[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-9565":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":10,"languages":10,"totalLinesOfCode":10,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":16,"stars7d":17,"stars30d":18,"stars90d":15,"forks30d":15,"starsTrendScore":19,"compositeScore":20,"rankGlobal":10,"rankLanguage":10,"license":21,"archived":22,"fork":22,"defaultBranch":23,"hasWiki":24,"hasPages":22,"topics":25,"createdAt":10,"pushedAt":10,"updatedAt":40,"readmeContent":41,"aiSummary":42,"trendingCount":15,"starSnapshotCount":15,"syncStatus":43,"lastSyncTime":44,"discoverSource":45},9565,"applied-ml","eugeneyan\u002Fapplied-ml","eugeneyan","📚 Papers & tech blogs by companies sharing their work on data science & machine learning in production.","",null,29736,3945,953,3,0,7,104,898,59,45,"MIT License",false,"main",true,[26,27,28,29,30,31,32,33,34,35,36,37,38,39],"applied-data-science","applied-machine-learning","computer-vision","data-discovery","data-engineering","data-quality","data-science","deep-learning","machine-learning","natural-language-processing","production","recsys","reinforcement-learning","search","2026-06-12 02:02:09","# applied-ml\nCurated papers, articles, and blogs on **data science & machine learning in production**. ⚙️\n\n[![contributions welcome](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fcontributions-welcome-brightgreen.svg?style=flat)](.\u002FCONTRIBUTING.md) [![Summaries](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fsummaries-in%20tweets-%2300acee.svg?style=flat)](https:\u002F\u002Ftwitter.com\u002Feugeneyan\u002Fstatus\u002F1350509546133811200) ![HitCount](http:\u002F\u002Fhits.dwyl.com\u002Feugeneyan\u002Fapplied-ml.svg)\n\nFiguring out how to implement your ML project? Learn how other organizations did it:\n\n- **How** the problem is framed 🔎(e.g., personalization as recsys vs. search vs. sequences)\n- **What** machine learning techniques worked ✅ (and sometimes, what didn't ❌)\n- **Why** it works, the science behind it with research, literature, and references 📂\n- **What** real-world results were achieved (so you can better assess ROI ⏰💰📈)\n\nP.S., Want a summary of ML advancements? 👉[`ml-surveys`](https:\u002F\u002Fgithub.com\u002Feugeneyan\u002Fml-surveys)\n\nP.P.S, Looking for guides and interviews on applying ML? 👉[`applyingML`](https:\u002F\u002Fapplyingml.com)\n\n**Table of Contents**\n\n1. [Data Quality](#data-quality)\n2. [Data Engineering](#data-engineering)\n3. [Data Discovery](#data-discovery)\n4. [Feature Stores](#feature-stores)\n5. [Classification](#classification)\n6. [Regression](#regression)\n7. [Forecasting](#forecasting)\n8. [Recommendation](#recommendation)\n9. [Search & Ranking](#search--ranking)\n10. [Embeddings](#embeddings)\n11. [Natural Language Processing](#natural-language-processing)\n12. [Sequence Modelling](#sequence-modelling)\n13. [Computer Vision](#computer-vision)\n14. [Reinforcement Learning](#reinforcement-learning)\n15. [Anomaly Detection](#anomaly-detection)\n16. [Graph](#graph)\n17. [Optimization](#optimization)\n18. [Information Extraction](#information-extraction)\n19. [Weak Supervision](#weak-supervision)\n20. [Generation](#generation)\n21. [Audio](#audio)\n22. [Privacy-Preserving Machine Learning](#privacy-preserving-machine-learning)\n23. [Validation and A\u002FB Testing](#validation-and-ab-testing)\n24. [Model Management](#model-management)\n25. [Efficiency](#efficiency)\n26. [Ethics](#ethics)\n27. [Infra](#infra)\n28. [MLOps Platforms](#mlops-platforms)\n29. [Practices](#practices)\n30. [Team Structure](#team-structure)\n31. [Fails](#fails)\n\n## Data Quality\n1. [Reliable and Scalable Data Ingestion at Airbnb](https:\u002F\u002Fwww.slideshare.net\u002FHadoopSummit\u002Freliable-and-scalable-data-ingestion-at-airbnb-63920989) `Airbnb` `2016`\n2. [Monitoring Data Quality at Scale with Statistical Modeling](https:\u002F\u002Feng.uber.com\u002Fmonitoring-data-quality-at-scale\u002F) `Uber` `2017`\n3. [Data Management Challenges in Production Machine Learning](https:\u002F\u002Fresearch.google\u002Fpubs\u002Fpub46178\u002F) ([Paper](https:\u002F\u002Fthodrek.github.io\u002FCS839_spring18\u002Fpapers\u002Fp1723-polyzotis.pdf)) `Google` `2017`\n4. [Automating Large-Scale Data Quality Verification](https:\u002F\u002Fwww.amazon.science\u002Fpublications\u002Fautomating-large-scale-data-quality-verification) ([Paper](https:\u002F\u002Fassets.amazon.science\u002Fa6\u002F88\u002Fad858ee240c38c6e9dce128250c0\u002Fautomating-large-scale-data-quality-verification.pdf))`Amazon` `2018`\n5. [Meet Hodor — Gojek’s Upstream Data Quality Tool](https:\u002F\u002Fwww.gojek.io\u002Fblog\u002Fmeet-hodor-gojeks-upstream-data-quality-tool) `Gojek` `2019`\n6. [Data Validation for Machine Learning](https:\u002F\u002Fresearch.google\u002Fpubs\u002Fpub47967\u002F) ([Paper](https:\u002F\u002Fmlsys.org\u002FConferences\u002F2019\u002Fdoc\u002F2019\u002F167.pdf)) `Google` `2019`\n6. [An Approach to Data Quality for Netflix Personalization Systems](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=t7vHpA39TXM) `Netflix` `2020`\n7. [Improving Accuracy By Certainty Estimation of Human Decisions, Labels, and Raters](https:\u002F\u002Fresearch.fb.com\u002Fblog\u002F2020\u002F08\u002Fimproving-the-accuracy-of-community-standards-enforcement-by-certainty-estimation-of-human-decisions\u002F) ([Paper](https:\u002F\u002Fresearch.fb.com\u002Fwp-content\u002Fuploads\u002F2020\u002F08\u002FCLARA-Confidence-of-Labels-and-Raters.pdf)) `Facebook` `2020`\n\n## Data Engineering\n1. [Zipline: Airbnb’s Machine Learning Data Management Platform](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=Tg5VEMEsC-0) `Airbnb` `2018`\n2. [Sputnik: Airbnb’s Apache Spark Framework for Data