Definition
A data structure used in GraphRAG to represent entities and their relationships as nodes and edges, both of which can store rich key-value metadata. It offers superior multi-hop reasoning and explainability for AI agents compared to flat vector stores, though it requires more complex schema management and higher computational overhead for traversal.
Unlike RDF-based Knowledge Graphs, Property Graphs allow internal attributes (properties) to be stored directly on the relationships (edges).
"A subway map where stations (nodes) have lists of facilities and the tracks (edges) between them specify the exact distance and current train speed."
- GraphRAG(Implementation Framework)
- Cypher(Query Language)
- Vector Database(Hybrid Complement)
- Knowledge Graph(Conceptual Superclass)
Conceptual Overview
A data structure used in GraphRAG to represent entities and their relationships as nodes and edges, both of which can store rich key-value metadata. It offers superior multi-hop reasoning and explainability for AI agents compared to flat vector stores, though it requires more complex schema management and higher computational overhead for traversal.
Disambiguation
Unlike RDF-based Knowledge Graphs, Property Graphs allow internal attributes (properties) to be stored directly on the relationships (edges).
Visual Analog
A subway map where stations (nodes) have lists of facilities and the tracks (edges) between them specify the exact distance and current train speed.