Learning Graph

The ikigize Learning Graph is a living knowledge ontology — a continuously growing map of Topics and Skills that connects every entity on the platform into a shared intelligence layer.

What Is the Learning Graph?

Rather than treating a course, a campus, and a learner profile as isolated records, ikigize connects them all through a shared graph of knowledge. When a course teaches JavaScript, a resource covers functional programming, and a learner is interested in web development, these facts live in the same graph — and the platform can reason across all of them at once.

The graph is not static. It grows as content is added, as learners engage, and as the AI cataloging pipeline maps new entities to existing knowledge. Like the concept of ikigai itself — a convergence of what you know, what you can do, and where you want to go — the Learning Graph continuously evolves to reflect the living state of knowledge across the platform.

Core Capabilities

Knowledge Exploration
Navigate the full topology of Topics and Skills with an interactive graph
Browse root domains and drill into subtopics
Search full-text across all Topics and Skills
Explore related topics via lateral relationships
"Surprise Me" serendipitous discovery of unmapped territories
Multiple layout modes (layered, stress, force-directed)
Entity Mapping
Connect every platform entity to the knowledge graph
Map courses, modules, sessions, and tasks to Topics they teach
Express campus and organisation specialisations via FOCUSES_ON
Track learner interests with INTERESTED_IN on profiles
Manual mapping by educators and admins
AI-assisted cataloging via the Librarian agent
Public Resource Matching
Surface globally available resources that teach the same topics
Graph-augmented search merges text and topic signals
Multi-path recommendation scoring across topic hierarchy
Subtopic and sibling traversal for broad discovery
Automatic fallback to domain-level resources
Resources saved to a library auto-link via resource_save
Living & Growing
The graph continuously expands as the platform learns
Bootstrapped from the ESCO EU taxonomy (16,000+ nodes)
AI cataloging adds new nodes during entity mapping
Every saved resource enriches the ontology
Manual curation keeps the graph accurate
Multilingual — all nodes available in 6 languages

Explore in Detail

  • Ontology — Topics and Skills, internal relationships, and how the knowledge taxonomy is structured
  • Entity Connections — how courses, campuses, resources, and profiles connect to the graph via DEVELOPS, FOCUSES_ON, and INTERESTED_IN
  • Signals & Sources — where mappings come from (manual, catalog, activity), strength, and behavioral edges
  • Recommendations — how resource matching and recommendation scoring work
  • Exploration — the interactive graph explorer, domain selection, search, and Discover mode
  • AI Cataloging — how the Librarian agent maps entities to the graph

See Also

  • Library — graph-augmented search and recommendations in practice
  • System Overview — where the Learning Graph fits in the platform
  • Librarian Agent — the AI that catalogues entities via the graph