Engineering](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=BQumogSBsUw) `Airbnb` `2020`\n3. [Unbundling Data Science Workflows with Metaflow and AWS Step Functions](https:\u002F\u002Fnetflixtechblog.com\u002Funbundling-data-science-workflows-with-metaflow-and-aws-step-functions-d454780c6280) `Netflix` `2020`\n4. [How DoorDash is Scaling its Data Platform to Delight Customers and Meet Growing Demand](https:\u002F\u002Fdoordash.engineering\u002F2020\u002F09\u002F25\u002Fhow-doordash-is-scaling-its-data-platform\u002F) `DoorDash` `2020`\n5. [Revolutionizing Money Movements at Scale with Strong Data Consistency](https:\u002F\u002Feng.uber.com\u002Fmoney-scale-strong-data\u002F) `Uber` `2020`\n6. [Zipline - A Declarative Feature Engineering Framework](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=LjcKCm0G_OY) `Airbnb` `2020`\n7. [Automating Data Protection at Scale, Part 1](https:\u002F\u002Fmedium.com\u002Fairbnb-engineering\u002Fautomating-data-protection-at-scale-part-1-c74909328e08) ([Part 2](https:\u002F\u002Fmedium.com\u002Fairbnb-engineering\u002Fautomating-data-protection-at-scale-part-2-c2b8d2068216)) `Airbnb` `2021`\n8. [Real-time Data Infrastructure at Uber](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2104.00087.pdf) `Uber` `2021`\n9. [Introducing Fabricator: A Declarative Feature Engineering Framework](https:\u002F\u002Fdoordash.engineering\u002F2022\u002F01\u002F11\u002Fintroducing-fabricator-a-declarative-feature-engineering-framework\u002F) `DoorDash` `2022`\n10. [Functions & DAGs: introducing Hamilton, a microframework for dataframe generation](https:\u002F\u002Fmultithreaded.stitchfix.com\u002Fblog\u002F2021\u002F10\u002F14\u002Ffunctions-dags-hamilton\u002F) `Stitch Fix` `2021`\n11. [Optimizing Pinterest’s Data Ingestion Stack: Findings and Learnings](https:\u002F\u002Fmedium.com\u002F@Pinterest_Engineering\u002Foptimizing-pinterests-data-ingestion-stack-findings-and-learnings-994fddb063bf) `Pinterest` `2022`\n12. [Lessons Learned From Running Apache Airflow at Scale](https:\u002F\u002Fshopifyengineering.myshopify.com\u002Fblogs\u002Fengineering\u002Flessons-learned-apache-airflow-scale) `Shopify` `2022`\n13. [Understanding Data Storage and Ingestion for Large-Scale Deep Recommendation Model Training](https:\u002F\u002Farxiv.org\u002Fabs\u002F2108.09373v4) `Meta` `2022`\n14. [Data Mesh — A Data Movement and Processing Platform @ Netflix](https:\u002F\u002Fnetflixtechblog.com\u002Fdata-mesh-a-data-movement-and-processing-platform-netflix-1288bcab2873) `Netflix` `2022`\n15. [Building Scalable Real Time Event Processing with Kafka and Flink](https:\u002F\u002Fdoordash.engineering\u002F2022\u002F08\u002F02\u002Fbuilding-scalable-real-time-event-processing-with-kafka-and-flink\u002F) `DoorDash` `2022`\n\n## Data Discovery\n1. [Apache Atlas: Data Goverance and Metadata Framework for Hadoop](https:\u002F\u002Fatlas.apache.org\u002F#\u002F) ([Code](https:\u002F\u002Fgithub.com\u002Fapache\u002Fatlas)) `Apache`\n2. [Collect, Aggregate, and Visualize a Data Ecosystem's Metadata](https:\u002F\u002Fmarquezproject.github.io\u002Fmarquez\u002F) ([Code](https:\u002F\u002Fgithub.com\u002FMarquezProject\u002Fmarquez)) `WeWork`\n3. [Discovery and Consumption of Analytics Data at Twitter](https:\u002F\u002Fblog.twitter.com\u002Fengineering\u002Fen_us\u002Ftopics\u002Finsights\u002F2016\u002Fdiscovery-and-consumption-of-analytics-data-at-twitter.html) `Twitter` `2016`\n4. [Democratizing Data at Airbnb](https:\u002F\u002Fmedium.com\u002Fairbnb-engineering\u002Fdemocratizing-data-at-airbnb-852d76c51770) `Airbnb` `2017`\n5. [Databook: Turning Big Data into Knowledge with Metadata at Uber](https:\u002F\u002Feng.uber.com\u002Fdatabook\u002F) `Uber` `2018`\n6. [Metacat: Making Big Data Discoverable and Meaningful at Netflix](https:\u002F\u002Fnetflixtechblog.com\u002Fmetacat-making-big-data-discoverable-and-meaningful-at-netflix-56fb36a53520) ([Code](https:\u002F\u002Fgithub.com\u002FNetflix\u002Fmetacat)) `Netflix` `2018`\n7. [Amundsen — Lyft’s Data Discovery & Metadata Engine](https:\u002F\u002Feng.lyft.com\u002Famundsen-lyfts-data-discovery-metadata-engine-62d27254fbb9) `Lyft` `2019`\n8. [Open Sourcing Amundsen: A Data Discovery And Metadata Platform](https:\u002F\u002Feng.lyft.com\u002Fopen-sourcing-amundsen-a-data-discovery-and-metadata-platform-2282bb436234) ([Code](https:\u002F\u002Fgithub.com\u002Flyft\u002Famundsen)) `Lyft` `2019`\n9. [DataHub: A Generalized Metadata Search & Discovery Tool](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2019\u002Fdata-hub) ([Code](https:\u002F\u002Fgithub.com\u002Flinkedin\u002Fdatahub)) `LinkedIn` `2019`\n10. [Amundsen: One Year Later](https:\u002F\u002Feng.lyft.com\u002Famundsen-1-year-later-7b60bf28602) `Lyft` `2020`\n11. [Using Amundsen to Support User Privacy via Metadata Collection at Square](https:\u002F\u002Fdeveloper.squareup.com\u002Fblog\u002Fusing-amundsen-to-support-user-privacy-via-metadata-collection-at-square\u002F) `Square` `2020`\n12. [Turning Metadata Into Insights with Databook](https:\u002F\u002Feng.uber.com\u002Fmetadata-insights-databook\u002F) `Uber` `2020`\n13. [DataHub: Popular Metadata Architectures Explained](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2020\u002Fdatahub-popular-metadata-architectures-explained) `LinkedIn` `2020`\n14. [How We Improved Data Discovery for Data Scientists at Spotify](https:\u002F\u002Fengineering.atspotify.com\u002F2020\u002F02\u002F27\u002Fhow-we-improved-data-discovery-for-data-scientists-at-spotify\u002F) `Spotify` `2020` \n15. [How We’re Solving Data Discovery Challenges at Shopify](https:\u002F\u002Fengineering.shopify.com\u002Fblogs\u002Fengineering\u002Fsolving-data-discovery-challenges-shopify) `Shopify` `2020`\n16. [Nemo: Data discovery at