Knowledge map
The Knowledge Map is an interactive visual representation of your workspace's knowledge base. It shows documents, concepts, relationships, and connections as a dynamic graph, helping you understand how information is connected across your workspace.
Key Features
Interactive graph of workspace knowledge
Dynamic node and connection display
Accesing
- From Tools & Tab → Knowledge Map
- Or by accesing: `/workspace/{slug}/tools/knowledge-ma
Requirements
- Workspace must have embedded documents
- Documents need to be vectorized for connections to appear
- If no documents exist, you'll see a message with a link to workspace settings
Two view modes
- Intelligence: AI-extracted concepts, topics, timelines
- Structure: Documents, tags, structural relationships
What It Shows
- Topics: AI-extracted concepts connecting related documents
- Timeline Nodes: Time-based groupings of documents
- RAG Chunks: Vector chunks from documents
- Conversations: Chat threads linked to documents they referenced
- Tags: Document tags and metadata
Best For
- Understanding conceptual relationships
- Finding related content across documents
- Discovering patterns and themes
- Exploring temporal connections
Intelligence Mode
What It Shows
- Topics: AI-extracted concepts connecting related documents
- Timeline Nodes: Time-based groupings of documents
- RAG Chunks: Vector chunks from documents
- Conversations: Chat threads linked to documents they referenced
- Tags: Document tags and metadata
Best For
- Understanding conceptual relationships
- Finding related content across documents
- Discovering patterns and themes
- Exploring temporal connections
Structure Mode
What It Shows
- Documents: All workspace documents
- Tags: Document tags and metadata
- Structural relationships: Direct document connections
Best For
- Viewing document organization
- Understanding tag-based groupings
- Simpler, cleaner view
- Focused document exploration
Node Types
- Documents (PDF, DOC, DATA, WEB)
- Tags, Topics, Timeline nodes
- RAG Chunks, Chat conversations
Tag Nodes
Appearance:
- Light blue circles
Purpose:
- Shows document groupings by tags
- Helps organize content
- Can be toggled on/off
Topic Nodes
Appearance:
- Purple diamonds
- Larger size indicates more connections
Purpose:
- AI-extracted concepts
- Groups related documents
- Shows conceptual relationships
- Up to 20 topics displayed
Chat Nodes
Appearance
- Rose/pink circles
Purpose
- Conversation threads
- Linked to documents they referenced
- Shows prompt preview in details
Interactions
- Click to select and view details
- Drag to organize layout
- Zoom and pan to explore
- Search and filter to focus
Discovery Tools
- Search functionality
- Dimension filters
- Connection highlighting
- Details panel with metadata
Best Practices
Exploration
Start Broad
- Begin with full graph view
- Use Structure mode for overview
- Switch to Intelligence for insights
Focus Gradually
- Use search to find specific items
- Filter dimensions to reduce clutter
- Zoom in on interesting areas
Follow Connections
- Click nodes to see relationships
- Explore connected nodes
- Discover hidden connections
Organization
Use Filters
- Hide irrelevant node types
- Focus on specific document types
- Toggle tags/topics as needed
Custom Layouts
- Drag nodes to organize
- Create visual groupings
- Fix important nodes in place
Search Effectively
- Use specific terms
- Search for concepts or topics
- Filter before searching
Understanding Relationships
Connection Colors
- Purple: Semantic relationships (topics)
- Orange: Temporal relationships (time)
- Beige/White: Structural relationships (tags, direct)
Node Sizes
- Larger nodes = more connections
- Small nodes = fewer connections
- Hubs indicate important content
Connection Patterns
- Clusters show related content
- Isolated nodes = unique content
- Dense areas = highly connected topics
Tips and Tricks
Performance
- Large workspaces may take time to load
- Processing indicator shows during mode switches
- Graph stabilizes automatically
- Dragging nodes may cause temporary movement
Navigation
- Use zoom to see details
- Pan to explore different areas
- Reset by deselecting nodes
- Use search to jump to specific content
Discovery
- Switch between modes to see different perspectives
- Explore topic nodes to find related documents
- Check timeline nodes for temporal patterns
- Review chat nodes to see document usage
Troubleshooting
Graph Not Loading
Check
- Workspace has embedded documents
- Documents are vectorized
- Sufficient data available
Solutions
- Add documents to workspace
- Ensure documents are processed
- Check workspace settings
Nodes Not Appearing
Check
- Dimension filters are active
- Search query isn't hiding nodes
- View mode supports node type
Solutions
- Enable relevant dimension filters
- Clear search query
- Switch view mode if needed
Performance Issues
For Large Workspaces
- Use filters to reduce nodes
- Focus on specific areas
- Use search to narrow view
- Allow graph to stabilize
Summary
Use the Knowledge Map to explore, understand, and discover relationships within your workspace's knowledge base.
The Knowledge Map provides:
Visualization
- Interactive graph of workspace knowledge
- Dynamic node and connection display
- Real-time updates and filtering
Two View Modes
- Intelligence: AI-extracted concepts, topics, timelines
- Structure: Documents, tags, structural relationships
Node Types
- Documents (PDF, DOC, DATA, WEB)
- Tags, Topics, Timeline nodes
- RAG Chunks, Chat conversations
Interactions
- Click to select and view details
- Drag to organize layout
- Zoom and pan to explore
- Search and filter to focus
Discovery Tools
- Search functionality
- Dimension filters
- Connection highlighting
- Details panel with metadata