly, term definitions can be replaced prior to processing, enabling adjustments that, for instance, ensure term definitions match previously protected terms or that they include additional type coercion information.

The following examples illustrate how `@import` can be used to express a type-scoped context that loads an imported context and sets `@propagate` to `true`, as a technique for making other similar modifications.

Suppose there was a context that could be referenced remotely via the URL https://json-ld.org/contexts/remote-context.jsonld:

  
  

A wrapping context could be used to source it and modify it:

  
  

The effect would be the same as if the entire imported context had been copied into the type-scoped context:

  
  

Similarly, the wrapping context may replace term definitions or set other context-wide keywords that may affect how the imported context term definitions will be processed:

  
  

Again, the effect would be the same as if the entire imported context had been copied into the context:

  
  

The result of loading imported contexts must be context definition, not an IRI or an array. Additionally, the imported context cannot include an `@import` entry.

Protected Term Definitions

JSON-LD is used in many specifications as the specified data format. However, there is also a desire to allow some JSON-LD contents to be processed as plain JSON, without using any of the JSON-LD algorithms. Because JSON-LD is very flexible, some terms from the original format may be locally overridden through the use of embedded contexts, and take a different meaning for JSON-LD based implementations. On the other hand, "plain JSON" implementations may not be able to interpret these embedded contexts, and hence will still interpret those terms with their original meaning. To prevent this divergence of interpretation, JSON-LD 1.1 allows term definitions to be protected.

A protected term definition is a term definition with an entry @protected set to true. It generally prevents further contexts from overriding this term definition, either through a new definition of the same term, or through clearing the context with "@context": null. Such attempts will raise an error and abort the processing (except in some specific situations described below).

    
  

When all or most term definitions of a context need to be protected, it is possible to add an entry @protected set to true to the context itself. It has the same effect as protecting each of its term definitions individually. Exceptions can be made by adding an entry @protected set to false in some term definitions.

While protected terms can in general not be overridden, there are two exceptions to this rule. The first exception is that a context is allowed to redefine a protected term if the new definition is identical to the protected term definition (modulo the @protected flag). The rationale is that the new definition does not violate the protection, as it does not change the semantics of the protected term. This is useful for widespread term definitions, such as aliasing @type to type, which may occur (including in a protected form) in several contexts.

The second exception is that a property-scoped context is not affected by protection, and can therefore override protected terms, either with a new term definition, or by clearing the context with "@context": null.

The rationale is that "plain JSON" implementations, relying on a given specification, will only traverse properties defined by that specification. Scoped contexts belonging to the specified properties are part of the specification, so the "plain JSON" implementations are expected to be aware of the change of semantics they induce. Scoped contexts belonging to other properties apply to parts of the document that "plain JSON" implementations will ignore. In both cases, there is therefore no risk of diverging interpretations between JSON-LD-aware implementations and "plain JSON" implementations, so overriding is permitted.

By preventing terms from being overridden, protection also prevents any adaptation of a term (e.g., defining a more precise datatype, restricting the term's use to lists, etc.). This kind of adaptation is frequent with some general purpose contexts, for which protection would therefore hinder their usability. As a consequence, context publishers should use this feature with care.

Protected term definitions are a new feature in JSON-LD 1.1.

Describing Values

Values are leaf nodes in a graph associated with scalar values such as strings, dates, times, and other such atomic values.

Typed Values

A value with an associated type, also known as a typed value, is indicated by associating a value with an IRI which indicates the value's type. Typed values may be expressed in JSON-LD in three ways:

  1. By utilizing the @type keyword when defining a term within an @context section.
  2. By utilizing a value object.
  3. By using a native JSON type such as number, true, or false.

The first example uses the @type keyword to associate a type with a particular term in the @context:

The modified key's value above is automatically interpreted as a dateTime value because of the information specified in the @context. The example tabs show how a JSON-LD processor will interpret the data.

The second example uses the expanded form of setting the type information in the body of a JSON-LD document:

Both examples above would generate the value 2010-05-29T14:17:39+02:00 with the type http://www.w3.org/2001/XMLSchema#dateTime. Note that it is also possible to use a term or a compact IRI to express the value of a type.

The @type keyword is also used to associate a type with a node. The concept of a node type and a value type are distinct. For more on adding types to nodes, see .

When expanding, an `@type` defined within a term definition can be associated with a string value to create an expanded value object, which is described in . Type coercion only takes place on string values, not for values which are maps, such as node objects and value objects in their expanded form.

A node type specifies the type of thing that is being described, like a person, place, event, or web page. A value type specifies the data type of a particular value, such as an integer, a floating point number, or a date.


The first use of @type associates a node type (http://schema.org/BlogPosting) with the node, which is expressed using the @id keyword. The second use of @type associates a value type (http://www.w3.org/2001/XMLSchema#dateTime) with the value expressed using the @value keyword. As a general rule, when @value and @type are used in the same map, the @type keyword is expressing a value type. Otherwise, the @type keyword is expressing a node type. The example above expresses the following data:

JSON Literals

At times, it is useful to include JSON within JSON-LD that is not interpreted as JSON-LD. Generally, a JSON-LD processor will ignore properties which don't map to IRIs, but this causes them to be excluded when performing various algorithmic transformations. But, when the data that is being described is, itself, JSON, it's important that it survives algorithmic transformations.

JSON-LD is intended to allow native JSON to be interpreted through the use of a context. The use of JSON literals creates blobs of data which are not available for interpretation. It is for use only in the rare cases that JSON cannot be represented as JSON-LD.

When a term is defined with @type set to @json, a JSON-LD processor will treat the value as a JSON literal, rather than interpreting it further as JSON-LD. In the expanded document form, such JSON will become the value of @value within a value object having "@type": "@json".

When transformed into RDF, the JSON literal will have a lexical form based on a specific serialization of the JSON, as described in Compaction algorithm of [[JSON-LD11-API]] and the JSON datatype.

The following example shows an example of a JSON Literal contained as the value of a property. Note that the RDF results use a canonicalized form of the JSON to ensure interoperability between different processors. JSON canonicalization is described in Data Round Tripping in [[JSON-LD11-API]].

Generally, when a JSON-LD processor encounters `null`, the associated entry or value is removed. However, `null` is a valid JSON token; when used as the value of a JSON literal, a `null` value will be preserved.

Type Coercion

JSON-LD supports the coercion of string values to particular data types. Type coercion allows someone deploying JSON-LD to use string property values and have those values be interpreted as typed values by associating an IRI with the value in the expanded value object representation. Using type coercion, string value representation can be used without requiring the data type to be specified explicitly with each piece of data.

Type coercion is specified within an expanded term definition using the @type key. The value of this key expands to an IRI. Alternatively, the keyword @id or @vocab may be used as value to indicate that within the body of a JSON-LD document, a string value of a term coerced to @id or @vocab is to be interpreted as an IRI. The difference between @id and @vocab is how values are expanded to IRIs. @vocab first tries to expand the value by interpreting it as term. If no matching term is found in the active context, it tries to expand it as an IRI or a compact IRI if there's a colon in the value; otherwise, it will expand the value using the active context's vocabulary mapping, if present. Values coerced to @id in contrast are expanded as an IRI or a compact IRI if a colon is present; otherwise, they are interpreted as relative IRI references.

The ability to coerce a value using a term definition is distinct from setting one or more types on a node object, as the former does not result in new data being added to the graph, while the latter manages node types through adding additional relationships to the graph.

Terms or compact IRIs used as the value of a @type key may be defined within the same context. This means that one may specify a term like xsd and then use xsd:integer within the same context definition.

The example below demonstrates how a JSON-LD author can coerce values to typed values and IRIs.

It is important to note that terms are only used in expansion for vocabulary-relative positions, such as for keys and values of map entries. Values of @id are considered to be document-relative, and do not use term definitions for expansion. For example, consider the following:

The unexpected result is that "barney" expands to both http://example1.com/barney and http://example2.com/barney, depending where it is encountered. String values interpreted as IRIs because of the associated term definitions are typically considered to be document-relative. In some cases, it makes sense to interpret these relative to the vocabulary, prescribed using "@type": "@vocab" in the term definition, though this can lead to unexpected consequences such as these.

In the previous example, "barney" appears twice, once as the value of @id, which is always interpreted as a document-relative IRI, and once as the value of "fred", which is defined to be vocabulary-relative, thus the different expanded values.

For more on this see .

A variation on the previous example using "@type": "@id" instead of @vocab illustrates the behavior of interpreting "barney" relative to the document:

The triple ex1:fred ex2:knows ex1:barney . is emitted twice, but exists only once in an output dataset, as it is a duplicate triple.

Terms may also be defined using IRIs or compact IRIs. This allows coercion rules to be applied to keys which are not represented as a simple term. For example:

In this case the @id definition in the term definition is optional. If it does exist, the IRI or compact IRI representing the term will always be expanded to IRI defined by the @id key—regardless of whether a prefix is defined or not.

Type coercion is always performed using the unexpanded value of the key. In the example above, that means that type coercion is done looking for foaf:age in the active context and not for the corresponding, expanded IRI http://xmlns.com/foaf/0.1/age.

Keys in the context are treated as terms for the purpose of expansion and value coercion. At times, this may result in multiple representations for the same expanded IRI. For example, one could specify that dog and cat both expanded to http://example.com/vocab#animal. Doing this could be useful for establishing different type coercion or language specification rules.

String Internationalization

At times, it is important to annotate a string with its language. In JSON-LD this is possible in a variety of ways. First, it is possible to define a default language for a JSON-LD document by setting the @language key in the context:

The example above would associate the ja language tag with the two strings 花澄 and 科学者. Languages tags are defined in [[BCP47]]. The default language applies to all string values that are not type coerced.

To clear the default language for a subtree, @language can be set to null in an intervening context, such as a scoped context as follows:

  
  

Second, it is possible to associate a language with a specific term using an expanded term definition:

  
  

The example above would associate 忍者 with the specified default language tag ja, Ninja with the language tag en, and Nindža with the language tag cs. The value of name, Yagyū Muneyoshi wouldn't be associated with any language tag since @language was reset to null in the expanded term definition.

