Understanding the Knowledge Graph
Your notes become a network of connected intelligence. See how ideas relate, discover hidden patterns, and navigate your knowledge visually.
Graph Structure
The knowledge graph represents your information as an interconnected network of nodes and edges.
Nodes
The elements in your knowledge graph:
- •Notes - Your individual pieces of content
- •Entities - Extracted people, organizations, concepts, etc.
Edges
Connections between nodes with:
- •Type - What kind of relationship exists
- •Strength - A value from 0 to 1 indicating connection strength
Connection Types
Lexic discovers and creates different types of connections between your content:
Notes with similar meaning, discovered through AI embeddings. Even if notes use different words, semantic connections find conceptual overlap.
Notes that share extracted entities. If two notes mention the same person, organization, or concept, they become connected.
Explicit references between notes. When you link one note to another, a reference connection is created.
Time-based relationships. Notes created around the same time or referencing similar dates are connected temporally.
User-created connections. You can manually connect any two notes when you see a relationship the AI might have missed.
Relationships between entities themselves. These describe how entities relate to each other across your entire knowledge base.
Entity Relationships
Entities extracted from your notes can have their own relationships, building a rich network of connections.
Relationship Types
Cross-Note Entity Resolution
When the same entity (like a person or company) appears in multiple notes, Lexic automatically recognizes and merges them. This means clicking on "Sarah Chen" shows you every note where she's mentioned, even if some notes use "Sarah" and others use "S. Chen".
Visualization Features
The graph viewer provides multiple ways to explore and filter your knowledge network.
View Modes
- All - See notes and entities together
- Notes Only - Focus on your content
- Entities Only - See the extracted network
Filtering Options
- Depth - 1, 2, or 3 levels from starting node
- Strength - Filter weak connections
- Entity Type - Show specific categories
Visual Legend
Each entity type has a unique color and shape. Shapes provide accessibility for colorblind users.
| Entity Type | Color | Shape |
|---|---|---|
Person | #3B82F6 | Circle |
Organization | #10B981 | Square |
Location | #F59E0B | Triangle |
Concept | #8B5CF6 | Hexagon |
Technology | #06B6D4 | Diamond |
Date | #EF4444 | Pentagon |
Money | #14B8A6 | Octagon |
Event | #EC4899 | Star |
* Shapes provide accessibility for colorblind users, ensuring entity types remain distinguishable regardless of color perception.
Graph Traversal
Navigating the graph uses words based on depth. Deeper exploration reveals more connections but costs more.
Depth Costs
5
words
Depth 1
Direct connections
15
words
Depth 2
Connections of connections
50
words
Depth 3 (Max)
Extended network
Traversal Strategies
- BFSBreadth-First Search - Explores all connections at each level before going deeper. Best for finding all nearby connections.
- DFSDepth-First Search - Follows connection chains as far as possible before backtracking. Best for finding distant but direct paths.
- Best-FirstBest-First Search - Prioritizes strongest connections. Best for finding the most relevant related content quickly.
Budget Mode Tip: When using Budget Mode, you can set a maximum word limit for graph traversal. The system will automatically stop exploring when the budget is reached.
How Do My Notes Connect?
When you save a note, Lexic automatically processes it to find connections:
- Entity extraction - Identifies people, places, organizations, and concepts
- Semantic embedding - Creates a mathematical representation of meaning
- Connection discovery - Compares against all other notes in your lexicon
- Graph update - Adds new nodes and edges to your knowledge graph
This happens automatically in the background. You write naturally; Lexic handles the rest.