Recommended Free Tools
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
A knowledge graph represents things and the relationships between them as connected data. In a typical graph, nodes stand for things, and labeled, directed edges state how those things relate. RDF is one important standards-based way to build such graphs, but it is not a requirement for every system called a knowledge graph.
This glossary separates the general idea from the terms used in RDF and related W3C standards, then explains how Google’s branded Knowledge Graph and API fit into the picture.
Start with the basic graph terms
Knowledge graph
A knowledge graph is a graph-structured representation of entities or other resources and the relationships between them. The graph makes connections explicit: rather than keeping facts only in isolated records, it links things with named relationships that can be interpreted in context.
The phrase covers different systems and designs. RDF, OWL, and SPARQL are useful standards to know, but not every knowledge graph uses them or follows the same database architecture.
#1 Best Overall
Node, resource, and entity
In an informal graph description, a node is an element or position in the graph. RDF uses the term resource for what statements describe or refer to. A resource can be a physical thing, a document, an abstract concept, a number, or a string; in the RDF Semantics context described by W3C, “entity” is used synonymously with resource. These words overlap in everyday explanations, but “node” is the broad graph term and “resource” is the RDF term. W3C’s RDF 1.2 Concepts and Abstract Data Model identifies itself as a Candidate Recommendation Snapshot, rather than a final Recommendation.
IRI
An Internationalized Resource Identifier (IRI) is an RDF term that can denote a resource. An IRI is an identifier, not a guarantee that a browser will open a useful web page when someone visits it.
Literal
A literal is a value term, such as a string or a typed value. RDF distinguishes literals from IRIs and blank nodes, even if their displayed text looks the same. For example, the text “Paris” as a literal value is not automatically the same thing as an identified resource representing the city of Paris.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →How an RDF statement works
Triple
An RDF triple is an ordered statement with a subject, predicate, and object. The subject is what the statement describes; the predicate names a property or relationship; and the object is the related resource or value. W3C describes RDF graphs as sets of subject–predicate–object triples, whose elements may include IRIs, blank nodes, datatyped literals, or triple terms. RDF 1.2 Concepts and Abstract Data Model
For example, a graph might contain a statement equivalent to “Mina knows Jo.” Mina is the subject, “knows” is the predicate, and Jo is the object. This is a directed statement: it does not, by itself, say that Jo knows Mina.
Predicate, property, and relationship
The predicate is the part of an RDF triple that identifies its property or relationship. “Property” and “relationship” are common explanatory terms for what a predicate expresses. Direction matters: a statement from one subject to one object does not automatically establish the reverse statement.
RDF graph
An RDF graph is a set of RDF triples used to describe resources. The set-based model means the graph is defined by its statements, rather than by a required visual layout or a particular database product.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsRDF dataset and named graph
An RDF dataset consists of one default graph and zero or more named graphs. A named graph is a graph paired with an IRI or blank-node name within a dataset. SPARQL can query named graphs. These are RDF dataset concepts, not features that every system described as a knowledge graph must provide. W3C RDF 1.2 Concepts and the W3C-hosted SPARQL 1.2 Query Language specification.
RDF, OWL, and SPARQL have different jobs
| Term | Role | What it helps answer |
|---|---|---|
| RDF | A framework and data model for representing information as graphs of triples. | How are statements and graph data represented? |
| OWL | A family of knowledge-representation and vocabulary-description languages for authoring ontologies. | How can domain concepts and their relationships be described formally? |
| SPARQL | A query language for RDF data. | How can matching information be retrieved from an RDF dataset? |
W3C’s Linked Data Glossary compares SPARQL’s role to SQL’s role for relational databases: both are query languages, but SPARQL is for RDF data. OWL is based on RDF and standardized by W3C; it is not another name for the RDF data model. The SPARQL 1.2 specification is hosted as a live draft and may change.
Triple store
A triple store is a colloquial name for an RDF database that stores triples. The term points to a storage system, not a query language or ontology language.
Ontology, taxonomy, vocabulary, and schema
Ontology
An ontology formally describes concepts and relationships in a domain. OWL is a W3C-standardized family of languages for authoring ontologies, with description-logic and RDF-based semantics. “Ontology” often implies a more formal account of a domain than a simple list of labels, though actual usage varies.
Taxonomy
A taxonomy arranges items in a hierarchy. It can be part of an ontology, but it is not a synonym for every ontology: an ontology may describe relationships beyond a hierarchy.
Vocabulary
A vocabulary is a collection of terms assembled for a purpose. In Linked Data practice, “vocabulary” and “ontology” can overlap; W3C notes this overlap in its Linked Data Glossary. When precision matters, check how a project defines the term rather than assuming a single universal distinction.
Schema
Schema has no single meaning across all graph systems. For RDF, an RDF vocabulary or schema describes terms used in data; an ontology language such as OWL provides a way to express formal descriptions and semantics. A schema is therefore not automatically an ontology, and “schema” in one product may mean something different from its use in another data model.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Choosing terms and approaches without conflating them
These labels describe different layers, not competing names for the same thing. When evaluating a graph system or reading its documentation, compare the concrete capabilities that matter:
Free tools Windows power users keep installed
One-click scans. No signup required.
- Data model and serialization: Is information represented as RDF triples, or does the system use another model? How is the data encoded for exchange?
- Schema and ontology expressiveness: What types, properties, constraints, or formal relationships can be described?
- Query and reasoning: What query language is supported, and what reasoning capabilities are available?
- Identifiers and integration: How are resources identified, and how can data from different sources be connected?
- Provenance, validation, and operations: How are sources and data quality handled, and what does the system require to deploy and maintain?
RDF, OWL, and SPARQL help name some of these dimensions, but the terms alone do not establish a particular product’s performance, operational model, or support for validation and reasoning.
Google’s Knowledge Graph is a specific example
Google’s Knowledge Graph is Google’s entity-based collection of known things; it is one branded example, not the generic definition of a knowledge graph. Google’s Knowledge Graph Search API searches entities using schema.org types and JSON-LD. Google documents uses including ranked entity matching, autocomplete, and content annotation. The API documentation describes it as read-only and cautions against relying on it for production-critical use. Those product-specific details should not be assumed to describe other knowledge graphs.
Google’s Search Console glossary also uses “Knowledge Graph” in the context of Google Search. That search feature and the developer API are related to Google’s branded entity system, but neither makes Google’s terminology a universal technical standard.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