Facebook](https:\u002F\u002Fengineering.fb.com\u002Fdata-infrastructure\u002Fnemo\u002F) `Facebook` `2020`\n17. [Exploring Data @ Netflix](https:\u002F\u002Fnetflixtechblog.com\u002Fexploring-data-netflix-9d87e20072e3) ([Code](https:\u002F\u002Fgithub.com\u002FNetflix\u002Fnf-data-explorer)) `Netflix` `2021`\n\n## Feature Stores\n1. [Distributed Time Travel for Feature Generation](https:\u002F\u002Fnetflixtechblog.com\u002Fdistributed-time-travel-for-feature-generation-389cccdd3907) `Netflix` `2016`\n2. [Building the Activity Graph, Part 2 (Feature Storage Section)](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2017\u002F07\u002Fbuilding-the-activity-graph--part-2) `LinkedIn` `2017`\n3. [Fact Store at Scale for Netflix Recommendations](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=DiwKg8KynVU) `Netflix` `2018`\n4. [Zipline: Airbnb’s Machine Learning Data Management Platform](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=Tg5VEMEsC-0) `Airbnb` `2018`\n5. [Feature Store: The missing data layer for Machine Learning pipelines?](https:\u002F\u002Fwww.hopsworks.ai\u002Fpost\u002Ffeature-store-the-missing-data-layer-in-ml-pipelines) `Hopsworks` `2018`\n6. [Introducing Feast: An Open Source Feature Store for Machine Learning](https:\u002F\u002Fcloud.google.com\u002Fblog\u002Fproducts\u002Fai-machine-learning\u002Fintroducing-feast-an-open-source-feature-store-for-machine-learning) ([Code](https:\u002F\u002Fgithub.com\u002Ffeast-dev\u002Ffeast)) `Gojek` `2019`\n7. [Michelangelo Palette: A Feature Engineering Platform at Uber](https:\u002F\u002Fwww.infoq.com\u002Fpresentations\u002Fmichelangelo-palette-uber\u002F) `Uber` `2019`\n8. [The Architecture That Powers Twitter's Feature Store](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=UNailXoiIrY) `Twitter` `2019`\n9. [Accelerating Machine Learning with the Feature Store Service](https:\u002F\u002Ftechnology.condenast.com\u002Fstory\u002Faccelerating-machine-learning-with-the-feature-store-service) `Condé Nast` `2019` \n10. [Feast: Bridging ML Models and Data](https:\u002F\u002Fwww.gojek.io\u002Fblog\u002Ffeast-bridging-ml-models-and-data) `Gojek` `2020`\n11. [Building a Scalable ML Feature Store with Redis, Binary Serialization, and Compression](https:\u002F\u002Fdoordash.engineering\u002F2020\u002F11\u002F19\u002Fbuilding-a-gigascale-ml-feature-store-with-redis\u002F) `DoorDash` `2020`\n12. [Rapid Experimentation Through Standardization: Typed AI features for LinkedIn’s Feed](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2020\u002Ffeed-typed-ai-features) `LinkedIn` `2020`\n13. [Building a Feature Store](https:\u002F\u002Fnlathia.github.io\u002F2020\u002F12\u002FBuilding-a-feature-store.html) `Monzo Bank` `2020`\n14. [Butterfree: A Spark-based Framework for Feature Store Building](https:\u002F\u002Fmedium.com\u002Fquintoandar-tech-blog\u002Fbutterfree-a-spark-based-framework-for-feature-store-building-48c3640522c7) ([Code](https:\u002F\u002Fgithub.com\u002Fquintoandar\u002Fbutterfree)) `QuintoAndar` `2020`\n15. [Building Riviera: A Declarative Real-Time Feature Engineering Framework](https:\u002F\u002Fdoordash.engineering\u002F2021\u002F03\u002F04\u002Fbuilding-a-declarative-real-time-feature-engineering-framework\u002F) `DoorDash` `2021`\n16. [Optimal Feature Discovery: Better, Leaner Machine Learning Models Through Information Theory](https:\u002F\u002Feng.uber.com\u002Foptimal-feature-discovery-ml\u002F) `Uber` `2021`\n17. [ML Feature Serving Infrastructure at Lyft](https:\u002F\u002Feng.lyft.com\u002Fml-feature-serving-infrastructure-at-lyft-d30bf2d3c32a) `Lyft` `2021`\n18. [Near real-time features for near real-time personalization](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2022\u002Fnear-real-time-features-for-near-real-time-personalization) `LinkedIn` `2022`\n19. [Building the Model Behind DoorDash’s Expansive Merchant Selection](https:\u002F\u002Fdoordash.engineering\u002F2022\u002F04\u002F19\u002Fbuilding-merchant-selection\u002F) `DoorDash` `2022`\n20. [Open sourcing Feathr – LinkedIn’s feature store for productive machine learning](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2022\u002Fopen-sourcing-feathr---linkedin-s-feature-store-for-productive-m) `LinkedIn` `2022`\n21. [Evolution of ML Fact Store](https:\u002F\u002Fnetflixtechblog.com\u002Fevolution-of-ml-fact-store-5941d3231762) `Netflix` `2022`\n22. [Developing scalable feature engineering DAGs](https:\u002F\u002Fouterbounds.com\u002Fblog\u002Fdeveloping-scalable-feature-engineering-dags) `Metaflow + Hamilton` via `Outerbounds` `2022`\n23. [Feature Store Design at Constructor](https:\u002F\u002Fmedium.com\u002Fconstructor-engineering\u002Ffeature-store-design-at-constructor-330b65f64b18) `Constructor.io` `2023`\n\n\n## Classification\n1. [Prediction of Advertiser Churn for Google AdWords](https:\u002F\u002Fresearch.google\u002Fpubs\u002Fpub36678\u002F) ([Paper](https:\u002F\u002Fstorage.googleapis.com\u002Fpub-tools-public-publication-data\u002Fpdf\u002F36678.pdf)) `Google` `2010`\n2. [High-Precision Phrase-Based Document Classification on a Modern Scale](https:\u002F\u002Fengineering.linkedin.com\u002Fresearch\u002F2011\u002Fhigh-precision-phrase-based-document-classification-on-a-modern-scale) ([Paper](http:\u002F\u002Fweb.stanford.edu\u002F~gavish\u002Fdocuments\u002Fphrase_based.pdf)) `LinkedIn` `2011`\n3. [Chimera: Large-scale Classification using Machine Learning, Rules, and Crowdsourcing](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.14778\u002F2733004.2733024) ([Paper](http:\u002F\u002Fpages.cs.wisc.edu\u002F%7Eanhai\u002Fpapers\u002Fchimera-vldb14.pdf)) `Walmart` `2014`\n4. [Large-scale Item Categorization in e-Commerce Using Multiple Recurrent Neural