Skip to content

Learning Graph Viewer

Open Learning Graph Viewer Fullscreen (recommended)

What This Viewer Shows

The picture above is the learning graph for Information Systems — all 580 concepts in the book and the prerequisite relationships between them, drawn as one connected diagram you can pan, zoom, search, and filter. Before you start dragging things around, it's worth knowing what each piece of the graph actually means.

What Is a Learning Graph?

A learning graph is a structured map of everything a course teaches and the order in which a student needs to learn it. Formally, it has two ingredients:

  1. Nodes — one node per concept (a single discrete idea, term, or skill the book wants you to learn).
  2. Edges — directed arrows from one concept to another that record "you have to understand A before B will make sense."

You can think of it as a wiring diagram for a curriculum. The chapters are the surface; the learning graph is the circuitry underneath.

What Is a Concept?

A concept is the smallest unit of knowledge the book treats as worth naming. In this book, concepts include things like:

  • Primary key
  • Normalization
  • Retrieval-Augmented Generation (RAG)
  • Information Technology Infrastructure Library (ITIL)
  • Prompt injection
  • Enterprise Knowledge Graph

Every concept has a precise definition in the Glossary, appears in at least one chapter, and shows up exactly once as a node in the learning graph. If you can't define a concept in one or two sentences without using the concept inside its own definition, it's either not a concept yet or it needs to be split into smaller ones.

What Are Concept Dependencies?

A concept dependency is the rule that says "to understand A, you must first understand B." In the learning graph, it shows up as an arrow:

A ─────► B

In this book, that arrow is read as "A depends on B" — the source of the arrow is the concept you're trying to learn, and the target is its prerequisite. (Said another way: the arrow points from the concept toward the prerequisite it needs.) For example, foreign key depends on primary key, which depends on table, which depends on row and column, which depend on relational model. You can't responsibly teach foreign keys to a student who has never seen a row.

A single concept usually has several prerequisites. GraphRAG, for example, depends on retrieval-augmented generation, knowledge graph, embedding, vector database, and a handful of others. The learning graph captures all of those dependencies so we can reason about them mechanically instead of guessing.

What Is Concept Enablement?

Concept enablement is the inverse of dependency. If concept A is a prerequisite for B, then learning A enables you to learn B. The same arrow tells both stories — it just depends on which direction you're reading it:

  • Following the arrow (toward prerequisites): "What do I need to know first?" → that's dependency.
  • Following the arrow backward (against the arrow): "What does knowing this unlock?" → that's enablement.

Enablement is the more motivating direction. It answers the student's question, "Why am I learning this?" — because it's the gateway to something else they want.

What Is a DAG?

The learning graph is a DAG, short for Directed Acyclic Graph. That mouthful breaks down into three properties:

  • Graph — a collection of nodes connected by edges.
  • Directed — every edge has a direction (the arrow), so the relationship is one-way.
  • Acyclic — there are no cycles. You cannot follow the arrows from any concept and end up back where you started.

The acyclic part matters more than it sounds. If concept A depends on B, and B depends on C, and C depends on A, then nothing in that loop can ever be the first thing you learn — and so none of the three can be learned at all. A cycle in the learning graph is a bug, full stop. Whenever this book's graph generator detects one, it refuses to ship until the cycle is broken.

What Is Learning Order?

A learning order is any ordering of the concepts in which every concept appears after all of its prerequisites. Mathematicians call this a topological sort of the DAG.

Because the graph is a DAG, at least one valid learning order always exists. Usually many do — for example, if normalization and indexing don't depend on each other, the graph doesn't care which one you teach first. The chapter outline you read in the book is one valid learning order; a personalized course that skipped concepts the student already knew would be another.

The chapter sequence in this book was generated by topologically sorting the learning graph and then grouping nearby concepts into cohesive chapters and parts. That's why each chapter feels like it has "earned" the next: every concept it introduces has all of its prerequisites already on the board.

Why This Matters Now — and Why It Will Matter More

Even at this book's current level, the learning graph is genuinely useful:

  • Author hygiene. It lets us prove that no chapter introduces a concept before its prerequisites exist somewhere earlier in the book.
  • Student navigation. You can search for any concept, see its upstream prerequisites, and jump to the chapter that introduces them if you hit a wall.
  • Curriculum mapping. Programs adopting the book for ABET accreditation can map specific student outcomes to specific subgraphs.

The learning graph becomes much more powerful in Level 3 (L3) intelligent textbooks — the next step in our textbook taxonomy. An L3 book treats the learning graph as a live structure the system reasons over, not just a static reference. In L3:

  • The book knows what each student has and hasn't mastered and shows them the next concept whose prerequisites they have already cleared.
  • Adaptive review kicks in automatically: if a quiz shows the student is shaky on normalization, the system surfaces every downstream concept that depends on it and offers a refresh before moving on.
  • Personalized learning paths become possible — a student who arrives already knowing relational databases gets a different topological sort than one starting from scratch, but both end at the same set of outcomes.
  • AI tutors use the graph as a grounding source so they never invent prerequisites or skip them, and so they can explain why a concept is being introduced now.

What you're looking at on this page is the foundation those features will be built on. Every concept node, every dependency arrow, and every category tag is a hook that an L3 system will eventually grab onto.

Features of This Viewer

  • Zoom and Pan — use your mouse and scroll wheel to navigate.
  • Search — type a concept name to find and focus on it.
  • Category Filtering — checkboxes in the sidebar show or hide groups of related concepts.
  • Interactive Navigation — drag to pan, scroll to zoom, click a node to highlight its connections.
  • Statistics — real-time counts of visible nodes and edges.

Suggested Overview Steps

  1. Uncheck all the categories on the left side.
  2. Check only the Foundations category.
  3. Zoom in on the cluster you see — these are the concepts with no prerequisites and are the entry points to the book.
  4. Re-enable one category at a time and watch the graph grow outward. This is, more or less, the order in which the book teaches.

Using the Viewer

  1. Search for Concepts — start typing in the search box. Click a result to focus on that node.
  2. Filter by Category — use the category checkboxes in the sidebar. "Check All" / "Uncheck All" do bulk operations.
  3. Navigate the Graph — drag to pan, scroll to zoom, click a node to select it and highlight its dependencies and downstream concepts.
  4. View Statistics — the sidebar shows counts of visible nodes, edges, and foundational concepts.

Graph Structure at a Glance

  • Foundational concepts (left side) — prerequisites with no incoming dependencies. These are where any valid learning order has to start.
  • Advanced concepts (right side) — topics that require many prerequisites. The deeper into the right side you go, the more of the book sits "underneath" that concept.
  • Edges — arrows point from a concept to its prerequisites. Read them as "needs," not "leads to."