Iterative Graph Expansion (MERGE)

Iterative Graph Expansion is a strategy for building knowledge graphs by gradually adding new nodes and relationships using the MERGE operation in Neo4j, rather than trying to build a complete graph from scratch. MERGE creates a node or relationship only if it doesn't exist — if it already exists, it updates it, preventing duplicates. This approach enables controlled, incremental graph development. The process begins by adding the central entity with basic attributes, then expands with sub-queries from PAA, competitor analysis data, and content gaps.

Each iteration enriches the graph without risking damage to the existing structure. Iterative graph expansion through MERGE follows the 'start small, iterate fast' principle — prioritizing accuracy over size ensures better graph quality.

For example, in the first iteration a law firm's graph has 1 entity and 10 attributes; after three iterations it contains 15 entities and 80 attributes with their relationships.

Best practice involves conducting the helicopter view after each iteration to ensure new nodes connect logically to the existing structure.