Networks](https:\u002F\u002Fwww.kdd.org\u002Fkdd2016\u002Fsubtopic\u002Fview\u002Flarge-scale-item-categorization-in-e-commerce-using-multiple-recurrent-neur\u002F) ([Paper](https:\u002F\u002Fwww.kdd.org\u002Fkdd2016\u002Fpapers\u002Ffiles\u002Fadf0392-haAemb.pdf)) `NAVER` `2016`\n5. [Learning to Diagnose with LSTM Recurrent Neural Networks](https:\u002F\u002Farxiv.org\u002Fabs\u002F1511.03677) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1511.03677.pdf)) `Google` `2017`\n6. [Discovering and Classifying In-app Message Intent at Airbnb](https:\u002F\u002Fmedium.com\u002Fairbnb-engineering\u002Fdiscovering-and-classifying-in-app-message-intent-at-airbnb-6a55f5400a0c) `Airbnb` `2019`\n7. [Teaching Machines to Triage Firefox Bugs](https:\u002F\u002Fhacks.mozilla.org\u002F2019\u002F04\u002Fteaching-machines-to-triage-firefox-bugs\u002F) `Mozilla` `2019`\n8. [Categorizing Products at Scale](https:\u002F\u002Fengineering.shopify.com\u002Fblogs\u002Fengineering\u002Fcategorizing-products-at-scale) `Shopify` `2020`\n9. [How We Built the Good First Issues Feature](https:\u002F\u002Fgithub.blog\u002F2020-01-22-how-we-built-good-first-issues\u002F) `GitHub` `2020`\n10. [Testing Firefox More Efficiently with Machine Learning](https:\u002F\u002Fhacks.mozilla.org\u002F2020\u002F07\u002Ftesting-firefox-more-efficiently-with-machine-learning\u002F) `Mozilla` `2020`\n11. [Using ML to Subtype Patients Receiving Digital Mental Health Interventions](https:\u002F\u002Fwww.microsoft.com\u002Fen-us\u002Fresearch\u002Fblog\u002Fa-path-to-personalization-using-ml-to-subtype-patients-receiving-digital-mental-health-interventions\u002F) ([Paper](https:\u002F\u002Fjamanetwork.com\u002Fjournals\u002Fjamanetworkopen\u002Ffullarticle\u002F2768347)) `Microsoft` `2020`\n12. [Scalable Data Classification for Security and Privacy](https:\u002F\u002Fengineering.fb.com\u002Fsecurity\u002Fdata-classification-system\u002F) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2006.14109.pdf)) `Facebook` `2020`\n13. [Uncovering Online Delivery Menu Best Practices with Machine Learning](https:\u002F\u002Fdoordash.engineering\u002F2020\u002F11\u002F10\u002Funcovering-online-delivery-menu-best-practices-with-machine-learning\u002F) `DoorDash` `2020`\n14. [Using a Human-in-the-Loop to Overcome the Cold Start Problem in Menu Item Tagging](https:\u002F\u002Fdoordash.engineering\u002F2020\u002F08\u002F28\u002Fovercome-the-cold-start-problem-in-menu-item-tagging\u002F) `DoorDash` `2020`\n15. [Deep Learning: Product Categorization and Shelving](https:\u002F\u002Fmedium.com\u002Fwalmartglobaltech\u002Fdeep-learning-product-categorization-and-shelving-630571e81e96) `Walmart` `2021`\n16. [Large-scale Item Categorization for e-Commerce](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F2396761.2396838) ([Paper](https:\u002F\u002Fwww.researchgate.net\u002Fprofile\u002FJean_David_Ruvini\u002Fpublication\u002F262270957_Large-scale_item_categorization_for_e-commerce\u002Flinks\u002F5512dc3d0cf270fd7e33a0d5\u002FLarge-scale-item-categorization-for-e-commerce.pdf)) `DianPing`, `eBay` `2012`\n17. [Semantic Label Representation with an Application on Multimodal Product Categorization](https:\u002F\u002Fmedium.com\u002Fwalmartglobaltech\u002Fsemantic-label-representation-with-an-application-on-multimodal-product-categorization-63d668b943b7) `Walmart` `2022`\n18. [Building Airbnb Categories with ML and Human-in-the-Loop](https:\u002F\u002Fmedium.com\u002Fairbnb-engineering\u002Fbuilding-airbnb-categories-with-ml-and-human-in-the-loop-e97988e70ebb) `Airbnb` `2022`\n\n\n## Regression\n1. [Using Machine Learning to Predict Value of Homes On Airbnb](https:\u002F\u002Fmedium.com\u002Fairbnb-engineering\u002Fusing-machine-learning-to-predict-value-of-homes-on-airbnb-9272d3d4739d) `Airbnb` `2017`\n2. [Using Machine Learning to Predict the Value of Ad Requests](https:\u002F\u002Fblog.twitter.com\u002Fengineering\u002Fen_us\u002Ftopics\u002Finsights\u002F2020\u002Fusing-machine-learning-to-predict-the-value-of-ad-requests.html) `Twitter` `2020`\n3. [Open-Sourcing Riskquant, a Library for Quantifying Risk](https:\u002F\u002Fnetflixtechblog.com\u002Fopen-sourcing-riskquant-a-library-for-quantifying-risk-6720cc1e4968) ([Code](https:\u002F\u002Fgithub.com\u002FNetflix-Skunkworks\u002Friskquant)) `Netflix` `2020`\n4. [Solving for Unobserved Data in a Regression Model Using a Simple Data Adjustment](https:\u002F\u002Fdoordash.engineering\u002F2020\u002F10\u002F14\u002Fsolving-for-unobserved-data-in-a-regression-model\u002F) `DoorDash` `2020`\n\n## Forecasting\n1. [Engineering Extreme Event Forecasting at Uber with RNN](https:\u002F\u002Feng.uber.com\u002Fneural-networks\u002F) `Uber` `2017`\n2. [Forecasting at Uber: An Introduction](https:\u002F\u002Feng.uber.com\u002Fforecasting-introduction\u002F) `Uber` `2018`\n3. [Transforming Financial Forecasting with Data Science and Machine Learning at Uber](https:\u002F\u002Feng.uber.com\u002Ftransforming-financial-forecasting-machine-learning\u002F) `Uber` `2018`\n4. [Under the Hood of Gojek’s Automated Forecasting Tool](https:\u002F\u002Fwww.gojek.io\u002Fblog\u002Funder-the-hood-of-gojeks-automated-forecasting-tool) `Gojek` `2019`\n5. [BusTr: Predicting Bus Travel Times from Real-Time Traffic](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fabs\u002F10.1145\u002F3394486.3403376) ([Paper](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fpdf\u002F10.1145\u002F3394486.3403376), [Video](https:\u002F\u002Fcrossminds.ai\u002Fvideo\u002F5f3369790576dd25aef288db\u002F)) `Google` `2020`\n6. [Retraining Machine Learning Models in the Wake of COVID-19](https:\u002F\u002Fdoordash.engineering\u002F2020\u002F09\u002F15\u002Fretraining-ml-models-covid-19\u002F) `DoorDash` `2020`\n7. [Automatic Forecasting using Prophet, Databricks, Delta Lake and