Concept mapping software, in context.
Concept mapping is the practice of representing ideas as nodes and the relationships between them as labeled edges. Modern concept mapping software comes in two flavors: drawing-first canvases and link-first knowledge bases. MindWiki is the link-first kind.
Two flavors of concept mapping software
- Canvas-first (Heptabase, Miro, Lucidchart, traditional CmapTools): you draw the nodes and edges by hand on an infinite canvas.
- Link-first (MindWiki, Obsidian, Roam, Logseq, Capacities): nodes are real pages, edges are wikilinks, and the graph is computed from how you write.
When the link-first kind is the better tool
- You want the concept map to grow as you write, not as a separate diagramming activity.
- Your concepts are long enough that they deserve real pages, not nodes with three lines of text.
- You want the "map" to stay accurate over years — drawn maps go stale; computed maps don't.
- You want AI clients reading the concept network.
How MindWiki implements link-first concept mapping
- Each concept is a markdown page. Frontmatter holds metadata (area, tags, type).
- [[wikilinks]] are the edges. Every page automatically tracks what links into it (backlinks panel).
- Graph view (macOS) renders the concept network — pan, zoom, filter by area or tag.
- API access: GET /vault/graph returns the network as JSON for external tools.
Frequently asked questions
Can I label the edges between concepts?
Indirectly. Wikilinks themselves don't carry edge labels, but the surrounding sentence does — "X is a cause of [[Y]]" — and the macOS graph view shows the linking sentence on hover.
Is this the same as a knowledge graph?
Effectively yes. "Concept map" and "knowledge graph" refer to the same shape — nodes and edges representing ideas and their relationships. Different fields prefer different names.
Can I export the concept map?
Yes — GET /vault/graph returns nodes and edges as JSON. Render it in whatever visualization tool you like.