skills.rest
cenjie
cenjieCOMMUNITY
Explore all skills by this author →
GitHub

Recommendation System

Develop end-to-end recommendation systems with collaborative filtering, matrix factorization, and A/B testing.

Updated May 17, 2026
Search Tags:#personalization#recommendations#content-based#matrix-factorization#collaborative-filtering#hybrid-models
One-click install
npx skills add https://github.com/cenjie/skills --skill recommendation-system
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: Recommendation System
Source: https://github.com/cenjie/skills/tree/main/skills/recommendation-system
Command: npx skills add https://github.com/cenjie/skills --skill recommendation-system

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Build and deploy end-to-end recommendation systems that predict user preferences and surface personalized item suggestions, improving engagement and conversions.

Core Features & Use Cases

  • Collaborative Filtering: Recommend items to users based on similar users' behavior.
  • Content-based: Recommend items using item attributes and user preferences.
  • Hybrid: Combine collaborative and content-based signals for better accuracy.
  • Matrix Factorization & Deep Learning: Learn latent representations to reveal hidden patterns in user-item data.
  • Production-ready evaluation: Use metrics like Precision@K, Recall@K, NDCG, coverage, and diversity; run A/B tests to measure business impact.

Quick Start

Train a small-scale, end-to-end pipeline to build a user-item matrix, compute similarity or latent factors, generate top-N recommendations for a sample user, and evaluate results.

Frequently Asked Questions about Recommendation System

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a recommendation system for cold-start scenarios in e-commerce?▼

Build an end-to-end recommendation pipeline by constructing a user-item matrix, computing similarity or latent factors via matrix factorization, generating top-N personalized recommendations, and evaluating results with metrics like Precision@K and NDCG.

What is the difference between collaborative filtering and content-based recommendations?▼

Collaborative filtering recommends items by analyzing similar users' behavior, whereas content-based recommendations match item attributes to user preferences. Hybrid models combine both signals to improve accuracy and mitigate cold-start limitations.

Does this approach support matrix factorization and deep learning for latent representations?▼

Evaluate recommendation systems using metrics like Precision@K, Recall@K, NDCG, coverage, and diversity. You can also run A/B tests to measure the actual business impact of personalized item suggestions on engagement and conversions.

How do I measure the impact of personalized recommendations with A/B testing?▼

Measure the impact of personalized recommendations by running A/B tests alongside evaluation metrics like Precision@K, Recall@K, NDCG, coverage, and diversity to quantify improvements in user engagement and conversions.

When should I use hybrid models instead of pure collaborative filtering?▼

Use hybrid models instead of pure collaborative filtering when dealing with sparse interactions and cold-start scenarios. Combining content-based attributes with behavioral signals yields better accuracy and coverage for personalized recommendations.

Related Skills

Explore More Data & Analytics Skills→
liangdabiaoliangdabiao
264

recommender-system

Analyze user behavior data to train recommender models and generate top-K recommendations.

Community
Advanced
h4vzzh4vzz
33

model-deployment

Deploy trained ML models as REST services on Docker or Kubernetes.

Community
Advanced
marchattonmarchatton
5

evaluate-rag

Evaluate RAG pipelines with Recall@k, Precision@k, MRR, and NDCG@k metrics.

Community
Advanced
dralkhdralkh
74

aeon

Train scikit-learn compatible time series classifiers, forecasters, and clusterers with aeon 1.x APIs.

Community
Advanced