Language associations are only applied to plain strings. Typed values or values that are subject to type coercion are not language tagged.

Just as in the example above, systems often need to express the value of a property in multiple languages. Typically, such systems also try to ensure that developers have a programmatically easy way to navigate the data structures for the language-specific data. In this case, language maps may be utilized.

  
  

The example above expresses exactly the same information as the previous example but consolidates all values in a single property. To access the value in a specific language in a programming language supporting dot-notation accessors for object properties, a developer may use the property.language pattern (when languages are limited to the primary language sub-tag, and do not depend on other sub-tags, such as `"en-us"`). For example, to access the occupation in English, a developer would use the following code snippet: obj.occupation.en.

Third, it is possible to override the default language by using a value object:

  
  

This makes it possible to specify a plain string by omitting the @language tag or setting it to null when expressing it using a value object:

  
  

See for a description of using language maps to set the language of mapped values.

Base Direction

It is also possible to annotate a string, or language-tagged string, with its base direction. As with language, it is possible to define a default base direction for a JSON-LD document by setting the `@direction` key in the context:

The example above would associate the ar-EG language tag and "rtl" base direction with the two strings HTML و CSS: تصميم و إنشاء مواقع الويب and مكتبة. The default base direction applies to all string values that are not type coerced.

To clear the default base direction for a subtree, @direction can be set to null in an intervening context, such as a scoped context as follows:

  
  

Second, it is possible to associate a base direction with a specific term using an expanded term definition:

  
  

The example above would create three properties:

SubjectPropertyValueLanguageDirection
_:b0http://example.com/vocab/publisherمكتبةar-EG
_:b0http://example.com/vocab/titleHTML و CSS: تصميم و إنشاء مواقع الويبar-EGrtl
_:b0http://example.com/vocab/titleHTML and CSS: Design and Build Websitesenltr

Base direction associations are only applied to plain strings and language-tagged strings. Typed values or values that are subject to type coercion are not given a base direction.

Third, it is possible to override the default base direction by using a value object:

  
  

See [[[string-meta]]] [[string-meta]] for a deeper discussion of base direction.

Value Ordering

A JSON-LD author can express multiple values in a compact way by using arrays. Since graphs do not describe ordering for links between nodes, arrays in JSON-LD do not convey any ordering of the contained elements by default. This is exactly the opposite from regular JSON arrays, which are ordered by default. For example, consider the following simple document:

Multiple values may also be expressed using the expanded form:

The example shown above would generate multiple statements, again with no inherent order.

Although multiple values of a property are typically of the same type, JSON-LD places no restriction on this, and a property may have values of different types:

When viewed as statements, the values have no inherent order.

Lists

As the notion of ordered collections is rather important in data modeling, it is useful to have specific language support. In JSON-LD, a list may be represented using the @list keyword as follows:

This describes the use of this array as being ordered, and order is maintained when processing a document. If every use of a given multi-valued property is a list, this may be abbreviated by setting @container to @list in the context:

The implementation of lists in RDF depends on linking anonymous nodes together using the properties rdf:first and rdf:rest, with the end of the list defined as the resource rdf:nil, as the "statements" tab illustrates. This allows order to be represented within an unordered set of statements.

Both JSON-LD and Turtle provide shortcuts for representing ordered lists.

In JSON-LD 1.1, lists of lists, where the value of a list object, may itself be a list object, are fully supported.

Note that the "@container": "@list" definition recursively describes array values of lists as being, themselves, lists. For example, in [[[RFC7946]]] (see [[RFC7946]]), coordinates are an ordered list of positions, which are represented as an array of two or more numbers:

{
  "type": "Feature",
  "bbox": [-10.0, -10.0, 10.0, 10.0],
  "geometry": {
    "type": "Polygon",
    "coordinates": [
        [
            [-10.0, -10.0],
            [10.0, -10.0],
            [10.0, 10.0],
            [-10.0, -10.0]
        ]
    ]
  }
  ####//...####
}

For these examples, it's important that values expressed within bbox and coordinates maintain their order, which requires the use of embedded list structures. In JSON-LD 1.1, we can express this using recursive lists, by simply adding the appropriate context definition:

Note that coordinates includes three levels of lists.

Values of terms associated with an @list container are always represented in the form of an array, even if there is just a single value or no value at all.

Sets

While @list is used to describe ordered lists, the @set keyword is used to describe unordered sets. The use of @set in the body of a JSON-LD document is optimized away when processing the document, as it is just syntactic sugar. However, @set is helpful when used within the context of a document. Values of terms associated with an @set container are always represented in the form of an array, even if there is just a single value that would otherwise be optimized to a non-array form in compact form (see ). This makes post-processing of JSON-LD documents easier as the data is always in array form, even if the array only contains a single value.

This describes the use of this array as being unordered, and order may change when processing a document. By default, arrays of values are unordered, but this may be made explicit by setting @container to @set in the context:

Since JSON-LD 1.1, the @set keyword may be combined with other container specifications within an expanded term definition to similarly cause compacted values of indexes to be consistently represented using arrays. See for a further discussion.

Using @set with @type

Unless the processing mode is set to `json-ld-1.0`, @type may be used with an expanded term definition with @container set to @set; no other entries may be set within such an expanded term definition. This is used by the Compaction algorithm to ensure that the values of @type (or an alias) are always represented in an array.

  
  

Nested Properties

Many JSON APIs separate properties from their entities using an intermediate object; in JSON-LD these are called nested properties. For example, a set of possible labels may be grouped under a common property:

By defining labels using the keyword @nest, a JSON-LD processor will ignore the nesting created by using the labels property and process the contents as if it were declared directly within containing object. In this case, the labels property is semantically meaningless. Defining it as equivalent to @nest causes it to be ignored when expanding, making it equivalent to the following:

Similarly, term definitions may contain a @nest property referencing a term aliased to @nest which will cause such properties to be nested under that aliased term when compacting. In the example below, both main_label and other_label are defined with "@nest": "labels", which will cause them to be serialized under labels when compacting.

  
  
  
  

Nested properties are a new feature in JSON-LD 1.1.

Embedding

Embedding is a JSON-LD feature that allows an author to use node objects as property values. This is a commonly used mechanism for creating a parent-child relationship between two nodes.

Without embedding, node objects can be linked by referencing the identifier of another node object. For example:

The previous example describes two node objects, for Manu and Gregg, with the knows property defined to treat string values as identifiers. Embedding allows the node object for Gregg to be embedded as a value of the knows property:

A node object, like the one used above, may be used in any value position in the body of a JSON-LD document.

While it is considered a best practice to identify nodes in a graph, at times this is impractical. In the data model, nodes without an explicit identifier are called blank nodes, which can be represented in a serialization such as JSON-LD using a blank node identifier. In the previous example, the top-level node for Manu does not have an identifier, and does not need one to describe it within the data model. However, if we were to want to describe a knows relationship from Gregg to Manu, we would need to introduce a blank node identifier (here _:b0).

Blank node identifiers may be automatically introduced by algorithms such as flattening, but they are also useful for authors to describe such relationships directly.

Identifying Blank Nodes

At times, it becomes necessary to be able to express information without being able to uniquely identify the node with an IRI. This type of node is called a blank node. JSON-LD does not require all nodes to be identified using @id. However, some graph topologies may require identifiers to be serializable. Graphs containing loops, e.g., cannot be serialized using embedding alone, @id must be used to connect the nodes. In these situations, one can use blank node identifiers, which look like IRIs using an underscore (_) as scheme. This allows one to reference the node locally within the document, but makes it impossible to reference the node from an external document. The blank node identifier is scoped to the document in which it is used.

The example above contains information about two secret agents that cannot be identified with an IRI. While expressing that agent 1 knows agent 2 is possible without using blank node identifiers, it is necessary to assign agent 1 an identifier so that it can be referenced from agent 2.

It is worth noting that blank node identifiers may be relabeled during processing. If a developer finds that they refer to the blank node more than once, they should consider naming the node using a dereferenceable IRI so that it can also be referenced from other documents.

Indexed Values

Sometimes multiple property values need to be accessed in a more direct fashion than iterating through multiple array values. JSON-LD provides an indexing mechanism to allow the use of an intermediate map to associate specific indexes with associated values.

Data Indexing
As described in , data indexing allows an arbitrary key to reference a node or value.
Language Indexing
As described in , language indexing allows a language to reference a string and be interpreted as the language associated with that string.
Node Identifier Indexing
As described in , node identifier indexing allows an IRI to reference a node and be interpreted as the identifier of that node.
Node Type Indexing
As described in , node type indexing allows an IRI to reference a node and be interpreted as a type of that node.

See for other uses of indexing in JSON-LD.

Data Indexing

Databases are typically used to make access to data more efficient. Developers often extend this sort of functionality into their application data to deliver similar performance gains. This data may have no meaning from a Linked Data standpoint, but is still useful for an application.

JSON-LD introduces the notion of index maps that can be used to structure data into a form that is more efficient to access. The data indexing feature allows an author to structure data using a simple key-value map where the keys do not map to IRIs. This enables direct access to data instead of having to scan an array in search of a specific item. In JSON-LD such data can be specified by associating the @index keyword with a @container declaration in the context:

In the example above, the athletes term has been marked as an index map. The catcher and pitcher keys will be ignored semantically, but preserved syntactically, by the JSON-LD Processor. If used in JavaScript, this can allow a developer to access a particular athlete using the following code snippet: obj.athletes.pitcher.

The interpretation of the data is expressed in the statements table. Note how the index keys do not appear in the statements, but would continue to exist if the document were compacted or expanded (see and ) using a JSON-LD processor.

As data indexes are not preserved when round-tripping to RDF; this feature should be used judiciously. Often, other indexing mechanisms, which are preserved, are more appropriate.

The value of @container can also be an array containing both @index and @set. When compacting, this ensures that a JSON-LD Processor will use the array form for all values of indexes.

Unless the processing mode is set to `json-ld-1.0`, the special index @none is used for indexing data which does not have an associated index, which is useful to maintain a normalized representation.

