Feature
Knowledge Graphs
See how concepts, entities, and ideas connect across your documents.
What Is a Knowledge Graph?
A knowledge graph is a way to organize information as a network of connected things.
Think of it like a map of ideas. Instead of reading documents from start to finish, you can see how different concepts relate to each other — who works with whom, what causes what, which ideas build on others.
A Simple Example
Imagine you upload a research paper about climate change. A knowledge graph might show:
- Entities: “Carbon dioxide”, “Global temperature”, “Paris Agreement”, “IPCC”
- Relationships: “Carbon dioxide → increases → Global temperature”, “Paris Agreement → addresses → Climate change”
Now you can see at a glance what the paper is really about and how its concepts connect.
Why This Matters
Traditional search finds documents that contain your keywords. Knowledge graphs go further:
| Traditional Search | Knowledge Graph |
|---|---|
| Finds documents with “carbon” | Shows what carbon dioxide connects to |
| Returns a list of matches | Reveals hidden relationships |
| You read to understand | You see the structure instantly |
What Onsomble Creates
When you add sources to a notebook, Onsomble automatically builds a knowledge graph by:
- Extracting entities — People, organizations, concepts, events, metrics
- Finding relationships — How entities connect to each other
- Tracking sources — Which documents mention each entity
The result is an interactive visualization you can explore.
When Knowledge Graphs Help
Research projects — See how papers, authors, and ideas connect across multiple sources.
Competitive analysis — Map companies, products, and market relationships.
Learning new topics — Understand how concepts build on each other.
Large document sets — Find connections you’d miss reading linearly.
Knowledge graphs become more powerful as you add more sources. Entities from different documents get linked together automatically.
Getting Started
Add Sources
Upload documents, add URLs, or import content to your notebook. Knowledge graphs are built from your sources automatically.
Open the Graph
Click Knowledge Graph in the Sources panel footer. A dialog opens showing your notebook’s graph.
Explore
Click nodes to see details. Filter by category. Search for specific entities. Change the layout to see different perspectives.
What’s in a Graph?
Nodes (Entities)
Each circle in the graph represents an entity — something extracted from your documents:
| Type | Examples |
|---|---|
| Entity | People, companies, products, locations |
| Concept | Ideas, theories, methodologies |
| Event | Meetings, launches, incidents |
| Process | Workflows, procedures, algorithms |
| Metric | Measurements, KPIs, statistics |
Edges (Relationships)
Lines connecting nodes show how entities relate:
| Type | Examples |
|---|---|
| Hierarchy | ”owns”, “contains”, “manages” |
| Causal | ”causes”, “prevents”, “enables” |
| Temporal | ”before”, “after”, “during” |
| Association | ”uses”, “implements”, “extends” |
Properties
Each entity includes additional information:
- Description — What this entity is
- Sources — Which documents mention it
- Mentions — How often it appears (affects node size)