--- title: Recommender Systems description: Recommender systems rank items for users at Netflix, Amazon and Spotify scale. You will learn collaborative filtering, matrix factorisation, deep and sequential models, evaluation and the feedback-loop pitfalls. category: programming-tech subcategory: artificial-intelligence difficulty: beginner, intermediate, advanced url: /subject/recommender-systems --- # Recommender Systems Recommender systems rank items for users at Netflix, Amazon and Spotify scale. You will learn collaborative filtering, matrix factorisation, deep and sequential models, evaluation and the feedback-loop pitfalls. ## Available Resources 2 Books • 1 Courses • 1 Websites • 5 Papers ## Papers ### 1. How Algorithmic Confounding in Recommendation Systems Increases Homogeneity and Decreases Utility **Author:** Allison J. B. Chaney, Brandon M. Stewart, Barbara E. Engelhardt RecSys 2018 paper using simulation to show how retraining recommenders on data their own recommendations shaped creates feedback loops that homogenise user behaviour, reduce utility, and bias offline evaluation. **Difficulty:** Advanced | **Language:** English | **Price:** Free **Link:** https://arxiv.org/abs/1710.11214 **Tags:** recommender-systems, feedback-loops, algorithmic-confounding, homogenization, offline-evaluation, simulation ### 2. Are We Really Making Much Progress? A Worrying Analysis of Recent Neural Recommendation Approaches **Author:** Maurizio Ferrari Dacrema, Paolo Cremonesi, Dietmar Jannach RecSys 2019 best paper that tried to reproduce 18 neural top-n recommendation methods and found most could not be reproduced or were beaten by tuned nearest-neighbour and simple baselines, exposing weak baselines and evaluation flaws. Covers the topic's evaluation pillar: teaches why offline comparisons mislead and why strong baselines matter. Essential corrective after reading about deep models. **Difficulty:** Advanced | **Language:** English | **Price:** Free **Link:** https://arxiv.org/abs/1907.06902 **Tags:** recommender-systems, reproducibility, neural-recommenders, evaluation-methodology, baselines, top-n-recommendation ### 3. Deep Neural Networks for YouTube Recommendations **Author:** Paul Covington, Jay Adams, Emre Sargin Primary source for deep recommenders at scale. RecSys 2016 paper describing YouTube's two-stage deep recommender: a candidate generation network trained as extreme multiclass classification and a separate ranking network, plus practical lessons on features, example age and watch-time objectives. **Difficulty:** Advanced | **Language:** English | **Price:** Free **Link:** https://research.google/pubs/deep-neural-networks-for-youtube-recommendations/ **Tags:** recommender-systems, deep-learning, candidate-generation, ranking-models, youtube, two-stage-recommenders ### 4. Matrix Factorization Techniques for Recommender Systems **Author:** Yehuda Koren, Robert Bell, Chris Volinsky The primary source for matrix factorization as practised in industry; short (8 pages) and readable. The 2009 IEEE Computer article by members of the Netflix Prize-winning team explaining latent factor models: SGD and ALS training, user and item biases, implicit feedback, temporal dynamics and confidence weighting, with results on Netflix data. **Difficulty:** Intermediate | **Language:** English | **Price:** Paid **Link:** https://ieeexplore.ieee.org/document/5197422 **Tags:** recommender-systems, matrix-factorization, collaborative-filtering, latent-factor-models, netflix-prize, implicit-feedback ### 5. Monolith: Real Time Recommendation System With Collisionless Embedding Table **Author:** Zhuoran Liu, Leqi Zou, Xuan Zou, et al. (ByteDance) A ByteDance systems paper describing Monolith, a real-time recommendation training system using a collisionless embedding table built on a Cuckoo hash map, expirable embeddings, frequency filtering, and online training that trades fault tolerance for fresher models. Readers learn how production recommenders handle sparse features. **Difficulty:** Advanced | **Language:** English | **Price:** Free **Link:** https://arxiv.org/abs/2209.07663 **Tags:** recommender-systems, embeddings, online-training, real-time-ml, ml-systems ## Websites ### 1. System Design for Recommendations and Search **Author:** Eugene Yan The best free bridge from textbook models to how real systems are wired. Synthesis of primary industry sources with links, not a rehash. Essay comparing how Alibaba, Facebook, JD, DoorDash, LinkedIn and others structure recommendation and search systems into offline and online components and retrieval and ranking stages, with links to each company's primary write-up. **Difficulty:** Intermediate | **Language:** English | **Price:** Free **Link:** https://eugeneyan.com/writing/system-design-for-discovery/ **Tags:** recommender-systems, search-systems, system-design, retrieval-and-ranking, ml-systems ## Books ### 1. Recommender Systems: The Textbook **Author:** Charu C. Aggarwal The single reference text for the field: comprehensive graduate textbook covering neighborhood and model-based collaborative filtering, latent factor models, content-based, knowledge-based and hybrid systems, evaluation methodology, context-aware and time-sensitive recommendation, and attack-resistant and social recommenders. **Difficulty:** Advanced | **Language:** English | **Price:** Paid **Link:** https://www.amazon.com/dp/3319296574?tag=edmonddante07-20 **Tags:** recommender-systems, collaborative-filtering, content-based-filtering, latent-factor-models, hybrid-recommenders, recommender-evaluation ### 2. Mining of Massive Datasets, Chapter 9: Recommendation Systems **Author:** Jure Leskovec, Anand Rajaraman, Jeffrey D. Ullman Free chapter from the Stanford CS246 textbook covering the utility matrix, content-based recommendation, user-user and item-item collaborative filtering, UV-decomposition by gradient descent, and a case study of the Netflix Prize. Compact, mathematically clean treatment of the classical toolkit (CF, UV-decomposition) with exercises. **Difficulty:** Intermediate | **Language:** English | **Price:** Free **Link:** http://infolab.stanford.edu/~ullman/mmds/ch9.pdf **Tags:** recommender-systems, collaborative-filtering, content-based-filtering, uv-decomposition, utility-matrix, netflix-prize ## Courses ### 1. Recommendation Systems (Google Machine Learning Course) Free four-hour Google course on how production recommenders are built: candidate generation, scoring and re-ranking, embeddings, content-based and collaborative filtering, matrix factorization and softmax deep models, with Colab exercises on MovieLens data. Best free on-ramp: short, architecture-first, and covers both classic CF/MF and neural retrieval with runnable notebooks. **Difficulty:** Beginner | **Language:** English | **Price:** Free **Link:** https://developers.google.com/machine-learning/recommendation **Tags:** recommender-systems, collaborative-filtering, matrix-factorization, embeddings, candidate-generation --- *This content is part of Dantes.io - Your Treasure Map to Knowledge* *Curated by humans at Dantes.io. Personal study use welcome; republishing this curation requires permission (team@dantes.io).* View this page online: https://dantes.io/subject/recommender-systems