Property-based data indexing

In its simplest form (as in the examples above), data indexing assigns no semantics to the keys of an index map. However, in some situations, the keys used to index objects are semantically linked to these objects, and should be preserved not only syntactically, but also semantically.

Unless the processing mode is set to `json-ld-1.0`, "@container": "@index" in a term description can be accompanied with an "@index" key. The value of that key must map to an IRI, which identifies the semantic property linking each object to its key.

When using property-based data indexing, index maps can only be used on node objects, not value objects or graph objects. Value objects are restricted to have only certain keys and do not support arbitrary properties.

Language Indexing

JSON which includes string values in multiple languages may be represented using a language map to allow for easily indexing property values by language tag. This enables direct access to language values instead of having to scan an array in search of a specific item. In JSON-LD such data can be specified by associating the @language keyword with a @container declaration in the context:

In the example above, the label term has been marked as a language map. The en and de keys are implicitly associated with their respective values by the JSON-LD Processor. This allows a developer to access the German version of the label using the following code snippet: obj.label.de, which, again, is only appropriate when languages are limited to the primary language sub-tag and do not depend on other sub-tags, such as `"de-at"`.

The value of @container can also be an array containing both @language and @set. When compacting, this ensures that a JSON-LD Processor will use the array form for all values of language tags.

Unless the processing mode is set to `json-ld-1.0`, the special index @none is used for indexing strings which do not have a language; this is useful to maintain a normalized representation for string values not having a datatype.

Node Identifier Indexing

In addition to index maps, JSON-LD introduces the notion of id maps for structuring data. The id indexing feature allows an author to structure data using a simple key-value map where the keys map to IRIs. This enables direct access to associated node objects instead of having to scan an array in search of a specific item. In JSON-LD such data can be specified by associating the @id keyword with a @container declaration in the context:

In the example above, the post term has been marked as an id map. The http://example.com/posts/1/en and http://example.com/posts/1/de keys will be interpreted as the @id property of the node object value.

The interpretation of the data above is exactly the same as that in using a JSON-LD processor.

The value of @container can also be an array containing both @id and @set. When compacting, this ensures that a JSON-LD processor will use the array form for all values of node identifiers.

The special index @none is used for indexing node objects which do not have an @id, which is useful to maintain a normalized representation. The @none index may also be a term which expands to @none, such as the term none used in the example below.

Id maps are a new feature in JSON-LD 1.1.

Node Type Indexing

In addition to id and index maps, JSON-LD introduces the notion of type maps for structuring data. The type indexing feature allows an author to structure data using a simple key-value map where the keys map to IRIs. This enables data to be structured based on the @type of specific node objects. In JSON-LD such data can be specified by associating the @type keyword with a @container declaration in the context:

In the example above, the affiliation term has been marked as a type map. The schema:Corporation and schema:ProfessionalService keys will be interpreted as the @type property of the node object value.

The value of @container can also be an array containing both @type and @set. When compacting, this ensures that a JSON-LD processor will use the array form for all values of types.

The special index @none is used for indexing node objects which do not have an @type, which is useful to maintain a normalized representation. The @none index may also be a term which expands to @none, such as the term none used in the example below.

As with id maps, when used with @type, a container may also include @set to ensure that key values are always contained in an array.

Type maps are a new feature in JSON-LD 1.1.

Included Nodes

Sometimes it is also useful to list node objects as part of another node object. For instance, to represent a set of resources which are used by some other resource. Included blocks may be also be used to collect such secondary node objects which can be referenced from a primary node object. For an example, consider a node object containing a list of different items, some of which share some common elements:

  
  

When flattened, this will move the `employee` and `contractor` elements from the included block into the outer array.

Included resources are described in Inclusion of Related Resources of [[[JSON.API]]] [[JSON.API]] as a way to include related resources associated with some primary resource; `@included` provides an analogous possibility in JSON-LD.

As a by product of the use of `@included` within node objects, a map may contain only `@included`, to provide a feature similar to that described in , where `@graph` is used to described disconnected nodes.

However, in contrast to `@graph`, `@included` does not interact with other properties contained within the same map, a feature discussed further in .

Reverse Properties

JSON-LD serializes directed graphs. That means that every property points from a node to another node or value. However, in some cases, it is desirable to serialize in the reverse direction. Consider for example the case where a person and its children should be described in a document. If the used vocabulary does not provide a children property but just a parent property, every node representing a child would have to be expressed with a property pointing to the parent as in the following example.

Expressing such data is much simpler by using JSON-LD's @reverse keyword:

The @reverse keyword can also be used in expanded term definitions to create reverse properties as shown in the following example:

Named Graphs

At times, it is necessary to make statements about a graph itself, rather than just a single node. This can be done by grouping a set of nodes using the @graph keyword. A developer may also name data expressed using the @graph keyword by pairing it with an @id keyword as shown in the following example:

The example above expresses a named graph that is identified by the IRI http://example.org/foaf-graph. That graph is composed of the statements about Manu and Gregg. Metadata about the graph itself is expressed via the generatedAt property, which specifies when the graph was generated.

When a JSON-LD document's top-level structure is a map that contains no other keys than @graph and optionally @context (properties that are not mapped to an IRI or a keyword are ignored), @graph is considered to express the otherwise implicit default graph. This mechanism can be useful when a number of nodes exist at the document's top level that share the same context, which is, e.g., the case when a document is flattened. The @graph keyword collects such nodes in an array and allows the use of a shared context.

In this case, embedding can not be used as the graph contains unrelated nodes. This is equivalent to using multiple node objects in array and defining the @context within each node object:

Graph Containers

In some cases, it is useful to logically partition data into separate graphs, without making this explicit within the JSON expression. For example, a JSON document may contain data against which other metadata is asserted and it is useful to separate this data in the data model using the notion of named graphs, without the syntactic overhead associated with the @graph keyword.

An expanded term definition can use @graph as the value of @container. This indicates that values of this term should be considered to be named graphs, where the graph name is an automatically assigned blank node identifier creating an implicitly named graph. When expanded, these become simple graph objects.

A different example uses an anonymously named graph as follows:

The example above expresses an anonymously named graph making a statement. The default graph includes a statement saying that the subject wrote that statement. This is an example of separating statements into a named graph, and then making assertions about the statements contained within that named graph.

Strictly speaking, the value of such a term is not a named graph, rather it is the graph name associated with the named graph, which exists separately within the dataset.

Graph Containers are a new feature in JSON-LD 1.1.

Named Graph Data Indexing

In addition to indexing node objects by index, graph objects may also be indexed by an index. By using the @graph container type, introduced in in addition to @index, an object value of such a property is treated as a key-value map where the keys do not map to IRIs, but are taken from an @index property associated with named graphs which are their values. When expanded, these must be simple graph objects

The following example describes a default graph referencing multiple named graphs using an index map.

As with index maps, when used with @graph, a container may also include @set to ensure that key values are always contained in an array.

The special index @none is used for indexing graphs which do not have an @index key, which is useful to maintain a normalized representation. Note, however, that compacting a document where multiple unidentified named graphs are compacted using the @none index will result in the content of those graphs being merged. To prevent this, give each graph a distinct @index key.

Named Graph Data Indexing is a new feature in JSON-LD 1.1.

Named Graph Indexing

In addition to indexing node objects by identifier, graph objects may also be indexed by their graph name. By using the @graph container type, introduced in in addition to @id, an object value of such a property is treated as a key-value map where the keys represent the identifiers of named graphs which are their values.

The following example describes a default graph referencing multiple named graphs using an id map.

As with id maps, when used with @graph, a container may also include @set to ensure that key values are always contained in an array.

As with id maps, the special index @none is used for indexing named graphs which do not have an @id, which is useful to maintain a normalized representation. The @none index may also be a term which expands to @none. Note, however, that if multiple graphs are represented without an @id, they will be merged on expansion. To prevent this, use @none judiciously, and consider giving graphs their own distinct identifier.

Graph Containers are a new feature in JSON-LD 1.1.

Loading Documents

The JSON-LD 1.1 Processing Algorithms and API specification [[JSON-LD11-API]] defines the interface to a JSON-LD Processor and includes a number of methods used for manipulating different forms of JSON-LD (see ). This includes a general mechanism for loading remote documents, including referenced JSON-LD documents and remote contexts, and potentially extracting embedded JSON-LD from other formats such as [[HTML]]. This is more fully described in Remote Document and Context Retrieval in [[JSON-LD11-API]].

A documentLoader can be useful in a number of contexts where loading remote documents can be problematic:

Forms of JSON-LD

As with many data formats, there is no single correct way to describe data in JSON-LD. However, as JSON-LD is used for describing graphs, certain transformations can be used to change the shape of the data, without changing its meaning as Linked Data.

Expanded Document Form
Expansion is the process of taking a JSON-LD document and applying a context so that the @context is no longer necessary. This process is described further in .
Compacted Document Form
Compaction is the process of applying a provided context to an existing JSON-LD document. This process is described further in .
Flattened Document Form
Flattening is the process of extracting embedded nodes to the top level of the JSON tree, and replacing the embedded node with a reference, creating blank node identifiers as necessary. This process is described further in .
Framed Document Form
Framing is used to shape the data in a JSON-LD document, using an example frame document which is used to both match the flattened data and show an example of how the resulting data should be shaped. This process is described further in .

Expanded Document Form

The JSON-LD 1.1 Processing Algorithms and API specification [[JSON-LD11-API]] defines a method for expanding a JSON-LD document. Expansion is the process of taking a JSON-LD document and applying a context such that all IRIs, types, and values are expanded so that the @context is no longer necessary.

For example, assume the following JSON-LD input document:

  
  

Running the JSON-LD Expansion algorithm against the JSON-LD input document provided above would result in the following output:

JSON-LD's media type defines a profile parameter which can be used to signal or request expanded document form. The profile URI identifying expanded document form is http://www.w3.org/ns/json-ld#expanded.

