Skip to content

knowledge-graph.md

Latest commit

 

History

History
126 lines (89 loc) · 2.83 KB

knowledge-graph.md

File metadata and controls

126 lines (89 loc) · 2.83 KB

Building a Knowledge Graph

Build a knowledge graph from scratch using Neumann's graph engine. By the end, you will have nodes representing people and topics, edges showing relationships, and queries that traverse the graph.

Prerequisites

Step 1: Start the Shell

neumann --wal-dir ./knowledge-graph-data

You should see the neumann> prompt.

Step 2: Create Nodes

Create nodes for people and topics:

NODE CREATE person { name: 'Alice', role: 'engineer' }
NODE CREATE person { name: 'Bob', role: 'designer' }
NODE CREATE person { name: 'Carol', role: 'manager' }
NODE CREATE topic { name: 'machine-learning' }
NODE CREATE topic { name: 'user-experience' }
NODE CREATE topic { name: 'project-planning' }

Verify the nodes exist:

NODE LIST

You should see 6 nodes with their auto-assigned IDs.

Step 3: Create Edges

Connect people to their expertise topics:

EDGE CREATE 'node:1' -> 'node:4' : expert_in
EDGE CREATE 'node:1' -> 'node:5' : interested_in
EDGE CREATE 'node:2' -> 'node:5' : expert_in
EDGE CREATE 'node:3' -> 'node:6' : expert_in
EDGE CREATE 'node:3' -> 'node:4' : interested_in

Connect people to each other:

EDGE CREATE 'node:1' -> 'node:2' : collaborates_with
EDGE CREATE 'node:2' -> 'node:3' : reports_to
EDGE CREATE 'node:1' -> 'node:3' : reports_to

Step 4: Query the Graph

Find all of Alice's connections:

TRAVERSE FROM 'node:1' DEPTH 1 DIRECTION OUTGOING

Find the shortest path between Alice and a topic:

PATH FROM 'node:1' TO 'node:6'

Find who is connected to machine-learning:

TRAVERSE FROM 'node:4' DEPTH 1 DIRECTION INCOMING

Step 5: Add Properties and Query by Property

Update a node with additional properties:

NODE UPDATE 'node:1' SET experience_years = 5

Step 6: Combine with Relational Data

Create a table to store project metadata alongside the graph:

CREATE TABLE projects (
    id INT PRIMARY KEY,
    name TEXT NOT NULL,
    status TEXT,
    lead_node TEXT
);
INSERT INTO projects VALUES (1, 'ML Pipeline', 'active', 'node:1');
INSERT INTO projects VALUES (2, 'Design System', 'active', 'node:2');

Now you can query both relational and graph data:

SELECT * FROM projects WHERE status = 'active';
TRAVERSE FROM 'node:1' DEPTH 2 DIRECTION OUTGOING

Verification

You should see:

  • 6 nodes (3 people, 3 topics)
  • 8 edges (5 expertise, 3 collaboration)
  • 2 projects in the relational table
  • Traversal results showing connected nodes and edges

Next Steps