Knowledge Graph

Executive Definition: Knowledge Graph — A structured representation of data that models real-world entities and the relationships between them using a graph topology.

A knowledge graph consists of nodes (representing entities like people, places, or concepts) and edges (representing the semantic relationships between those entities). By explicitly mapping how data points are connected, knowledge graphs enable complex querying and reasoning. They are commonly used in search engines, recommendation systems, and increasingly, as a factual foundation for large language models to mitigate hallucinations.

Frequently Asked Questions

How is a knowledge graph different from a relational database?

Relational databases use tables and foreign keys to link data, which can make complex relationship queries slow. Knowledge graphs natively store relationships as edges, allowing for highly efficient traversal of highly connected data.

What is GraphRAG?

GraphRAG is an architecture that retrieves context from a knowledge graph to augment large language models, providing more factual and relationship-aware responses than standard vector-based RAG.

What technologies are used to build knowledge graphs?

Knowledge graphs typically use graph databases like Neo4j, Amazon Neptune, or TigerGraph, and rely on query languages such as Cypher, SPARQL, or Gremlin.