Compacted Document Form

The JSON-LD 1.1 Processing Algorithms and API specification [[JSON-LD11-API]] defines a method for compacting a JSON-LD document. Compaction is the process of applying a developer-supplied context to shorten IRIs to terms or compact IRIs and JSON-LD values expressed in expanded form to simple values such as strings or numbers. Often this makes it simpler to work with document as the data is expressed in application-specific terms. Compacted documents are also typically easier to read for humans.

For example, assume the following JSON-LD input document:

  
  

Additionally, assume the following developer-supplied JSON-LD context:

  
  

Running the JSON-LD Compaction algorithm given the context supplied above against the JSON-LD input document provided above would result in the following output:

JSON-LD's media type defines a profile parameter which can be used to signal or request compacted document form. The profile URI identifying compacted document form is http://www.w3.org/ns/json-ld#compacted.

The details of Compaction are described in the Compaction algorithm in [[JSON-LD11-API]]. This section provides a short description of how the algorithm operates as a guide to authors creating contexts to be used for compacting JSON-LD documents.

The purpose of compaction is to apply the term definitions, vocabulary mapping, default language, and base IRI to an existing JSON-LD document to cause it to be represented in a form that is tailored to the use of the JSON-LD document directly as JSON. This includes representing values as strings, rather than value objects, where possible, shortening the use of list objects into simple arrays, reversing the relationship between nodes, and using data maps to index into multiple values instead of representing them as an array of values.

Shortening IRIs

In an expanded JSON-LD document, IRIs are always represented as absolute IRIs. In many cases, it is preferable to use a shorter version, either a relative IRI reference, compact IRI, or term. Compaction uses a combination of elements in a context to create a shorter form of these IRIs. See , , and for more details.

The vocabulary mapping can be used to shorten IRIs that may be vocabulary relative by removing the IRI prefix that matches the vocabulary mapping. This is done whenever an IRI is determined to be vocabulary relative, i.e., used as a property, or a value of @type, or as the value of a term described as "@type": "@vocab".

Representing Values as Strings

To be unambiguous, the expanded document form always represents nodes and values using node objects and value objects. Moreover, property values are always contained within an array, even when there is only one value. Sometimes this is useful to maintain a uniformity of access, but most JSON data use the simplest possible representation, meaning that properties have single values, which are represented as strings or as structured values such as node objects. By default, compaction will represent values which are simple strings as strings, but sometimes a value is an IRI, a date, or some other typed value for which a simple string representation would loose information. By specifying this within a term definition, the semantics of a string value can be inferred from the definition of the term used as a property. See for more details.

Representing Lists as Arrays

As described in , JSON-LD has an expanded syntax for representing ordered values, using the @list keyword. To simplify the representation in JSON-LD, a term can be defined with "@container": "@list" which causes all values of a property using such a term to be considered ordered.

Reversing Node Relationships

In some cases, the property used to relate two nodes may be better expressed if the nodes have a reverse direction, for example, when describing a relationship between two people and a common parent. See for more details.

Reverse properties can be even more useful when combined with framing, which can actually make node objects defined at the top-level of a document to become embedded nodes. JSON-LD provides a means to index such values, by defining an appropriate @container definition within a term definition.

Indexing Values

Properties with multiple values are typically represented using an unordered array. This means that an application working on an internalized representation of that JSON would need to iterate through the values of the array to find a value matching a particular pattern, such as a language-tagged string using the language en.

Data can be indexed on a number of different keys, including @id, @type, @language, @index and more. See and for more details.

Normalizing Values as Objects

Sometimes it's useful to compact a document, but keep the node object and value object representations. For this, a term definition can set "@type": "@none". This causes the Value Compaction algorithm to always use the object form of values, although components of that value may be compacted.

Representing Singular Values as Arrays

Generally, when compacting, properties having only one value are represented as strings or maps, while properties having multiple values are represented as an array of strings or maps. This means that applications accessing such properties need to be prepared to accept either representation. To force all values to be represented using an array, a term definition can set "@container": "@set". Moreover, @set can be used in combination with other container settings, for example looking at our language-map example from :

Term Selection

When compacting, the Compaction algorithm will compact using a term for a property only when the values of that property match the @container, @type, and @language specifications for that term definition. This can actually split values between different properties, all of which have the same IRI. In case there is no matching term definition, the compaction algorithm will compact using the absolute IRI of the property.

Flattened Document Form

The JSON-LD 1.1 Processing Algorithms and API specification [[JSON-LD11-API]] defines a method for flattening a JSON-LD document. Flattening collects all properties of a node in a single map and labels all blank nodes with blank node identifiers. This ensures a shape of the data and consequently may drastically simplify the code required to process JSON-LD in certain applications.

For example, assume the following JSON-LD input document:

  
  

Running the JSON-LD Flattening algorithm against the JSON-LD input document in the example above and using the same context would result in the following output:

JSON-LD's media type defines a profile parameter which can be used to signal or request flattened document form. The profile URI identifying flattened document form is http://www.w3.org/ns/json-ld#flattened. It can be combined with the profile URI identifying expanded document form or compacted document form.

Framed Document Form

The JSON-LD 1.1 Framing specification [[JSON-LD11-FRAMING]] defines a method for framing a JSON-LD document. Framing is used to shape the data in a JSON-LD document, using an example frame document which is used to both match the flattened data and show an example of how the resulting data should be shaped.

For example, assume the following JSON-LD frame:

  
  

This frame document describes an embedding structure that would place objects with type Library at the top, with objects of type Book that were linked to the library object using the contains property embedded as property values. It also places objects of type Chapter within the referencing Book object as embedded values of the Book object.

When using a flattened set of objects that match the frame components:

  
  

The Frame Algorithm can create a new document which follows the structure of the frame:

JSON-LD's media type defines a profile parameter which can be used to signal or request framed document form. The profile URI identifying framed document form is http://www.w3.org/ns/json-ld#framed.

JSON-LD's media type also defines a profile parameter which can be used to identify a script element in an HTML document containing a frame. The first script element of type application/ld+json;profile=http://www.w3.org/ns/json-ld#frame will be used to find a frame..

Modifying Behavior with Link Relationships

Certain aspects of JSON-LD processing can be modified using HTTP Link Headers [[RFC8288]]. These can be used when retrieving resources that are not, themselves, JSON-LD, but can be interpreted as JSON-LD by using information in a Link Relation.

When processing normal JSON documents, a link relation can be specified using the HTTP Link Header returned when fetching a remote document, as described in .

In other cases, a resource may be returned using a representation that cannot easily be interpreted as JSON-LD. Normally, HTTP content negotiation would be used to allow a client to specify a preference for JSON-LD over another representation, but in certain situations, it is not possible or practical for a server to respond appropriately to such requests. For this, an HTTP Link Header can be used to provide an alternate location for a document to be used in place of the originally requested resource, as described in .

Interpreting JSON as JSON-LD

Ordinary JSON documents can be interpreted as JSON-LD by providing an explicit JSON-LD context document. One way to provide this is by using referencing a JSON-LD context document in an HTTP Link Header. Doing so allows JSON to be unambiguously machine-readable without requiring developers to drastically change their documents and provides an upgrade path for existing infrastructure without breaking existing clients that rely on the application/json media type or a media type with a +json suffix as defined in [[RFC6839]].

In order to use an external context with an ordinary JSON document, when retrieving an ordinary JSON document via HTTP, processors MUST attempt to retrieve any JSON-LD document referenced by a Link Header with:

The referenced document MUST have a top-level JSON object. The @context entry within that object is added to the top-level JSON object of the referencing document. If an array is at the top-level of the referencing document and its items are JSON objects, the @context subtree is added to all array items. All extra information located outside of the @context subtree in the referenced document MUST be discarded. Effectively this means that the active context is initialized with the referenced external context. A response MUST NOT contain more than one HTTP Link Header using the http://www.w3.org/ns/json-ld#context link relation.

Other mechanisms for providing a JSON-LD Context MAY be described for other URI schemes.

The JSON-LD 1.1 Processing Algorithms and API specification [[JSON-LD11-API]] provides for an expandContext option for specifying a context to use when expanding JSON documents programmatically.

The following example demonstrates the use of an external context with an ordinary JSON document over HTTP:

  
  

Please note that JSON-LD documents served with the application/ld+json media type MUST have all context information, including references to external contexts, within the body of the document. Contexts linked via a http://www.w3.org/ns/json-ld#context HTTP Link Header MUST be ignored for such documents.

Alternate Document Location

Documents which can't be directly interpreted as JSON-LD can provide an alternate location containing JSON-LD. One way to provide this is by referencing a JSON-LD document in an HTTP Link Header. This might be useful, for example, when the URL associated with a namespace naturally contains an HTML document, but the JSON-LD context associated with that URL is located elsewhere.

To specify an alternate location, a non-JSON resource (i.e., one using a media type other than `application/json` or a derivative) can return the alternate location using a Link Header with:

A response MUST NOT contain more than one HTTP Link Header using the alternate link relation with type="application/ld+json" .

Other mechanisms for providing an alternate location MAY be described for other URI schemes.

The following example demonstrates the use of an alternate location with an ordinary HTML document over HTTP:

  
  

A processor seeing a non-JSON result will note the presence of the link header and load that document instead.

Embedding JSON-LD in HTML Documents

This section describes features available with a documentLoader supporting HTML script extraction. See Remote Document and Context Retrieval for more information.

JSON-LD content can be easily embedded in HTML [[HTML]] by placing it in a script element with the type attribute set to application/ld+json. Doing so creates a data block.

Defining how such data may be used is beyond the scope of this specification. The embedded JSON-LD document might be extracted as is or, e.g., be interpreted as RDF.

If JSON-LD content is extracted as RDF [[RDF11-CONCEPTS]], it MUST be expanded into an RDF Dataset using the Deserialize JSON-LD to RDF Algorithm [[JSON-LD11-API]]. Unless a specific script is targeted (see ), all script elements with type application/ld+json MUST be processed and merged into a single dataset with equivalent blank node identifiers contained in separate script elements treated as if they were in a single document (i.e., blank nodes are shared between different JSON-LD script elements).

Inheriting base IRI from HTML's base element

When processing a JSON-LD script element, the Document Base URL of the containing HTML document, as defined in [[HTML]], is used to establish the default base IRI of the enclosed JSON-LD content.