MLflow](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=TkcpjnLh690) ([Paper](https:\u002F\u002Fpeerj.com\u002Fpreprints\u002F3190.pdf), [Code](https:\u002F\u002Fgithub.com\u002Ffacebook\u002Fprophet)) `Atlassian` `2020`\n8. [Introducing Orbit, An Open Source Package for Time Series Inference and Forecasting](https:\u002F\u002Feng.uber.com\u002Forbit\u002F) ([Paper](https:\u002F\u002Farxiv.org\u002Fabs\u002F2004.08492), [Video](https:\u002F\u002Fyoutu.be\u002FLXDpq_iwcWY), [Code](https:\u002F\u002Fgithub.com\u002Fuber\u002Forbit)) `Uber` `2021`\n9. [Managing Supply and Demand Balance Through Machine Learning](https:\u002F\u002Fdoordash.engineering\u002F2021\u002F06\u002F29\u002Fmanaging-supply-and-demand-balance-through-machine-learning\u002F) `DoorDash` `2021`\n10. [Greykite: A flexible, intuitive, and fast forecasting library](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2021\u002Fgreykite--a-flexible--intuitive--and-fast-forecasting-library) `LinkedIn` `2021`\n11. [The history of Amazon’s forecasting algorithm](https:\u002F\u002Fwww.amazon.science\u002Flatest-news\u002Fthe-history-of-amazons-forecasting-algorithm) `Amazon` `2021`\n11. [DeepETA: How Uber Predicts Arrival Times Using Deep Learning](https:\u002F\u002Feng.uber.com\u002Fdeepeta-how-uber-predicts-arrival-times\u002F) `Uber` `2022`\n12. [Forecasting Grubhub Order Volume At Scale](https:\u002F\u002Fbytes.grubhub.com\u002Fforecasting-grubhub-order-volume-at-scale-a966c2f901d2) `Grubhub` `2022`\n13. [Causal Forecasting at Lyft (Part 1)](https:\u002F\u002Feng.lyft.com\u002Fcausal-forecasting-at-lyft-part-1-14cca6ff3d6d) `Lyft` `2022`\n\n## Recommendation\n1. [Amazon.com Recommendations: Item-to-Item Collaborative Filtering](https:\u002F\u002Fieeexplore.ieee.org\u002Fdocument\u002F1167344) ([Paper](https:\u002F\u002Fwww.cs.umd.edu\u002F~samir\u002F498\u002FAmazon-Recommendations.pdf)) `Amazon` `2003`\n2. [Netflix Recommendations: Beyond the 5 stars (Part 1](https:\u002F\u002Fnetflixtechblog.com\u002Fnetflix-recommendations-beyond-the-5-stars-part-1-55838468f429) ([Part 2](https:\u002F\u002Fnetflixtechblog.com\u002Fnetflix-recommendations-beyond-the-5-stars-part-2-d9b96aa399f5)) `Netflix` `2012`\n3. [How Music Recommendation Works — And Doesn’t Work](https:\u002F\u002Fnotes.variogram.com\u002F2012\u002F12\u002F11\u002Fhow-music-recommendation-works-and-doesnt-work\u002F) `Spotify` `2012`\n4. [Learning to Rank Recommendations with the k -Order Statistic Loss](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F2507157.2507210) ([Paper](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fpdf\u002F10.1145\u002F2507157.2507210)) `Google` `2013`\n5. [Recommending Music on Spotify with Deep Learning](https:\u002F\u002Fbenanne.github.io\u002F2014\u002F08\u002F05\u002Fspotify-cnns.html) `Spotify` `2014`\n6. [Learning a Personalized Homepage](https:\u002F\u002Fnetflixtechblog.com\u002Flearning-a-personalized-homepage-aa8ec670359a) `Netflix` `2015`\n7. [The Netflix Recommender System: Algorithms, Business Value, and Innovation](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F2843948) ([Paper](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fpdf\u002F10.1145\u002F2843948)) `Netflix` `2015`\n7. [Session-based Recommendations with Recurrent Neural Networks](https:\u002F\u002Farxiv.org\u002Fabs\u002F1511.06939) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1511.06939.pdf)) `Telefonica` `2016`\n8. [Deep Neural Networks for YouTube Recommendations](https:\u002F\u002Fstatic.googleusercontent.com\u002Fmedia\u002Fresearch.google.com\u002Fen\u002F\u002Fpubs\u002Farchive\u002F45530.pdf) `YouTube` `2016`\n9. [E-commerce in Your Inbox: Product Recommendations at Scale](https:\u002F\u002Farxiv.org\u002Fabs\u002F1606.07154) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1606.07154.pdf)) `Yahoo` `2016`\n10. [To Be Continued: Helping you find shows to continue watching on Netflix](https:\u002F\u002Fnetflixtechblog.com\u002Fto-be-continued-helping-you-find-shows-to-continue-watching-on-7c0d8ee4dab6) `Netflix` `2016`\n11. [Personalized Recommendations in LinkedIn Learning](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2016\u002F12\u002Fpersonalized-recommendations-in-linkedin-learning) `LinkedIn` `2016`\n12. [Personalized Channel Recommendations in Slack](https:\u002F\u002Fslack.engineering\u002Fpersonalized-channel-recommendations-in-slack\u002F) `Slack` `2016`\n13. [Recommending Complementary Products in E-Commerce Push Notifications](https:\u002F\u002Farxiv.org\u002Fabs\u002F1707.08113) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1707.08113.pdf)) `Alibaba` `2017`\n14. [Artwork Personalization at Netflix](https:\u002F\u002Fnetflixtechblog.com\u002Fartwork-personalization-c589f074ad76) `Netflix` `2017`\n15. [A Meta-Learning Perspective on Cold-Start Recommendations for Items](https:\u002F\u002Fpapers.nips.cc\u002Fpaper\u002F7266-a-meta-learning-perspective-on-cold-start-recommendations-for-items) ([Paper](https:\u002F\u002Fpapers.nips.cc\u002Fpaper\u002F7266-a-meta-learning-perspective-on-cold-start-recommendations-for-items.pdf)) `Twitter` `2017`\n16. [Pixie: A System for Recommending 3+ Billion Items to 200+ Million Users in Real-Time](https:\u002F\u002Farxiv.org\u002Fabs\u002F1711.07601) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1711.07601.pdf)) `Pinterest` `2017`\n17. [Powering Search & Recommendations at DoorDash](https:\u002F\u002Fdoordash.news\u002Fcompany\u002Fpowering-search-recommendations-at-doordash\u002F) `DoorDash` `2017`\n17. [How 20th Century Fox uses ML to predict a movie audience](https:\u002F\u002Fcloud.google.com\u002Fblog\u002Fproducts\u002Fai-machine-learning\u002Fhow-20th-century-fox-uses-ml-to-predict-a-movie-audience) ([Paper](https:\u002F\u002Farxiv.org\u002Fabs\u002F1810.08189)) `20th Century Fox` `2018`\n18. [Calibrated