Recommendation Systems
Recommendation Systems (16): Industrial Architecture and Best Practices
Production recommendation systems serve hundreds of millions of users with sub-100ms latency. This final article covers the industrial multi-stage pipeline (recall, coarse ranking, fine ranking, reranking), feature …
Recommendation Systems (15): Real-Time Recommendation and Online Learning
A practitioner's guide to real-time recommendation: streaming pipelines (Kafka + Flink), online learning (SGD, FTRL, AdaGrad), bandits (UCB, Thompson Sampling, LinUCB), latency budgets, feature freshness, concept drift, …
Recommendation Systems (14): Cross-Domain Recommendation and Cold-Start Solutions
Cold-start and cross-domain recommendation in depth: the three faces of cold-start, EMCDR/PTUPCDR cross-domain bridges, MeLU/MAML meta-learning, UCB bandits for exploration, and the cold-to-warm production stack.
Recommendation Systems (13): Fairness, Debiasing, and Explainability
A practical deep dive into trustworthy recommendation: the seven biases (popularity, position, selection, exposure, conformity, demographic, confirmation), causal inference (RCTs, IPS, doubly robust estimators), …
Recommendation Systems (12): Large Language Models and Recommendation
How LLMs reshape recommendation: enhancers (P5, M6Rec), predictors (TallRec, GenRec), and agents (LlamaRec, ChatREC). Hybrid pipelines, cold-start wins, prompt design, and the cost/quality Pareto frontier.
Recommendation Systems (11): Contrastive Learning and Self-Supervised Learning
A practitioner's guide to contrastive learning for recommendations: InfoNCE and the role of temperature, SimCLR vs MoCo negatives, SGL graph augmentations, CL4SRec sequence augmentations, XSimGCL's noise-only trick, with …
Recommendation Systems (10): Deep Interest Networks and Attention Mechanisms
From DIN's target attention to DIEN's AUGRU and BST's Transformer — how Alibaba taught CTR models to read a user's history like a chef reads the room.
Recommendation Systems (9): Multi-Task Learning and Multi-Objective Optimization
How real recommenders juggle clicks, conversions, watch time and revenue at once. Shared-Bottom, ESMM, MMoE, PLE explained from first principles, with PyTorch code, loss-balancing strategies and the gradient-conflict …
Recommendation Systems (8): Knowledge Graph-Enhanced Recommendation
Learn how knowledge graphs supercharge recommendation systems by adding semantic understanding. Covers RippleNet, KGCN, KGAT, CKE, and path-based reasoning -- with intuitive explanations, real-world analogies, and …
Recommendation Systems (7): Graph Neural Networks and Social Recommendation
A deep, intuition-first walkthrough of graph neural networks for recommendation: GCN, GAT, GraphSAGE, PinSage, LightGCN, NGCF, social signals, scalable sampling, and cold start. Diagrams plus working PyTorch.
Recommendation Systems (6): Sequential Recommendation and Session-based Modeling
How recommenders use the order of user actions to predict the next one. Markov chains, GRU4Rec, Caser, SASRec, BERT4Rec, BST, and SR-GNN, with implementations and intuition.
Recommendation Systems (5): Embedding and Representation Learning
How modern recommenders learn dense vector representations for users and items: Word2Vec / Item2Vec, Node2Vec, two-tower DSSM and YouTube DNN, negative sampling, FAISS/HNSW serving, and how to evaluate embedding quality. …
Recommendation Systems (4): CTR Prediction and Click-Through Rate Modeling
A practical guide to CTR prediction models -- from Logistic Regression and Factorization Machines to DeepFM, xDeepFM, DCN, AutoInt, and FiBiNet -- with intuitive explanations and PyTorch implementations.
Recommendation Systems (3): Deep Learning Foundations
From MLPs to embeddings to NeuMF, YouTube DNN, and Wide & Deep -- a progressive walkthrough of the deep learning building blocks that power every modern recommender, with verified architectures and runnable PyTorch code.
Recommendation Systems (2): Collaborative Filtering and Matrix Factorization
An in-depth tour of collaborative filtering and matrix factorization: User-CF and Item-CF, similarity metrics, latent-factor models, SVD++, ALS, BPR, and factorization machines — with intuitions, derivations, and …
Recommendation Systems (1): Fundamentals and Core Concepts
A beginner-friendly guide to recommendation systems: the three core paradigms (collaborative filtering, content-based, hybrid), evaluation metrics, the multi-stage funnel architecture used in production, and the open …