HTML allows for Dynamic changes to base URLs. This specification does not require any specific behavior, and to ensure that all systems process the base IRI equivalently, authors SHOULD either use IRIs, or explicitly as defined in . Implementations (particularly those natively operating in the [[!DOM]]) MAY take into consideration Dynamic changes to base URLs.

Restrictions for contents of JSON-LD script elements

Due to the HTML Restrictions for contents of <script> elements additional encoding restrictions are placed on JSON-LD data contained in script elements.

Authors should avoid using character sequences in scripts embedded in HTML which may be confused with a comment-open, script-open, comment-close, or script-close.

Such content should be escaped as indicated below, however the content will remain escaped after processing through the JSON-LD API [[JSON-LD11-API]].
  • &amp; → & (ampersand, U+0026)
  • &lt; → < (less-than sign, U+003C)
  • &gt; → > (greater-than sign, U+003E)
  • &quot; → " (quotation mark, U+0022)
  • &apos; → ' (apostrophe, U+0027)

Locating a Specific JSON-LD Script Element

A specific script element within an HTML document may be located using a fragment identifier matching the unique identifier of the script element within the HTML document located by a URL (see [[!DOM]]). A JSON-LD processor MUST extract only the specified data block's contents parsing it as a standalone JSON-LD document and MUST NOT merge the result with any other markup from the same HTML document.

For example, given an HTML document located at http://example.com/document, a script element identified by "dave" can be targeted using the URL http://example.com/document#dave.

Data Model

JSON-LD is a serialization format for Linked Data based on JSON. It is therefore important to distinguish between the syntax, which is defined by JSON in [[RFC8259]], and the data model which is an extension of the RDF data model [[RDF11-CONCEPTS]]. The precise details of how JSON-LD relates to the RDF data model are given in .

To ease understanding for developers unfamiliar with the RDF model, the following summary is provided:

JSON-LD documents MAY contain data that cannot be represented by the data model defined above. Unless otherwise specified, such data is ignored when a JSON-LD document is being processed. One result of this rule is that properties which are not mapped to an IRI, a blank node, or keyword will be ignored.

Additionally, the JSON serialization format is internally represented using the JSON-LD internal representation, which uses the generic concepts of lists, maps, strings, numbers, booleans, and null to describe the data represented by a JSON document.

The image depicts a linked data dataset with a default graph and two named graphs.

An illustration of a linked data dataset.
A description of the linked data dataset diagram is available in the Appendix. Image available in SVG and PNG formats.

The dataset described in this figure can be represented as follows:

Note the use of @graph at the outer-most level to describe three top-level resources (two of them named graphs). The named graphs use @graph in addition to @id to provide the name for each graph.

JSON-LD Grammar

This section restates the syntactic conventions described in the previous sections more formally.

A JSON-LD document MUST be valid JSON text as described in [[RFC8259]], or some format that can be represented in the JSON-LD internal representation that is equivalent to valid JSON text.

A JSON-LD document MUST be a single node object, a map consisting of only the entries @context and/or @graph, or an array of zero or more node objects.

In contrast to JSON, in JSON-LD the keys in objects MUST be unique.

Whenever a keyword is discussed in this grammar, the statements also apply to an alias for that keyword.

JSON-LD allows keywords to be aliased (see for details). For example, if the active context defines the term id as an alias for @id, that alias may be legitimately used as a substitution for @id. Note that keyword aliases are not expanded during context processing.

Terms

A term is a short-hand string that expands to an IRI, blank node identifier, or keyword.

A term MUST NOT equal any of the JSON-LD keywords, other than `@type`.

When used as the prefix in a Compact IRI, to avoid the potential ambiguity of a prefix being confused with an IRI scheme, terms SHOULD NOT come from the list of URI schemes as defined in [[IANA-URI-SCHEMES]]. Similarly, to avoid confusion between a Compact IRI and a term, terms SHOULD NOT include a colon (:) and SHOULD be restricted to the form of isegment-nz-nc as defined in [[RFC3987]].

To avoid forward-compatibility issues, a term SHOULD NOT start with an @ character followed exclusively by one or more ALPHA characters (see [[RFC5234]]) as future versions of JSON-LD may introduce additional keywords. Furthermore, the term MUST NOT be an empty string ("") as not all programming languages are able to handle empty JSON keys.

See and for further discussion on mapping terms to IRIs.

Node Objects

A node object represents zero or more properties of a node in the graph serialized by the JSON-LD document. A map is a node object if it exists outside of a JSON-LD context and:

The properties of a node in a graph may be spread among different node objects within a document. When that happens, the keys of the different node objects need to be merged to create the properties of the resulting node.

A node object MUST be a map. All keys which are not IRIs, compact IRIs, terms valid in the active context, or one of the following keywords (or alias of such a keyword) MUST be ignored when processed:

If the node object contains the @context key, its value MUST be null, an IRI reference, a context definition, or an array composed of any of these.

If the node object contains the @id key, its value MUST be an IRI reference, or a compact IRI (including blank node identifiers). See , , and for further discussion on @id values.

If the node object contains the @graph key, its value MUST be a node object or an array of zero or more node objects. If the node object also contains an @id keyword, its value is used as the graph name of a named graph. See for further discussion on @graph values. As a special case, if a map contains no keys other than @graph and @context, and the map is the root of the JSON-LD document, the map is not treated as a node object; this is used as a way of defining node objects that may not form a connected graph. This allows a context to be defined which is shared by all of the constituent node objects.

If the node object contains the @type key, its value MUST be either an IRI reference, a compact IRI (including blank node identifiers), a term defined in the active context expanding into an IRI, or an array of any of these. See for further discussion on @type values.

If the node object contains the @reverse key, its value MUST be a map containing entries representing reverse properties. Each value of such a reverse property MUST be an IRI reference, a compact IRI, a blank node identifier, a node object or an array containing a combination of these.

If the node object contains the `@included` key, its value MUST be an included block. See for further discussion on included blocks.

If the node object contains the @index key, its value MUST be a string. See for further discussion on @index values.

If the node object contains the @nest key, its value MUST be a map or an array of map which MUST NOT include a value object. See for further discussion on @nest values.

Keys in a node object that are not keywords MAY expand to an IRI using the active context. The values associated with keys that expand to an IRI MUST be one of the following:

Frame Objects

When framing, a frame object extends a node object to allow entries used specifically for framing.

See [[JSON-LD11-FRAMING]] for a description of how frame objects are used.

Graph Objects

A graph object represents a named graph, which MAY include an explicit graph name. A map is a graph object if it exists outside of a JSON-LD context, it contains an @graph entry (or an alias of that keyword), it is not the top-most map in the JSON-LD document, and it consists of no entries other than @graph, @index, @id and @context, or an alias of one of these keywords.

If the graph object contains the @context key, its value MUST be null, an IRI reference, a context definition, or an array composed of any of these.

If the graph object contains the @id key, its value is used as the identifier (graph name) of a named graph, and MUST be an IRI reference, or a compact IRI (including blank node identifiers). See , , and for further discussion on @id values.

A graph object without an @id entry is also a simple graph object and represents a named graph without an explicit identifier, although in the data model it still has a graph name, which is an implicitly allocated blank node identifier.

The value of the @graph key MUST be a node object or an array of zero or more node objects. See for further discussion on @graph values..

Value Objects

A value object is used to explicitly associate a type or a language with a value to create a typed value or a language-tagged string and possibly associate a base direction.

A value object MUST be a map containing the @value key. It MAY also contain an @type, an @language, an `@direction`, an @index, or an @context key but MUST NOT contain both an @type and either @language or `@direction` keys at the same time. A value object MUST NOT contain any other keys that expand to an IRI or keyword.

The value associated with the @value key MUST be either a string, a number, true, false or null. If the value associated with the @type key is @json, the value MAY be either an array or an object.

The value associated with the @type key MUST be a term, an IRI, a compact IRI, a string which can be turned into an IRI using the vocabulary mapping, @json, or null.

The value associated with the @language key MUST have the lexical form described in [[BCP47]], or be null.

The value associated with the @direction key MUST be one of "ltr" or "rtl", or be null.

The value associated with the @index key MUST be a string.

See and for more information on value objects.

Value Patterns

When framing, a value pattern extends a value object to allow entries used specifically for framing.

Lists and Sets

A list represents an ordered set of values. A set represents an unordered set of values. Unless otherwise specified, arrays are unordered in JSON-LD. As such, the @set keyword, when used in the body of a JSON-LD document, represents just syntactic sugar which is optimized away when processing the document. However, it is very helpful when used within the context of a document. Values of terms associated with an @set or @list container will always be represented in the form of an array when a document is processed—even if there is just a single value that would otherwise be optimized to a non-array form in compacted document form. This simplifies post-processing of the data as the data is always in a deterministic form.

A list object MUST be a map that contains no keys that expand to an IRI or keyword other than @list and @index.

A set object MUST be a map that contains no keys that expand to an IRI or keyword other than @set and @index. Please note that the @index key will be ignored when being processed.

In both cases, the value associated with the keys @list and @set MUST be one of the following types:

See for further discussion on sets and lists.

Language Maps

A language map is used to associate a language with a value in a way that allows easy programmatic access. A language map may be used as a term value within a node object if the term is defined with @container set to @language, or an array containing both @language and @set . The keys of a language map MUST be strings representing [[BCP47]] language tags, the keyword @none, or a term which expands to @none, and the values MUST be any of the following types:

See for further discussion on language maps.

Index Maps

An index map allows keys that have no semantic meaning, but should be preserved regardless, to be used in JSON-LD documents. An index map may be used as a term value within a node object if the term is defined with @container set to @index, or an array containing both @index and @set . The values of the entries of an index map MUST be one of the following types:

See for further information on this topic.

Index Maps may also be used to map indexes to associated named graphs, if the term is defined with @container set to an array containing both @graph and @index, and optionally including @set. The value consists of the node objects contained within the named graph which is indexed using the referencing key, which can be represented as a simple graph object if the value does not include @id, or a named graph if it includes @id.

