Knowledge Graphs & Semantic Search

Structure is the missing half of retrieval

Entities, relations, and provenance — how knowledge graphs ground LLMs, power GraphRAG, and answer the questions vector search can't.

Intermediate4 modules · 12 lessons~119 min1160 XPAligned with Neo4j docs + GraphRAG researchStart with: NLP Foundations

Progress0/12

Module

Thinking in Triples

The atom of knowledge, the ontology contract, and queries that match shapes

Entities, Relations & Triples9 min · 90 XPEntities, Relations & Triples

Ontology Design10 min · 95 XPOntology Design

Querying the Graph10 min · 95 XPQuerying the Graph

Module

From Documents to Graph

Entity resolution, LLM extraction pipelines, and provenance that survives audits

Entity Linking & Resolution10 min · 100 XPEntity Linking & Resolution

Building KGs from Documents with LLMs11 min · 105 XPBuilding KGs from Documents with LLMs

Provenance & Grounding9 min · 90 XPProvenance & Grounding

Module

Graphs, Vectors & Search

Topology vs geometry — where each index wins and why production is hybrid

Graph Databases vs Vector Databases9 min · 90 XPGraph Databases vs Vector Databases

Graph Embeddings10 min · 100 XPGraph Embeddings

Semantic Search Needs Structure9 min · 90 XPSemantic Search Needs Structure

Module

GraphRAG & KG+LLM Systems

Structured + unstructured retrieval, coupling depths, and operating the graph as a product

GraphRAG: Retrieval with Structure11 min · 105 XPGraphRAG: Retrieval with Structure

KG + LLM Architectures11 min · 105 XPKG + LLM Architectures

Operating a Knowledge Graph10 min · 95 XPOperating a Knowledge Graph

Capstone project

Map a domain

Model a small domain you know well — a game, a hobby, your course schedule — as a knowledge graph: entities, relationships, and three queries it can answer.

Deliverable: A graph schema (diagram or code) plus the queries and their answers.

Milestones

Finish every lesson to earn the Knowledge Graphs & Semantic Search Certificate.