Appearance
Creating a Taxonomy with Termboard
Taxonomies are fundamental for organizing knowledge. They classify concepts into a hierarchical structure, typically using "is-a" relationships (e.g., a "Car" is a "Vehicle"). Termboard makes it easy to create, manage, and visualize these hierarchies.
What is a Taxonomy?
A Taxonomy is a scheme of classification, especially a hierarchical classification, in which things are organized into groups or types.
In Termboard, a taxonomy is built using:
- Terms (representing concepts or categories).
- Relations (specifically "Generalization" relations) that link a child term to a parent term.
1. Setting up the Relation Type
To create a proper taxonomy in Termboard, you need a relation type that represents the hierarchy.
- Map the relation names you want to use to the
Generalizationtype, as described here - Ensure there is a relation type (usually named
Generalizationoris a) where the Type is set toGeneralization.- This specific setting tells Termboard to treat this relation as a hierarchical link.
- It enables special features like Relation Inheritance and the Hierarchy View.
2. Creating the Hierarchy
You can build your taxonomy manually or use AI to generate it.
Manual Creation
- Create Terms: Add terms for your categories (e.g., "Animal", "Mammal", "Dog").
- Link them: Draw a relation from the specific concept (Child) to the general concept (Parent).
- Example: Draw a line from "Dog" to "Mammal".
- Set Relation Type: Give it (one of the) name you mapped to
Generalization. It is handy to setColor Relation typesin . The color of the relation arrow should now automatically be set to blue.
Using the Hierarchy View
For large taxonomies, the Hierarchy View is the most efficient tool.
- Open the Hierarchy View tab in the right sidebar.
- This view displays your terms in a tree structure based on the Generalization relations.
- Drag and Drop: You can often drag terms within this view to re-parent them (depending on your specific Termboard version's capabilities).
- Navigation: Click on any item in the tree to locate it on the canvas.
3. AI-Assisted Taxonomy Generation
From within Termboard
Termboard's AI capabilities can jumpstart your taxonomy.
- Go to .
- Select the prompt "Taxonomy for selected terms" (if you have terms selected)
- The AI will generate the terms and the necessary
Generalizationrelations automatically.
You can also easily copy paste taxonomies or import One way of doing this is
- Give the AI a role of taxonomist. Here is a Claude Skill file
- Ask it to generate first the taxonmoy with terms and relations: source|is a|target
- This you can paste in the bulk add relation field
- (optional) Improve you taxonomy in Termboard, then export it in excel for example
- Give the taxonomy to the AI and ask it to generate definitions in the format:
an X is an Y that Zto make a complete and consice definition. It should give as output: name|definition - Paste this in the bulk add sidebar term field
From outside Termboard
If you want more control, you can use the bulk add sidebar to add terms and relations. See here.
Here is an example prompt you can use to get the right format:
Create an elaborate, fully connected, and deep taxonomy about [TOPIC].
Write it to [FILE_PATH].
Format each line as: source|relation name|target
# Core Requirements
1. **Single Ultimate Root:** Define exactly one top-level concept (e.g., "[TOPIC] Concept") that serves as the root.
2. **Strict "Is A" Hierarchy:** Focus **exclusively** on `is a` (hierarchical) relationships for this pass. Do not include complex relations like `is part of`, `measures`, or `causes` yet. We are building the classification backbone.
3. **Full Connectivity:** Every single concept must trace a path back to the Ultimate Root. No orphan concepts.
# Structural Constraints (Crucial)
4. **Max Children Rule:** **NO parent concept should have more than ~25 direct children.**
* *Bad:* "Asset" has 50 types of assets directly linked to it.
* *Good:* "Asset" -> "Current Asset" (has 10 children) & "Non-Current Asset" (has 10 children).
* *Action:* If a category gets too broad, you **MUST** invent logical intermediate groupings (e.g., instead of listing 50 "Risks", group them into "Financial Risk", "Operational Risk", "Strategic Risk").
5. **Proper Layering:** Avoid flat lists. Use meaningful grouping categories to ensure depth.4. Visualizing and Managing
Tree Layout
When working with taxonomies, the layout of your graph matters.
- Use the Auto Arrange feature.
- Select Hierarchical layout algorithm to visually arrange your terms in a clear top-down or left-right structure.
- If there are many terms, you can also try the
widerlayout.
Semantic Checks
Use Semantic Checks to validate your taxonomy:
- Circular Hierarchy: Ensures you don't have loops (A is a B, B is a A).
- Multiple Parents: While allowed in some taxonomies (poly-hierarchy), often you want a strict tree. Termboard can warn you about this.
5. Enriching with Hierarchy Data
You can automatically add useful metadata to your taxonomy using Extra Fields. Two specific field types are designed for hierarchies:
Hierarchy-Level
The hierarchy-level field type automatically calculates the depth of each term. You can use this to:
- Assign labels like "Domain" (depth 0), "Category" (depth 1), "Subcategory" (depth 2), etc.
- Color-code terms based on their level to visually distinguish the hierarchy structure.
Hierarchy-Branch
The hierarchy-branch field type identifies which main branch a term belongs to.
- This is perfect for "coloring by sub-tree". For example, all "Mammals" and their descendants can be one color, while "Birds" and their descendants are another.
- It simplifies visual analysis of large taxonomies by grouping related concepts.
6. Exporting
You can export your taxonomy in standard formats:
- OWL (Web Ontology Language): Your generalization relations will be exported as
rdfs:subClassOf. - JSON: For custom processing.
See Import/Export for details.