Property-based Index Maps

A property-based index map is a variant of index map where indexes are semantically preserved in the graph as property values. A property-based index map may be used as a term value within a node object if the term is defined with @container set to @index, or an array containing both @index and @set, and with @index set to a string. The values of a property-based index map MUST be node objects or strings which expand to node objects.

When expanding, if the active context contains a term definition for the value of @index, this term definition will be used to expand the keys of the index map. Otherwise, the keys will be expanded as simple value objects. Each node object in the expanded values of the index map will be added an additional property value, where the property is the expanded value of @index, and the value is the expanded referencing key.

See for further information on this topic.

Id Maps

An id map is used to associate an IRI with a value that allows easy programmatic access. An id map may be used as a term value within a node object if the term is defined with @container set to @id, or an array containing both @id and @set. The keys of an id map MUST be IRIs (IRI references or compact IRIs (including blank node identifiers)), the keyword @none, or a term which expands to @none, and the values MUST be node objects.

If the value contains a property expanding to @id, its value MUST be equivalent to the referencing key. Otherwise, the property from the value is used as the @id of the node object value when expanding.

Id Maps may also be used to map graph names to their named graphs, if the term is defined with @container set to an array containing both @graph and @id, and optionally including @set. The value consists of the node objects contained within the named graph which is named using the referencing key.

Type Maps

A type map is used to associate an IRI with a value that allows easy programmatic access. A type map may be used as a term value within a node object if the term is defined with @container set to @type, or an array containing both @type and @set. The keys of a type map MUST be IRIs (IRI references or compact IRI (including blank node identifiers)), terms, or the keyword @none, and the values MUST be node objects or strings which expand to node objects.

If the value contains a property expanding to @type, and its value contains the referencing key after suitable expansion of both the referencing key and the value, then the node object already contains the type. Otherwise, the property from the value is added as a @type of the node object value when expanding.

Included Blocks

An included block is used to provide a set of node objects. An included block MAY appear as the value of a member of a node object with either the key of `@included` or an alias of `@included`. An included block is either a node object or an array of node objects.

When expanding, multiple included blocks will be coalesced into a single included block.

Property Nesting

A nested property is used to gather properties of a node object in a separate map, or array of maps which are not value objects. It is semantically transparent and is removed during the process of expansion. Property nesting is recursive, and collections of nested properties may contain further nesting.

Semantically, nesting is treated as if the properties and values were declared directly within the containing node object.

Context Definitions

A context definition defines a local context in a node object.

A context definition MUST be a map whose keys MUST be either terms, compact IRIs, IRIs, or one of the keywords @base, @import, @language, @propagate, @protected, @type, @version, or @vocab.

If the context definition has an @base key, its value MUST be an IRI reference, or null.

If the context definition has an @direction key, its value MUST be one of "ltr" or "rtl", or be null.

If the context definition contains the @import keyword, its value MUST be an IRI reference. When used as a reference from an `@import`, the referenced context definition MUST NOT include an `@import` key, itself.

If the context definition has an @language key, its value MUST have the lexical form described in [[BCP47]] or be null.

If the context definition has an @propagate key, its value MUST be true or false.

If the context definition has an @protected key, its value MUST be true or false.

If the context definition has an @type key, its value MUST be a map with only the entry @container set to @set, and optionally an entry `@protected`.

If the context definition has an @version key, its value MUST be a number with the value 1.1.

If the context definition has an @vocab key, its value MUST be an IRI reference, a compact IRI, a blank node identifier, a term, or null.

The value of keys that are not keywords MUST be either an IRI, a compact IRI, a term, a blank node identifier, a keyword, null, or an expanded term definition.

Expanded term definition

An expanded term definition is used to describe the mapping between a term and its expanded identifier, as well as other properties of the value associated with the term when it is used as key in a node object.

An expanded term definition MUST be a map composed of zero or more keys from @id, @reverse, @type, @language, @container, @context, @prefix, @propagate, or @protected. An expanded term definition SHOULD NOT contain any other keys.

When the associated term is `@type`, the expanded term definition MUST NOT contain keys other than `@container` and `@protected`. The value of `@container` is limited to the single value `@set`.

If the term being defined is not an IRI or a compact IRI and the active context does not have an @vocab mapping, the expanded term definition MUST include the @id key.

Term definitions with keys which are of the form of an IRI or a compact IRI MUST NOT expand to an IRI other than the expansion of the key itself.

If the expanded term definition contains the @id keyword, its value MUST be null, an IRI, a blank node identifier, a compact IRI, a term, or a keyword.

If an expanded term definition has an @reverse entry, it MUST NOT have @id or @nest entries at the same time, its value MUST be an IRI, a blank node identifier, a compact IRI, or a term. If an @container entry exists, its value MUST be null, @set, or @index.

If the expanded term definition contains the @type keyword, its value MUST be an IRI, a compact IRI, a term, null, or one of the keywords @id, @json, @none, or @vocab.

If the expanded term definition contains the @language keyword, its value MUST have the lexical form described in [[BCP47]] or be null.

If the expanded term definition contains the `@index` keyword, its value MUST be an IRI, a compact IRI, or a term.

If the expanded term definition contains the @container keyword, its value MUST be either @list, @set, @language, @index, @id, @graph, @type, or be null or an array containing exactly any one of those keywords, or a combination of @set and any of @index, @id, @graph, @type, @language in any order . @container may also be an array containing @graph along with either @id or @index and also optionally including @set. If the value is @language, when the term is used outside of the @context, the associated value MUST be a language map. If the value is @index, when the term is used outside of the @context, the associated value MUST be an index map.

If an expanded term definition has an @context entry, it MUST be a valid context definition.

If the expanded term definition contains the @nest keyword, its value MUST be either @nest, or a term which expands to @nest.

If the expanded term definition contains the @prefix keyword, its value MUST be true or false.

If the expanded term definition contains the @propagate keyword, its value MUST be true or false.

If the expanded term definition contains the @protected keyword, its value MUST be true or false.

Terms MUST NOT be used in a circular manner. That is, the definition of a term cannot depend on the definition of another term if that other term also depends on the first term.

See for further discussion on contexts.

Keywords

JSON-LD keywords are described in , this section describes where each keyword may appear within different JSON-LD structures.

Within node objects, value objects, graph objects, list objects, set objects, and nested properties keyword aliases MAY be used instead of the corresponding keyword, except for `@context`. The `@context` keyword MUST NOT be aliased. Within local contexts and expanded term definitions, keyword aliases MAY NOT be used.

@base
The unaliased @base keyword MAY be used as a key in a context definition. Its value MUST be an IRI reference, or null.
@container
The unaliased @container keyword MAY be used as a key in an expanded term definition. Its value MUST be either @list, @set, @language, @index, @id, @graph, @type, or be null, or an array containing exactly any one of those keywords, or a combination of @set and any of @index, @id, @graph, @type, @language in any order. The value may also be an array containing @graph along with either @id or @index and also optionally including @set.
@context
The @context keyword MUST NOT be aliased, and MAY be used as a key in the following objects: The value of @context MUST be null, an IRI reference, a context definition, or an array composed of any of these.
`@graph`
The `@graph` keyword MAY be aliased and MAY be used as a key in a node object or a graph object, where its value MUST be a value object, node object, or an array of either value objects or node objects.

The unaliased `@graph` MAY be used as the value of the @container key within an expanded term definition.

See .

@id
The @id keyword MAY be aliased and MAY be used as a key in a node object or a graph object.

The unaliased @id MAY be used as a key in an expanded term definition, or as the value of the @container key within an expanded term definition.

The value of the @id key MUST be an IRI reference, or a compact IRI (including blank node identifiers).

See , , and for further discussion on @id values.

@index
The @index keyword MAY be aliased and MAY be used as a key in a node object, value object, graph object, set object, or list object. Its value MUST be a string.

The unaliased @index MAY be used as the value of the @container key within an expanded term definition and as an entry in an expanded term definition, where the value is an IRI, a compact IRI, or a term.

See , and for a further discussion.

@language
The @language keyword MAY be aliased and MAY be used as a key in a value object. Its value MUST be a string with the lexical form described in [[BCP47]] or be null.

The unaliased @language MAY be used as a key in a context definition, or as the value of the @container key within an expanded term definition.

See , .

`@direction`
The @direction keyword MAY be aliased and MAY be used as a key in a value object. Its value MUST be one of "ltr" or "rtl", or be null.

The unaliased @direction MAY be used as a key in a context definition.

See for a further discussion.

@list
The @list keyword MAY be aliased and MUST be used as a key in a list object. The unaliased @list MAY be used as the value of the @container key within an expanded term definition. Its value MUST be one of the following:

See for further discussion on sets and lists.

@nest
The @nest keyword MAY be aliased and MAY be used as a key in a node object, where its value must be a map.

The unaliased @nest MAY be used as the value of a simple term definition, or as a key in an expanded term definition, where its value MUST be a string expanding to @nest.

See for a further discussion.

@none
The @none keyword MAY be aliased and MAY be used as a key in an index map, id map, language map, type map. See , , , , , or for a further discussion.
@prefix
The unaliased @prefix keyword MAY be used as a key in an expanded term definition. Its value MUST be true or false. See and for a further discussion.
`@propagate`
The unaliased `@propagate` keyword MAY be used in a context definition. Its value MUST be true or false. See for a further discussion.
`@protected`
The unaliased `@protected` keyword MAY be used in a context definition, or an expanded term definition. Its value MUST be true or false. See for a further discussion.
@reverse
The @reverse keyword MAY be aliased and MAY be used as a key in a node object.

The unaliased @reverse MAY be used as a key in an expanded term definition.

The value of the @reverse key MUST be an IRI reference, or a compact IRI (including blank node identifiers).

See and for further discussion.

@set
The @set keyword MAY be aliased and MUST be used as a key in a set object. Its value MUST be one of the following:

The unaliased @set MAY be used as the value of the @container key within an expanded term definition.

See for further discussion on sets and lists.

