Reference

Knowledge Graph Glossary

The key terms you need to know — from nodes and edges to RAG, SPARQL, and MCP. Plain definitions, no jargon.

Knowledge Graph

A structured map of entities and the relationships between them. Unlike a database that stores records, a knowledge graph stores connections as first-class citizens.

Node

An entity in a knowledge graph — a person, concept, document, event, or any thing that has an identity and can be related to other nodes.

Edge

A typed relationship between two nodes in a knowledge graph. Examples: "depends on," "contradicts," "builds on," "relates to."

Ontology

The schema that defines what types of entities exist in a knowledge graph and what types of relationships are allowed between them.

RDF

Resource Description Framework — a W3C standard for representing information about resources in the form of subject-predicate-object triples.

Triple

The fundamental unit of RDF data: subject-predicate-object. For example: "Server-A runs-on Port-8080."

SPARQL

A query language for RDF knowledge graphs, analogous to SQL for relational databases.

Vector Database

A database that stores embeddings — mathematical representations of meaning. Used for similarity search and semantic retrieval.

Embedding

A numerical vector representation of text, images, or other data that captures semantic meaning. Similar items have similar vectors.

RAG (Retrieval-Augmented Generation)

A technique that retrieves relevant document chunks from a vector database and injects them into an LLM prompt to improve answer quality.

Graph RAG

An emerging approach that combines knowledge graphs with RAG — using the graph's structure to retrieve more relevant context for LLMs.

LLM (Large Language Model)

A neural network model trained on vast text corpora that can generate, summarize, and transform text. Examples: GPT-4, Claude, Gemini.

Hallucination

When an LLM generates confident-sounding but factually incorrect information — often because it lacks grounded, structured context.

Provenance

The record of where a piece of information came from. In a knowledge graph, every node and edge traces back to its source document and section.

Contradiction Detection

The process of identifying when two or more sources make incompatible claims about the same entity or relationship.

Dependency Mapping

The process of identifying prerequisite relationships — what must be understood before something else can make sense.

Isomorphism

A structure-preserving mapping between two systems. In knowledge graphs, isomorphic structures have the same relational shape even if the content differs.

Cross-Domain Mapping

The application of structural patterns from one domain to another. For example, mapping the structure of a software architecture onto the structure of an organizational chart.

MCP (Model Context Protocol)

An open protocol that allows AI models to access external tools and data sources — including knowledge graphs — in a standardized way.

Semantic Search

Search that understands meaning rather than just keywords. Vector databases enable semantic search through embeddings.

Ready to build your knowledge graph?

SILKLEARN reads your documents and builds the graph automatically. No manual ontology design needed.