From RecSysWiki
Jump to navigation Jump to search

Amazon is the largest online retailer as well, with its subsidiary Amazon Web Services (AWS), the largest cloud provider in the Western world. They were also one of the first, if not the first, commercial user of recommendation systems. AWS also offers recommendations as a service with their product AWS Personalize.


  1. all search and information retrieval publications by Amazon
  2. Recommendations: Item-to-Item Collaborative Filtering, IEEE Internet Computing, 2003
  3. Estimating the Causal Impact of Recommendation Systems from Observational Data, EC 2015
  4. One-Pass Ranking Models for Low-Latency Product Recommendations, KDD 2015
  5. πŸ‘“ Adaptive, personalized diversity for visual discovery, RecSys 2016 (best short paper)
  6. Diversifying Music Recommendations, ICML 2016
  7. Sustainability at Scale: Towards Bridging the Intention-Behavior Gap with Sustainable Recommendations, RecSys 2017
  8. Recommending Product Sizes to Customers, RecSys 2017
  9. πŸ•΄ Two Decades of Recommender Systems at, 2017
  10. Intent Based Relevance Estimation from Click Logs, CIKM 2017
  11. MRNet-Product2Vec: A Multi-task Recurrent Neural Network for Product Embeddings, ECML-PKDD 2017
  12. πŸ“‘πŸŽ° Learning to Rank in the Position Based Model with Bandit Feedback (Amazon Music)
  13. 🎰 A Linear Bandit for Seasonal Environments (Amazon Music)
  14. An efficient neighborhood-based interaction model for recommendation on heterogeneous graph
  15. 🎰 An Efficient Bandit Algorithm for Realtime Multivariate Optimization, KDD 2017
  16. 🧠 The Effectiveness of a Two-layer Neural Network for Recommendations, ICLR 2018
  17. 🎰 Contextual Multi-Armed Bandits for Causal Marketing, ICML 2018
  18. Buy It Again: Modeling Repeat Purchase Recommendations, KDD 2018
  19. LORE: A Large-Scale Offer Recommendation Engine Through the Lens of an Online Subscription Service
  20. πŸ” Learning Robust Models for e-Commerce Product Search
  21. πŸ” Treating Cold Start in Product Search by Priors
  22. πŸƒ Scalable Feature Selection for (Multitask) Gradient Boosted Trees
  23. πŸ“‘ Multi-objective Relevance Ranking via Constrained Optimization
  24. Language-Agnostic Representation Learning for Product Search on E-Commerce Platforms
  25. πŸ“„ Whole page optimization with local and global constraints, KDD 2019 (Amazon Video)
  26. πŸƒ Large-scale Collaborative Filtering with Product Embeddings, 2019
  27. P-Companion: A principled framework for diversified complementary product recommendation, CIKM 2020
  28. πŸ” Why Do People Buy Seemingly Irrelevant Items in Voice Product Search?, WSDM 2020, blog post
  29. 🧠 Temporal-Contextual Recommendation in Real-Time, KDD 2020 (best applied data science paper), notes
  30. Challenges and research opportunities in ecommerce search and recommendations, SIGIR Forum 2020
  31. A flexible large-scale similar product identification system in e-commerce, KDD 1st International Workshop on Industrial Recommendation 2020
  32. πŸ“‘ CPR: Collaborative pairwise ranking for online list recommendations, RecSys 2020 Workshop on Online Recommender Systems and User Modeling
  33. πŸ‘— Fashion Outfit Complementary Item Retrieva, CVPR 2020
  34. Amazon SageMaker Automatic Model Tuning: Scalable Black-box Optimization, 2020
  35. TabTransformer: Tabular Data Modeling Using Contextual Embeddings, arXiv preprint, 2020
  36. 🎰 Learning from eXtreme bandit feedback
  37. 🧠 Heterogeneous graph neural networks with neighbor-SIM attention mechanism for substitute product recommendation, DLG-AAAI 2021
  38. πŸ” Seasonal relevance in e-commerce search, CIKM 2021
  39. πŸ‘— Contrastive Learning for Interactive Recommendation in Fashion, SIGIR 2022
  40. πŸ”πŸ”¬ N. Bi, P. Castells, D. Gilbert, S. Galperin, P. Tardif, S. Ahuja: Debiased balanced interleaving at Amazon Search, CIKM 2022
  41. A Data-Driven State Aggregation Approach for Dynamic Discrete Choice Models, UAI 2023
  42. Neural Insights for Digital Marketing Content Design, KDD 2023

Blog posts

Articles about Amazon

  1. πŸ” Amazon is stuffing its search results pages with ads


  • open source software library for training and deploying recommendation models with sparse inputs, fully connected hidden layers, and sparse outputs. Models with weight matrices that are too large for a single GPU can still be trained on a single host. DSSTNE has been used at Amazon to generate personalized product recommendations for our customers at Amazon’s scale.

External links