Recommendations](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3240323.3240372) ([Paper](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fpdf\u002F10.1145\u002F3240323.3240372)) `Netflix` `2018`\n19. [Food Discovery with Uber Eats: Recommending for the Marketplace](https:\u002F\u002Feng.uber.com\u002Fuber-eats-recommending-marketplace\u002F) `Uber` `2018`\n20. [Explore, Exploit, and Explain: Personalizing Explainable Recommendations with Bandits](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3240323.3240354) ([Paper](https:\u002F\u002Fstatic1.squarespace.com\u002Fstatic\u002F5ae0d0b48ab7227d232c2bea\u002Ft\u002F5ba849e3c83025fa56814f45\u002F1537755637453\u002FBartRecSys.pdf)) `Spotify` `2018`\n21. [Talent Search and Recommendation Systems at LinkedIn: Practical Challenges and Lessons Learned](https:\u002F\u002Farxiv.org\u002Fabs\u002F1809.06481) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1809.06481.pdf)) `LinkedIn` `2018`\n21. [Behavior Sequence Transformer for E-commerce Recommendation in Alibaba](https:\u002F\u002Farxiv.org\u002Fabs\u002F1905.06874) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1905.06874.pdf)) `Alibaba` `2019`\n22. [SDM: Sequential Deep Matching Model for Online Large-scale Recommender System](https:\u002F\u002Farxiv.org\u002Fabs\u002F1909.00385) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1909.00385.pdf)) `Alibaba` `2019`\n23. [Multi-Interest Network with Dynamic Routing for Recommendation at Tmall](https:\u002F\u002Farxiv.org\u002Fabs\u002F1904.08030) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1904.08030.pdf)) `Alibaba` `2019`\n24. [Personalized Recommendations for Experiences Using Deep Learning](https:\u002F\u002Fwww.tripadvisor.com\u002Fengineering\u002Fpersonalized-recommendations-for-experiences-using-deep-learning\u002F) `TripAdvisor` `2019`\n25. [Powered by AI: Instagram’s Explore recommender system](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fpowered-by-ai-instagrams-explore-recommender-system\u002F) `Facebook` `2019`\n26. [Marginal Posterior Sampling for Slate Bandits](https:\u002F\u002Fwww.ijcai.org\u002Fproceedings\u002F2019\u002F308) ([Paper](https:\u002F\u002Fwww.ijcai.org\u002Fproceedings\u002F2019\u002F0308.pdf)) `Netflix` `2019`\n27. [Food Discovery with Uber Eats: Using Graph Learning to Power Recommendations](https:\u002F\u002Feng.uber.com\u002Fuber-eats-graph-learning\u002F) `Uber` `2019`\n28. [Music recommendation at Spotify](http:\u002F\u002Fsigir.org\u002Fafirm2019\u002Fslides\u002F16.%20Friday%20-%20Music%20Recommendation%20at%20Spotify%20-%20Ben%20Carterette.pdf) `Spotify` `2019`\n29. [Using Machine Learning to Predict what File you Need Next (Part 1)](https:\u002F\u002Fdropbox.tech\u002Fmachine-learning\u002Fcontent-suggestions-machine-learning) `Dropbox` `2019`\n30. [Using Machine Learning to Predict what File you Need Next (Part 2)](https:\u002F\u002Fdropbox.tech\u002Fmachine-learning\u002Fusing-machine-learning-to-predict-what-file-you-need-next-part-2) `Dropbox` `2019`\n31. [Learning to be Relevant: Evolution of a Course Recommendation System](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fpdf\u002F10.1145\u002F3357384.3357817) (**PAPER NEEDED**)`LinkedIn` `2019`\n32. [Temporal-Contextual Recommendation in Real-Time](https:\u002F\u002Fwww.amazon.science\u002Fpublications\u002Ftemporal-contextual-recommendation-in-real-time) ([Paper](https:\u002F\u002Fassets.amazon.science\u002F96\u002F71\u002Fd1f25754497681133c7aa2b7eb05\u002Ftemporal-contextual-recommendation-in-real-time.pdf)) `Amazon` `2020`\n33. [P-Companion: A Framework for Diversified Complementary Product Recommendation](https:\u002F\u002Fwww.amazon.science\u002Fpublications\u002Fp-companion-a-principled-framework-for-diversified-complementary-product-recommendation) ([Paper](https:\u002F\u002Fassets.amazon.science\u002Fd5\u002F16\u002F3f7809974a899a11bacdadefdf24\u002Fp-companion-a-principled-framework-for-diversified-complementary-product-recommendation.pdf)) `Amazon` `2020`\n34. [Deep Interest with Hierarchical Attention Network for Click-Through Rate Prediction](https:\u002F\u002Farxiv.org\u002Fabs\u002F2005.12981) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2005.12981.pdf)) `Alibaba` `2020`\n35. [TPG-DNN: A Method for User Intent Prediction with Multi-task Learning](https:\u002F\u002Farxiv.org\u002Fabs\u002F2008.02122) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2008.02122.pdf)) `Alibaba` `2020`\n36. [PURS: Personalized Unexpected Recommender System for Improving User Satisfaction](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3383313.3412238) ([Paper](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fpdf\u002F10.1145\u002F3383313.3412238)) `Alibaba` `2020`\n37. [Controllable Multi-Interest Framework for Recommendation](https:\u002F\u002Farxiv.org\u002Fabs\u002F2005.09347) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2005.09347)) `Alibaba` `2020`\n38. [MiNet: Mixed Interest Network for Cross-Domain Click-Through Rate Prediction](https:\u002F\u002Farxiv.org\u002Fabs\u002F2008.02974) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2008.02974.pdf)) `Alibaba` `2020`\n39. [ATBRG: Adaptive Target-Behavior Relational Graph Network for Effective Recommendation](https:\u002F\u002Farxiv.org\u002Fabs\u002F2005.12002) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2005.12002.pdf)) `Alibaba` `2020`\n40. [For Your Ears Only: Personalizing Spotify Home with Machine Learning](https:\u002F\u002Fengineering.atspotify.com\u002F2020\u002F01\u002F16\u002Ffor-your-ears-only-personalizing-spotify-home-with-machine-learning\u002F) `Spotify` `2020`\n41. [Reach for the Top: How Spotify Built Shortcuts in Just Six Months](https:\u002F\u002Fengineering.atspotify.com\u002F2020\u002F04\u002F15\u002Freach-for-the-top-how-spotify-built-shortcuts-in-just-six-months\u002F) `Spotify` `2020`\n42. [Contextual