Knowledge Graphs

A knowledge graph represents information as entities and the relationships between them. CocoIndex builds a knowledge graph by extracting relationships from documents and writing them to a graph database.

LLM extraction turns unstructured text into structured data. CocoIndex uses an LLM to read a document and produce a summary together with a set of (subject, predicate, object) triples about the concepts it covers.

A triple becomes a relationship in the graph. The subject and object become entities, and the predicate labels the edge between them. CocoIndex also records which document mentioned which entity.

Neo4j stores the resulting property graph. CocoIndex declares Document nodes, Entity nodes, and the edges between them, and incremental processing keeps the knowledge graph up to date as the source documents change.