JSON-LD (JSON for Linking Data)

JSON-LD separates syntax from semantics through a @context mechanism. This maps JSON keys to globally unique IRIs (Internationalized Resource Identifiers), allowing data to be unambiguous and self-describing. For example, a name property can be expanded to http://schema.org/name, ensuring machines understand it refers to a person or organization's name, not a file name. This prevents key collisions and enables data integration across heterogeneous sources without prior coordination.

JSON-LD is a concrete syntax for the Resource Description Framework (RDF). Every JSON-LD document can be losslessly converted to RDF triples (subject-predicate-object statements) and back. This makes it a first-class citizen in the Semantic Web stack, compatible with SPARQL query engines and OWL reasoners. Developers can work in familiar JSON while the underlying data participates in a global graph of linked knowledge.

JSON-LD is the W3C-recommended format for structured data in web pages. It is injected into a <script type="application/ld+json"> tag, cleanly separating structured data from HTML markup. This contrasts with Microdata and RDFa, which interleave metadata with presentation tags. Major search engines like Google, Bing, and Yandex parse JSON-LD to power rich results, knowledge panels, and entity understanding.

JSON-LD provides Framing and Shaping APIs to control the structure of output documents. A JSON-LD Frame is a template that reshapes an RDF graph into a specific JSON tree layout, solving the 'graph-to-tree' problem. This allows developers to enforce a deterministic JSON structure for their applications, even when the underlying data model is a flexible graph. The SHACL standard is often used alongside JSON-LD for advanced data validation.

JSON-LD can be adopted incrementally. A plain JSON document is already valid JSON-LD if interpreted with a default context. Developers can start by adding a simple @context to existing APIs, making them semantically interoperable without breaking existing clients. This 'zero-edit' compatibility dramatically lowers the barrier to entry for Linked Data publishing compared to migrating to a pure RDF/XML or Turtle format.

Unlike traditional JSON which represents a strict tree hierarchy, JSON-LD natively supports a directed, labeled graph model. Using the @id keyword, any node can be referenced from multiple places, creating complex relational networks without data duplication. This is essential for representing interconnected entities like a Person who worksFor an Organization and knows other Persons, all within a single, coherent document.

The atomic unit of data in RDF and the underlying structure of JSON-LD. A triple consists of:

• Subject: The resource being described (an IRI or blank node).
• Predicate: The property or relationship (an IRI).
• Object: The value or another resource (an IRI, blank node, or literal). In JSON-LD, the @id keyword identifies subjects, property keys are predicates, and string values are literal objects. Understanding this decomposition is essential for debugging graph serialization.