and Sequential User Embeddings for Large-Scale Music Recommendation](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3383313.3412248) ([Paper](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fpdf\u002F10.1145\u002F3383313.3412248)) `Spotify` `2020`\n43. [The Evolution of Kit: Automating Marketing Using Machine Learning](https:\u002F\u002Fengineering.shopify.com\u002Fblogs\u002Fengineering\u002Fevolution-kit-automating-marketing-machine-learning) `Shopify` `2020`\n44. [A Closer Look at the AI Behind Course Recommendations on LinkedIn Learning (Part 1)](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2020\u002Fcourse-recommendations-ai-part-one) `LinkedIn` `2020`\n45. [A Closer Look at the AI Behind Course Recommendations on LinkedIn Learning (Part 2)](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2020\u002Fcourse-recommendations-ai-part-two) `LinkedIn` `2020`\n46. [Building a Heterogeneous Social Network Recommendation System](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2020\u002Fbuilding-a-heterogeneous-social-network-recommendation-system) `LinkedIn` `2020`\n47. [How TikTok recommends videos #ForYou](https:\u002F\u002Fnewsroom.tiktok.com\u002Fen-us\u002Fhow-tiktok-recommends-videos-for-you) `ByteDance` `2020`\n48. [Zero-Shot Heterogeneous Transfer Learning from RecSys to Cold-Start Search Retrieval](https:\u002F\u002Farxiv.org\u002Fabs\u002F2008.02930) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2008.02930.pdf)) `Google` `2020`\n49. [Improved Deep & Cross Network for Feature Cross Learning in Web-scale LTR Systems](https:\u002F\u002Farxiv.org\u002Fabs\u002F2008.13535) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2008.13535.pdf)) `Google` `2020`\n50. [Mixed Negative Sampling for Learning Two-tower Neural Networks in Recommendations](https:\u002F\u002Fresearch.google\u002Fpubs\u002Fpub50257\u002F) ([Paper](https:\u002F\u002Fstorage.googleapis.com\u002Fpub-tools-public-publication-data\u002Fpdf\u002Fb9f4e78a8830fe5afcf2f0452862fb3c0d6584ea.pdf)) `Google` `2020`\n51. [Future Data Helps Training: Modeling Future Contexts for Session-based Recommendation](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1906.04473.pdf) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1906.04473.pdf)) `Tencent` `2020`\n52. [A Case Study of Session-based Recommendations in the Home-improvement Domain](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3383313.3412235) ([Paper](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fpdf\u002F10.1145\u002F3383313.3412235)) `Home Depot` `2020`\n53. [Balancing Relevance and Discovery to Inspire Customers in the IKEA App](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3383313.3411550) ([Paper](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fpdf\u002F10.1145\u002F3383313.3411550)) `Ikea` `2020`\n54. [How we use AutoML, Multi-task learning and Multi-tower models for Pinterest Ads](https:\u002F\u002Fmedium.com\u002Fpinterest-engineering\u002Fhow-we-use-automl-multi-task-learning-and-multi-tower-models-for-pinterest-ads-db966c3dc99e) `Pinterest` `2020`\n55. [Multi-task Learning for Related Products Recommendations at Pinterest](https:\u002F\u002Fmedium.com\u002Fpinterest-engineering\u002Fmulti-task-learning-for-related-products-recommendations-at-pinterest-62684f631c12) `Pinterest` `2020`\n56. [Improving the Quality of Recommended Pins with Lightweight Ranking](https:\u002F\u002Fmedium.com\u002Fpinterest-engineering\u002Fimproving-the-quality-of-recommended-pins-with-lightweight-ranking-8ff5477b20e3) `Pinterest` `2020`\n57. [Multi-task Learning and Calibration for Utility-based Home Feed Ranking](https:\u002F\u002Fmedium.com\u002Fpinterest-engineering\u002Fmulti-task-learning-and-calibration-for-utility-based-home-feed-ranking-64087a7bcbad) `Pinterest` `2020`\n57. [Personalized Cuisine Filter Based on Customer Preference and Local Popularity](https:\u002F\u002Fdoordash.engineering\u002F2020\u002F01\u002F27\u002Fpersonalized-cuisine-filter\u002F) `DoorDash` `2020`\n58. [How We Built a Matchmaking Algorithm to Cross-Sell Products](https:\u002F\u002Fwww.gojek.io\u002Fblog\u002Fhow-we-built-a-matchmaking-algorithm-to-cross-sell-products) `Gojek` `2020`\n59. [Lessons Learned Addressing Dataset Bias in Model-Based Candidate Generation](https:\u002F\u002Farxiv.org\u002Fabs\u002F2105.09293) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2105.09293.pdf)) `Twitter` `2021`\n60. [Self-supervised Learning for Large-scale Item Recommendations](https:\u002F\u002Farxiv.org\u002Fabs\u002F2007.12865) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2007.12865.pdf)) `Google` `2021`\n61. [Deep Retrieval: End-to-End Learnable Structure Model for Large-Scale Recommendations](https:\u002F\u002Farxiv.org\u002Fabs\u002F2007.07203) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2007.07203.pdf)) `ByteDance` `2021`\n62. [Using AI to Help Health Experts Address the COVID-19 Pandemic](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fusing-ai-to-help-health-experts-address-the-covid-19-pandemic\u002F) `Facebook` `2021`\n63. [Advertiser Recommendation Systems at Pinterest](https:\u002F\u002Fmedium.com\u002Fpinterest-engineering\u002Fadvertiser-recommendation-systems-at-pinterest-ccb255fbde20) `Pinterest` `2021`\n64. [On YouTube's Recommendation System](https:\u002F\u002Fblog.youtube\u002Finside-youtube\u002Fon-youtubes-recommendation-system\u002F) `YouTube` `2021`\n65. [\"Are you sure?