`@import`
The unaliased `@import` keyword MAY be used in a context definition. Its value MUST be an IRI reference. See for a further discussion.
@type
The @type keyword MAY be aliased and MAY be used as a key in a node object or a value object, where its value MUST be a term, IRI reference, or a compact IRI (including blank node identifiers).

The unaliased @type MAY be used as a key in an expanded term definition, where its value may also be either @id or @vocab, or as the value of the @container key within an expanded term definition.

Within a context, `@type` may be used as the key for an expanded term definition, whose entries are limited to `@container` and `@protected`.

This keyword is described further in and .

@value
The @value keyword MAY be aliased and MUST be used as a key in a value object. Its value key MUST be either a string, a number, true, false or null. This keyword is described further in .
@version
The unaliased @version keyword MAY be used as a key in a context definition. Its value MUST be a number with the value 1.1. This keyword is described further in .
@vocab
The unaliased @vocab keyword MAY be used as a key in a context definition or as the value of @type in an expanded term definition. Its value MUST be an IRI reference, a compact IRI, a blank node identifier, a term, or null. This keyword is described further in , and .
`@included`
The `@included` keyword MAY be aliased and its value MUST be an included block. This keyword is described further in , and .
`@json`
The @json keyword MAY be aliased and MAY be used as the value of the `@type` key within a value object or an expanded term definition.

See .

Relationship to RDF

JSON-LD is a concrete RDF syntax as described in [[RDF11-CONCEPTS]]. Hence, a JSON-LD document is both an RDF document and a JSON document and correspondingly represents an instance of an RDF data model. However, JSON-LD also extends the RDF data model to optionally allow JSON-LD to serialize generalized RDF Datasets. The JSON-LD extensions to the RDF data model are:

The use of blank node identifiers to label properties is obsolete, and may be removed in a future version of JSON-LD, as is the support for generalized RDF Datasets.

Summarized, these differences mean that JSON-LD is capable of serializing any RDF graph or dataset and most, but not all, JSON-LD documents can be directly interpreted as RDF as described in RDF 1.1 Concepts [[RDF11-CONCEPTS]].

Authors are strongly encouraged to avoid labeling properties using blank node identifiers, instead, consider one of the following mechanisms:

The normative algorithms for interpreting JSON-LD as RDF and serializing RDF as JSON-LD are specified in the JSON-LD 1.1 Processing Algorithms and API specification [[JSON-LD11-API]].

Even though JSON-LD serializes RDF Datasets, it can also be used as a graph source. In that case, a consumer MUST only use the default graph and ignore all named graphs. This allows servers to expose data in languages such as Turtle and JSON-LD using HTTP content negotiation.

Publishers supporting both dataset and graph syntaxes have to ensure that the primary data is stored in the default graph to enable consumers that do not support datasets to process the information.

Serializing/Deserializing RDF

The process of serializing RDF as JSON-LD and deserializing JSON-LD to RDF depends on executing the algorithms defined in RDF Serialization-Deserialization Algorithms in the JSON-LD 1.1 Processing Algorithms and API specification [[JSON-LD11-API]]. It is beyond the scope of this document to detail these algorithms any further, but a summary of the necessary operations is provided to illustrate the process.

The procedure to deserialize a JSON-LD document to RDF involves the following steps:

  1. Expand the JSON-LD document, removing any context; this ensures that properties, types, and values are given their full representation as IRIs and expanded values. Expansion is discussed further in .
  2. Flatten the document, which turns the document into an array of node objects. Flattening is discussed further in .
  3. Turn each node object into a series of triples.

For example, consider the following JSON-LD document in compact form:

    
    

Running the JSON-LD Expansion and Flattening algorithms against the JSON-LD input document in the example above would result in the following output:

    
    

Deserializing this to RDF now is a straightforward process of turning each node object into one or more triples. This can be expressed in Turtle as follows:

    
    

The process of serializing RDF as JSON-LD can be thought of as the inverse of this last step, creating an expanded JSON-LD document closely matching the triples from RDF, using a single node object for all triples having a common subject, and a single property for those triples also having a common predicate. The result may then be framed by using the Framing Algorithm described in [[JSON-LD11-FRAMING]] to create the desired object embedding.

The rdf:JSON Datatype

RDF provides for JSON content as a possible literal value. This allows markup in literal values. Such content is indicated in a graph using a literal whose datatype is set to rdf:JSON.

The rdf:JSON datatype is defined as follows:

The IRI denoting this datatype
is http://www.w3.org/1999/02/22-rdf-syntax-ns#JSON.
The lexical space
is the set of UNICODE [[UNICODE]] strings which conform to the JSON Grammar as described in Section 2 JSON Grammar of [[RFC8259]].
The value space
is the set of UNICODE [[UNICODE]] strings which conform to the JSON Grammar as described in Section 2 JSON Grammar of [[RFC8259]], and furthermore comply with the following constraints:
  • It MUST NOT contain any unnecessary whitespace,
  • Keys in objects MUST be in code point order,
  • Native Numeric values MUST be serialized according to Section 7.1.12.1 of [[?ECMASCRIPT]],
  • Strings MUST be serialized with Unicode codepoints from U+0000 through U+001F using lower case hexadecimal Unicode notation (\uhhhh) unless in the set of predefined JSON control characters U+0008, U+0009, U+000A, U+000C or U+000D which SHOULD be serialized as \b, \t, \n, \f and \r respectively. All other Unicode characters SHOULD be serialized "as is", other than U+005C (\) and U+0022 (") which SHOULD be serialized as \\ and \" respectively.
The JSON Canonicalization Scheme (JCS) [[RFC8785]] is an emerging standard for JSON canonicalization. This specification will likely be updated to require such a canonical representation. Users are cautioned from depending on the JSON literal lexical representation as an RDF literal, as the specifics of serialization may change in a future revision of this document.
Despite being defined as a set of strings, this value space is considered distinct from the value space of `xsd:string`, in order to avoid side effects with existing specifications.
The lexical-to-value mapping
maps any element of the lexical space to the result of
  1. parsing it into an internal representation consistent with [[ECMASCRIPT]] representation created by using the JSON.parse function as defined in Section 24.5 The JSON Object of [[ECMASCRIPT]],
  2. then serializing it in the JSON format [[RFC8259]] in compliance with the constraints of the value space described above.
The canonical mapping
maps any element of the value space to the identical string in the lexical space.

The `i18n` Namespace

The `i18n` namespace is used for describing combinations of language tag and base direction in RDF literals. It is used as an alternative mechanism for describing the [[BCP47]] language tag and base direction of RDF literals that would otherwise use the `xsd:string` or `rdf:langString` datatypes.

Datatypes based on this namespace allow round-tripping of JSON-LD documents using base direction, although the mechanism is not otherwise standardized.

The Deserialize JSON-LD to RDF Algorithm can be used with the rdfDirection option set to `i18n-datatype` to generate RDF literals using the `i18n` base to create an IRI encoding the base direction along with optional language tag (normalized to lower case) from value objects containing `@direction` by appending to `https://www.w3.org/ns/i18n#` the value of `@language`, if any, followed by an underscore (`"_"`) followed by the value of `@direction`.

For improved interoperability, the language tag is normalized to lower case when creating the datatype IRI.

The following example shows two statements with literal values of `i18n:ar-EG_rtl`, which encodes the language tag `ar-EG` and the base direction `rtl`.

    
    

See for more details on using base direction for strings.

The `rdf:CompoundLiteral` class and the `rdf:language` and `rdf:direction` properties

This specification defines the `rdf:CompoundLiteral` class, which is in the domain of `rdf:language` and `rdf:direction` to be used for describing RDF literal values containing base direction and a possible language tag to be associated with the string value of `rdf:value` on the same subject.

`rdf:CompoundLiteral`
A class representing a compound literal.
`rdf:language`
An RDF property. The range of the property is an `rdfs:Literal`, whose value MUST be a well-formed [[BCP47]] language tag. The domain of the property is `rdf:CompoundLiteral`.
`rdf:direction`
An RDF property. The range of the property is an `rdfs:Literal`, whose value MUST be either `"ltr"` or `"rtl"`. The domain of the property is `rdf:CompoundLiteral`.

The Deserialize JSON-LD to RDF Algorithm can be used with the rdfDirection option set to `compound-literal` to generate RDF literals using these properties to describe the base direction and optional language tag (normalized to lower case) from value objects containing `@direction` and optionally `@language`.

For improved interoperability, the language tag is normalized to lower case when creating the datatype IRI.

The following example shows two statements with compound literals representing strings with the language tag `ar-EG` and base direction `rtl`.

    
    

See for more details on using base direction for strings.

Security Considerations

See RFC 8259, section 12 [[RFC8259]].

Since JSON-LD is intended to be a pure data exchange format for directed graphs, the serialization SHOULD NOT be passed through a code execution mechanism such as JavaScript's eval() function to be parsed. An (invalid) document may contain code that, when executed, could lead to unexpected side effects compromising the security of a system.

When processing JSON-LD documents, links to remote contexts and frames are typically followed automatically, resulting in the transfer of files without the explicit request of the user for each one. If remote contexts are served by third parties, it may allow them to gather usage patterns or similar information leading to privacy concerns. Specific implementations, such as the API defined in the JSON-LD 1.1 Processing Algorithms and API specification [[JSON-LD11-API]], may provide fine-grained mechanisms to control this behavior.

JSON-LD contexts that are loaded from the Web over non-secure connections, such as HTTP, run the risk of being altered by an attacker such that they may modify the JSON-LD active context in a way that could compromise security. It is advised that any application that depends on a remote context for mission critical purposes vet and cache the remote context before allowing the system to use it.

Given that JSON-LD allows the substitution of long IRIs with short terms, JSON-LD documents may expand considerably when processed and, in the worst case, the resulting data might consume all of the recipient's resources. Applications should treat any data with due skepticism.

As JSON-LD places no limits on the IRI schemes that may be used, and vocabulary-relative IRIs use string concatenation rather than IRI resolution, it is possible to construct IRIs that may be used maliciously, if dereferenced.

Future versions of this specification may incorporate subresource integrity [[?SRI]] as a means of ensuring that cached and retrieved content matches data retrieved from remote servers; see issue 86.

