People who liked this talk also liked ... Building Recommendation...
summary
Shows a concrete recipe: build a memory‑based user‑item matrix, compute Pearson correlation coefficient (example implemented in ~40 lines of Ruby) and pick top‑k neighbors (k=2 demo) to predict ratings. Covers content models with K‑means, a simple hybrid intersection/backup combiner, and practical failures: cold start, data sparsity and memory cost. Explains evaluation with precision/recall and click‑through rate, and points to Apache Mahout/JRuby Mahout, SC Ruby and recommenderlab (R).
AI-written summary · about
chapters (11)
- 0:00Introduction
- 2:54What is a recommendation system
- 5:44Collaborative filtering basics
- 10:57Similarity to recommendations
- 12:41Problems with collaborative filtering
- 14:25Content-based systems
- 17:18Limitations of content-based
- 18:19Hybrid algorithms
- 21:05Evaluating recommendations
- 23:54Libraries and tools
- 26:03Q&A
related talks
23mFrom Data to Recommendations: Building an Intelligent System with ROR
55mBuilding a Recommendation Engine with Machine Learning Techniques
43mBuilding your first recommender
41mRecommender Systems: Crafting Effective Personalised Experiences
58mRecommender Systems – A Walkthrough
33mPicking What to Watch Next: Build a Recommendation System
37mA primer on recommendation systems.
31mCreating recommender systems with Spark, Scala and Prediction.Io