Knowledge-Graph Recommender for a Niche Online Bookstore
Overview
What this challenge is about.
Model the catalog as a knowledge graph (nodes: books, authors, genres, themes, eras, awards; edges: wrote, in-genre, has-theme, won, similar-to). Use Neo4j or a simple Python in-memory graph. Implement a recommender that takes a book ID and returns top-10 recommendations via traversal rules (e.g. same-theme + different-era; same-author + lesser-known; award-adjacent). Add a similarity scoring layer (Jaccard over themes + simple text embedding cosine for blurbs). Evaluate against staff picks via a precision-at-10 study with 30 user-history triples (provided). Deliver: graph schema + populated DB, recommender Python service, 5-page evaluation report, integration spec for the Next.js storefront.
The Brief
What you'll do, and what you'll demonstrate.
Build a knowledge-representation-based recommender that matches staff-pick precision-at-10 within 15 percent, using a 8,000-title knowledge graph.
Earning criteria — what you'll demonstrate
- Model a real domain as a knowledge graph with explicit semantics
- Implement rule-based + similarity-based recommendations on a graph
- Evaluate recommenders against human-curated baselines with precision-at-k
- Specify a recommender integration for a production storefront
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
Careers
Roles this prepares you for.
Real titles. Real skill bridges. Pick the one closest to your trajectory.
Product Manager
PMs at curation-driven commerce companies need this knowledge-graph fluency to scope recommendations that fit the brand voice.
This challenge sharpens
- recommender-systems
- knowledge-representation
- algorithm-evaluation