Sensemaking Semantic Web · Roadmap

Where this is, where it's going

An honest tracker for a self-directed curriculum being worked in public. Updated as the work moves.

← Back to the curriculum map
Right now

Module 1 is underway. All three submodule synthesis pages are published, along with five hands-on workbooks (Naruto and mythology variants for submodules 1.1 and 1.2, plus the 1.3 resume workbook), a reading companion for the Allemang and DuCharme chapters, and the Module 1 cheat sheet. The primary project — a Resume Graph RDF slice — is in progress. Modules 2 through 4 are sketched in the syllabus but not yet built.

The four modules

What's coming, in rough order

The full curriculum is twelve weeks of self-directed work, organized in four modules. Each module produces real artifacts — published ontologies, evaluation notebooks, deployed demos. By the end, the curriculum is the portfolio.

The default rhythm is five to seven hours per week, but the modules are independent units. Compression to eight weeks or expansion to twenty are both fine. Pauses between modules are expected.

01

Foundations

In progress

RDF, Turtle, basic SPARQL — the conceptual core of the stack.

  • Resume Graph Explorer slice in both Cypher and Turtle — side-by-side comparison on real data
  • Naruto Network Graph Turtle slice — one arc, converted to RDF
  • Blog post: "LPG vs RDF for resume graphs: a side-by-side"
Weeks 1-3 · ~5-7 hours/week · Light new ground
02

Modeling

Not yet started

RDFS, OWL, ontology design — where modeling judgment develops.

  • Naruto ontology v1.0 — published OWL ontology with documented design decisions, REUSE.md, and SHACL shapes
  • Blog post: "Designing an anime ontology in public: the small modeling decisions that compound"
  • Optional: career narrative ontology layer for Resume Graph Explorer
Weeks 4-6 · ~6-8 hours/week · Medium new ground
03

Reasoning at the edge

Not yet started

Inference, reification, SHACL — the conceptually hardest segment.

  • Reified Naruto knowledge graph with full provenance — handling contested fan canon honestly
  • Skill inference notebook on Resume Graph — formal reasoning vs. LLM reasoning, compared honestly
  • Blog post: "Where OWL reasoners beat (and lose to) LLM reasoning"
Weeks 7-9 · ~7-9 hours/week · Heavy new ground · Most likely to slip
04

Shipping

Not yet started

SPARQL UPDATE, deployment, LLM + knowledge graph integration.

  • TwinKit v2.0 — RAG framework with optional semantic layer, released on GitHub
  • Naruto KG Explorer — publicly deployed demo combining vector and graph retrieval
  • Case study: "Building a hybrid semantic + vector knowledge twin (with Naruto)"
Weeks 10-12 · ~8-10 hours/week · Mixed new ground · The capstone

A note on honesty

When this might pause

The curriculum is voluntary, self-directed, and worked alongside consulting practice. That means a few conditions could legitimately stop or delay it, and naming them up front is more honest than pretending the path is linear.

None of these are failure modes. They are "this isn't the right vehicle right now" modes. If the curriculum pauses, the pause and its reason will be noted here.


Follow along

If you want updates

Two low-friction ways to follow what's shipping, without anything resembling a newsletter or email capture.

Questions, corrections, and respectful disagreement about the curriculum design are welcome via GitHub Issues. This is a personal learning project — pull requests on the curriculum itself aren't accepted, but feedback shapes how it evolves.