Knowledge Graph
The Knowledge Graph is an interactive force-directed visualization showing the relationships between people, topics, documents, and knowledge sessions.
Node Types
The graph contains four types of nodes, each with a distinct color and size:
| Node Type | Color | Size | Represents |
|---|---|---|---|
| Expert | Purple (#6366f1) | Large (24px) | Team members with expertise |
| Doc | Blue (#3b82f6) | Medium (16px) | Documents that contain knowledge |
| Session | Green (#10b981) | Medium (18px) | Recorded knowledge sessions |
| Topic | Amber (#f59e0b) | Small (14px) | Knowledge areas/subjects |
Nodes flagged as at risk (single point of failure) are highlighted in red (#ef4444) regardless of type.
Edge Types
Connections between nodes represent relationships:
- Solid lines — direct expertise or authorship (expert → topic, expert → doc)
- Dashed lines — reference relationships (doc references topic, session covers topic)
Interacting with the Graph
- Click a node to select it and see its connections in a detail panel below the graph
- The detail panel shows the node's label, type, risk status, and all connected nodes with relationship types
- Click a selected node again to deselect it
- The legend at the top shows node type colors and a count of at-risk nodes
Layout
The graph uses a spring-based force layout that runs 120 iterations to position nodes. Connected nodes attract each other, while all nodes repel to avoid overlap. The layout automatically adjusts to the container size.
Data Source
The graph data comes from goSvc.getKnowledgeGraph(), which returns nodes and edges built from:
- Knowledge profiles (who knows what)
- Documents authored by each person
- Recorded knowledge sessions
- Topic assignments (automatic and manual)
Access
Available under Knowledge > Graph tab. All authenticated users can view the graph. The at-risk indicators help managers identify knowledge concentration risks.