Knowledge Graph and Connections
How BrainTube discovers and visualizes connections between your content.
BrainTube builds a Knowledge Graph — a network of relationships between your saved content, entities, and concepts.
How Connections Form
Automatic Discovery
The auto-linker creates edges based on shared entities (2+ shared trigger a connection), semantic similarity (>85%), and co-mention patterns.
Relationship Types
The graph supports 22 edge types including related_to, builds_on, contradicts, reinforces, and synonym_of.
Exploring the Graph
Knowledge Map
Visit Knowledge Boxes to see a visual map of how your content connects.
Item Connections
On any item's detail page, scroll to "Related Content" to see connected items.
Graph-Enhanced Search
Search traverses the graph to discover contextually related items beyond keyword matches.
Graph Health
- Orphan rate — Goal: <25% items with no connections
- Edge density — Average connections per item
- Community detection — Clusters of densely connected items become Second Brains
Personalized PageRank
BrainTube uses Personalized PageRank (inspired by HippoRAG) to discover multi-hop connections — finding items connected through chains of relationships.