Workshop: Building a Music Recommender System in Java
summary
A hands-on workshop showing how to build a music recommender system from scratch in Java using an open dataset, demonstrating which naive approaches fail in real systems. It explains how ML tasks are handled in heterogeneous production environments (mixing Java and Python) with examples from Odnoklassniki, aimed at developers new to machine learning.
AI-written summary · about
chapters (18)
- 0:00Workshop intro
- 2:31Speaker intro
- 3:34Recommendation systems types
- 7:59Task and dataset
- 9:04Java and KNN choice
- 10:51Collaborative filtering
- 13:52Sparse matrices & libraries
- 16:24Java libraries selection
- 20:22Reading and processing data
- 21:30Smile matrix operations
- 24:44Testing the recommender
- 26:53Popularity bias
- 29:04Production vs local
- 31:12Vector search in prod
- 35:20Content-based bonus
- 37:29LLMs in recommendations
- 39:55Key takeaways
- 41:01Q&A
related talks
43mBuilding your first recommender
55mBuilding a Recommendation Engine with Machine Learning Techniques
01h 38mBuilding Recommender Systems Using Python
01h 31mHands on - Build a Recommender system
34mBuilding a Large Music Recommender Leveraging AI, Deep Learning and Human Expertise
20mBuilding Recommender system
40mBuilding A Folk Music Recommendation System with LLMs
58mRecommender Systems – A Walkthrough