Privacy Considerations

The retrieval of external contexts can expose the operation of a JSON-LD processor, allow intermediate nodes to fingerprint the client application through introspection of retrieved resources (see [[?fingerprinting-guidance]]), and provide an opportunity for a man-in-the-middle attack. To protect against this, publishers should consider caching remote contexts for future use, or use the documentLoader to maintain a local version of such contexts.

Internationalization Considerations

As JSON-LD uses the RDF data model, it is restricted by design in its ability to properly record JSON-LD Values which are strings with left-to-right or right-to-left direction indicators. Both JSON-LD and RDF provide a mechanism for specifying the language associated with a string (language-tagged string), but do not provide a means of indicating the base direction of the string.

Unicode provides a mechanism for signaling direction within a string (see [[[UAX9]]] [[UAX9]]), however, when a string has an overall base direction which cannot be determined by the beginning of the string, an external indicator is required, such as the [[HTML]] dir attribute, which currently has no counterpart for RDF literals.

The issue of properly representing base direction in RDF is not something that this Working Group can handle, as it is a limitation or the core RDF data model. This Working Group expects that a future RDF Working Group will consider the matter and add the ability to specify the base direction of language-tagged strings.

Until a more comprehensive solution can be addressed in a future version of this specification, publishers should consider this issue when representing strings where the base direction of the string cannot otherwise be correctly inferred based on the content of the string. See [[?string-meta]] for a discussion best practices for identifying language and base direction for strings used on the Web.

Image Descriptions

Linked Data Dataset

This section describes the Linked Data Dataset figure in .

The image consists of three dashed boxes, each describing a different linked data graph. Each box consists of shapes linked with arrows describing the linked data relationships.

The first box is titled "default graph: <no name>" describes two resources: http://example.com/people/alice and http://example.com/people/bob (denoting "Alice" and "Bob" respectively), which are connected by an arrow labeled schema:knows which describes the knows relationship between the two resources. Additionally, the "Alice" resource is related to three different literals:

Alice
an RDF literal with no datatype or language.
weiblich | de
a language-tagged string with the value "weiblich" and language tag "de".
female | en
a language-tagged string with the value "female" and language tag "en".

The second and third boxes describe two named graphs, with the graph names "http://example.com/graphs/1" and "http://example.com/graphs/2", respectively.

The second box consists of two resources: http://example.com/people/alice and http://example.com/people/bob related by the schema:parent relationship, and names the http://example.com/people/bob "Bob".

The third box consists of two resources, one named http://example.com/people/bob and the other unnamed. The two resources related to each other using schema:sibling relationship with the second named "Mary".

Relationship to Other Linked Data Formats

The JSON-LD examples below demonstrate how JSON-LD can be used to express semantic data marked up in other linked data formats such as Turtle, RDFa, and Microdata. These sections are merely provided as evidence that JSON-LD is very flexible in what it can express across different Linked Data approaches.

Turtle

The following are examples of transforming RDF expressed in [[Turtle]] into JSON-LD.

Prefix definitions

The JSON-LD context has direct equivalents for the Turtle @prefix declaration:

      
      
      
      

Embedding

Both [[Turtle]] and JSON-LD allow embedding, although [[Turtle]] only allows embedding of blank nodes.

      
      
      
      

Conversion of native data types

In JSON-LD numbers and boolean values are native data types. While [[Turtle]] has a shorthand syntax to express such values, RDF's abstract syntax requires that numbers and boolean values are represented as typed literals. Thus, to allow round-tripping, Section 8.6 of the the JSON-LD 1.1 Processing Algorithms and API specification [[JSON-LD11-API]] defines conversion rules between JSON-LD's native data types and RDF's counterparts. Numbers without fractions are converted to xsd:integer-typed literals, numbers with fractions to xsd:double-typed literals and the two boolean values true and false to a xsd:boolean-typed literal. All typed literals are in canonical lexical form.

      
      
      
      
Note that this interpretation differs from [[Turtle]], in which the literal `2.78` translates to an `xsd:decimal`. The rationale is that most JSON tools parse numbers with fractions as floating point numbers, so `xsd:double` is the most appropriate datatype to render them back in RDF.
Native JSON numbers are sometimes converted to `xsd:double` (numbers with a fractional part, or very big integers) and this conversion can be lossy. Therefore, the use of native JSON numbers is discouraged when the exact value of the numbers must be preserved. Instead, it is advised to use value objects with the appropriate datatype.

Lists

Both JSON-LD and [[Turtle]] can represent sequential lists of values.

      
      
      
      

RDFa

The following example describes three people with their respective names and homepages in RDFa [[RDFA-CORE]].

    
    

An example JSON-LD implementation using a single context is described below.

    
    

Microdata

The HTML Microdata [[MICRODATA]] example below expresses book information as a Microdata Work item.

    
    

Note that the JSON-LD representation of the Microdata information stays true to the desires of the Microdata community to avoid contexts and instead refer to items by their full IRI.

    
    

IANA Considerations

This section has been submitted to the Internet Engineering Steering Group (IESG) for review, approval, and registration with IANA.

Media Type Registration: `application/ld+json`

Type name:
application
Subtype name:
ld+json
Required parameters:
N/A
Optional parameters:
profile

A non-empty list of space-separated URIs identifying specific constraints or conventions that apply to a JSON-LD document according to [[RFC6906]]. A profile does not change the semantics of the resource representation when processed without profile knowledge, so that clients both with and without knowledge of a profiled resource can safely use the same representation. The profile parameter MAY be used by clients to express their preferences in the content negotiation process. If the profile parameter is given, a server SHOULD return a document that honors the profiles in the list which it recognizes, and MUST ignore the profiles in the list which it does not recognize. It is RECOMMENDED that profile URIs are dereferenceable and provide useful documentation at that URI. For more information and background please refer to [[RFC6906]].

This specification defines six values for the profile parameter.

http://www.w3.org/ns/json-ld#expanded
To request or specify expanded JSON-LD document form.
http://www.w3.org/ns/json-ld#compacted
To request or specify compacted JSON-LD document form.
http://www.w3.org/ns/json-ld#context
To request or specify a JSON-LD context document.
http://www.w3.org/ns/json-ld#flattened
To request or specify flattened JSON-LD document form.
http://www.w3.org/ns/json-ld#frame
To request or specify a JSON-LD frame document.
http://www.w3.org/ns/json-ld#framed
To request or specify framed JSON-LD document form.

All other URIs starting with http://www.w3.org/ns/json-ld are reserved for future use by JSON-LD specifications.

Other specifications may publish additional `profile` parameter URIs with their own defined semantics. This includes the ability to associate a file extension with a `profile` parameter.

When used as a media type parameter [[RFC4288]] in an HTTP Accept header [[RFC7231]], the value of the profile parameter MUST be enclosed in quotes (") if it contains special characters such as whitespace, which is required when multiple profile URIs are combined.

When processing the "profile" media type parameter, it is important to note that its value contains one or more URIs and not IRIs. In some cases it might therefore be necessary to convert between IRIs and URIs as specified in section 3 Relationship between IRIs and URIs of [[RFC3987]].

Encoding considerations:
See RFC 8259, section 11.
Security considerations:
See .
Interoperability considerations:
Not Applicable
Published specification:
http://www.w3.org/TR/json-ld
Applications that use this media type:
Any programming environment that requires the exchange of directed graphs. Implementations of JSON-LD have been created for JavaScript, Python, Ruby, PHP, and C++.
Additional information:
Magic number(s):
Not Applicable
File extension(s):
.jsonld
Macintosh file type code(s):
TEXT
Person & email address to contact for further information:
Pierre-Antoine Champin <pierre-antoine@w3.org>
Intended usage:
Common
Restrictions on usage:
N/A
Author(s):
Manu Sporny, Dave Longley, Gregg Kellogg, Markus Lanthaler, Niklas Lindström
Change controller:
W3C

Fragment identifiers used with application/ld+json are treated as in RDF syntaxes, as per RDF 1.1 Concepts and Abstract Syntax [[RDF11-CONCEPTS]].

This registration is an update to the original definition for application/ld+json in [[JSON-LD10]].

Examples

The following examples illustrate different ways in which the profile parameter may be used to describe different acceptable responses.

      
      

Requests the server to return the requested resource as JSON-LD in expanded document form.

      
      

Requests the server to return the requested resource as JSON-LD in compacted document form. As no explicit context resource is specified, the server compacts using an application-specific default context.

      
      

Requests the server to return the requested resource as JSON-LD in both compacted document form and flattened document form. Note that as whitespace is used to separate the two URIs, they are enclosed in double quotes (").

Structured Suffix Registration: `+ld+json`

This section will be submitted to the Internet Engineering Steering Group (IESG) for review, approval, and registration with IANA.

The following section is based on the requirements defined in .

Name
JSON-based Serialization for Linked Data Structured Suffix
+suffix
+ld+json
References
A JSON-based Serialization for Linked Data
Encoding considerations
See RFC 8259, section 11.
Interoperability considerations
Not Applicable
Fragment identifier considerations

Fragment identifiers used with +ld+json are treated as in RDF syntaxes, as per RDF 1.1 Concepts and Abstract Syntax [[RDF11-CONCEPTS]].

Security considerations
See .
Contact
Pierre-Antoine Champin <pierre-antoine@w3.org>
Author/Change controller
W3C

Examples

The following examples illustrate different ways in which this structured suffix may be used to request and respond with more specific content types for a resource.

      
      
      
      

Open Issues

The following is a list of issues open at the time of publication.

Consider context by reference with metadata.

Compact IRI expansion support for non-trivial prefix term definitions.

Language-maps don't allow separate base direction.

`@default` in `@context` in JSON-LD core syntax.

Suggestion about `@prefix`.

Type Coercion / Node Conversion: @coerce keyword or similar.

Changes since 1.0 Recommendation of 16 January 2014

Additionally, see .

Changes since JSON-LD Community Group Final Report

Changes since Candidate Release of 12 December 2019

Changes since Proposed Recommendation Release of 7 May 2020

Changes since Recommendation of 16 July 2020