Definition
A knowledge graph is a programmatic way to represent a network of real-world entities and their interrelations. It uses nodes (entities) and edges (relationships) to store data in a way that is easily queryable by both humans and machines.
Why it matters (in Poovi’s context)
Fundamental to the memex project and tools like graphify. It allows for “declarative memory” that an AI agent can query without loading every raw source file into context.
Key properties or components
- Nodes: Represent entities (e.g., people, tools, concepts).
- Edges: Represent relationships (e.g., “mentions”, “implements”, “works-on”).
- Provenance: Tracking the origin of a fact (e.g., graphify’s EXTRACTED vs INFERRED tags).