\": Preliminary Insights from Scaling Product Comparisons to Multiple Shops](https:\u002F\u002Farxiv.org\u002Fabs\u002F2107.03256) `Coveo` `2021`\n66. [Mozrt, a Deep Learning Recommendation System Empowering Walmart Store Associates](https:\u002F\u002Fmedium.com\u002Fwalmartglobaltech\u002Fmozrt-a-deep-learning-recommendation-system-empowering-walmart-store-associates-with-a-5d42c08d88da) `Walmart` `2021`\n67. [Understanding Data Storage and Ingestion for Large-Scale Deep Recommendation Model Training](https:\u002F\u002Farxiv.org\u002Fabs\u002F2108.09373) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2108.09373.pdf)) `Meta` `2021`\n67. [The Amazon Music conversational recommender is hitting the right notes](https:\u002F\u002Fwww.amazon.science\u002Flatest-news\u002Fhow-amazon-music-uses-recommendation-system-machine-learning) `Amazon` `2022`\n68. [Personalized complementary product recommendation](https:\u002F\u002Fwww.amazon.science\u002Fpublications\u002Fpersonalized-complementary-product-recommendation) ([Paper](https:\u002F\u002Fassets.amazon.science\u002F6c\u002Fd9\u002Fa0ec3eda4f0fb4312ce0ada41771\u002Fpersonalized-complementary-product-recommendation.pdf)) `Amazon` `2022`\n69. [Building a Deep Learning Based Retrieval System for Personalized Recommendations](https:\u002F\u002Ftech.ebayinc.com\u002Fengineering\u002Fbuilding-a-deep-learning-based-retrieval-system-for-personalized-recommendations\u002F) `eBay` `2022`\n70. [How We Built: An Early-Stage Machine Learning Model for Recommendations](https:\u002F\u002Fwww.onepeloton.com\u002Fpress\u002Farticles\u002Fhow-we-built-machine-learning) `Peloton` `2022`\n71. [Lessons Learned from Building out Context-Aware Recommender Systems](https:\u002F\u002Fwww.onepeloton.com\u002Fpress\u002Farticles\u002Flessons-learned-from-building-context-aware-recommender-systems) `Peloton` `2022`\n72. [Beyond Matrix Factorization: Using hybrid features for user-business recommendations](https:\u002F\u002Fengineeringblog.yelp.com\u002F2022\u002F04\u002Fbeyond-matrix-factorization-using-hybrid-features-for-user-business-recommendations.html) `Yelp` `2022`\n73. [Improving job matching with machine-learned activity features](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2022\u002Fimproving-job-matching-with-machine-learned-activity-features-) `LinkedIn` `2022`\n74. [Understanding Data Storage and Ingestion for Large-Scale Deep Recommendation Model Training](https:\u002F\u002Farxiv.org\u002Fabs\u002F2108.09373v4) `Meta` `2022`\n75. [Blueprints for recommender system architectures: 10th anniversary edition](https:\u002F\u002Famatriain.net\u002Fblog\u002FRecsysArchitectures) `Xavier Amatriain` `2022`\n76. [How Pinterest Leverages Realtime User Actions in Recommendation to Boost Homefeed Engagement Volume](https:\u002F\u002Fmedium.com\u002Fpinterest-engineering\u002Fhow-pinterest-leverages-realtime-user-actions-in-recommendation-to-boost-homefeed-engagement-volume-165ae2e8cde8) `Pinterest` `2022`\n77. [RecSysOps: Best Practices for Operating a Large-Scale Recommender System](https:\u002F\u002Fnetflixtechblog.medium.com\u002Frecsysops-best-practices-for-operating-a-large-scale-recommender-system-95bbe195a841) `Netflix` `2022`\n78. [Recommend API: Unified end-to-end machine learning infrastructure to generate recommendations](https:\u002F\u002Fslack.engineering\u002Frecommend-api\u002F) `Slack` `2022`\n79. [Evolving DoorDash’s Substitution Recommendations Algorithm](https:\u002F\u002Fdoordash.engineering\u002F2022\u002F09\u002F08\u002Fevolving-doordashs-substitution-recommendations-algorithm\u002F) `DoorDash` `2022`\n80. [Homepage Recommendation with Exploitation and Exploration](https:\u002F\u002Fdoordash.engineering\u002F2022\u002F10\u002F05\u002Fhomepage-recommendation-with-exploitation-and-exploration\u002F) `DoorDash` `2022`\n81. [GPU-accelerated ML Inference at Pinterest](https:\u002F\u002Fmedium.com\u002F@Pinterest_Engineering\u002Fgpu-accelerated-ml-inference-at-pinterest-ad1b6a03a16d) `Pinterest` `2022`\n82. [Addressing Confounding Feature Issue for Causal Recommendation](https:\u002F\u002Farxiv.org\u002Fabs\u002F2205.06532) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2205.06532.pdf)) `Tencent` `2022`\n\n\n## Search & Ranking\n1. [Amazon Search: The Joy of Ranking Products](https:\u002F\u002Fwww.amazon.science\u002Fpublications\u002Famazon-search-the-joy-of-ranking-products) ([Paper](https:\u002F\u002Fassets.amazon.science\u002F89\u002Fcd\u002F34289f1f4d25b5857d776bdf04d5\u002Famazon-search-the-joy-of-ranking-products.pdf), [Video](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=NLrhmn-EZ88), [Code](https:\u002F\u002Fgithub.com\u002Fdariasor\u002FTreeExtra)) `Amazon` `2016`\n2. [How Lazada Ranks Products to Improve Customer Experience and Conversion](https:\u002F\u002Fwww.slideshare.net\u002Feugeneyan\u002Fhow-lazada-ranks-products-to-improve-customer-experience-and-conversion) `Lazada` `2016`\n3. [Ranking Relevance in Yahoo Search](https:\u002F\u002Fwww.kdd.org\u002Fkdd2016\u002Fsubtopic\u002Fview\u002Franking-relevance-in-yahoo-search) ([Paper](https:\u002F\u002Fwww.kdd.org\u002Fkdd2016\u002Fpapers\u002Ffiles\u002Fadf0361-yinA.pdf)) `Yahoo` `2016`\n4. [Learning to Rank Personalized Search Results in Professional Networks](https:\u002F\u002Farxiv.org\u002Fabs\u002F1605.04624) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1605.04624.pdf)) `LinkedIn` `2016`\n5. [Using Deep Learning at Scale in Twitter’s Timelines](https:\u002F\u002Fblog.twitter.com\u002Fengineering\u002Fen_us\u002Ftopics\u002Finsights\u002F2017\u002Fusing-deep-learning-at-scale-in-twitters-timelines.html) `Twitter` `2017`\n6. [An Ensemble-based Approach to Click-Through Rate Prediction for Promoted Listings at Etsy](https:\u002F\u002Farxiv.org\u002Fabs\u002F1711.01377) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1711.01377.pdf)) `Etsy` `2017`\n7. [Powering Search & Recommendations at DoorDash](https:\u002F\u002Fdoordash.engineering\u002F2017\u002F07\u002F06\u002Fpowering-search-recommendations-at-doordash\u002F) `DoorDash` `2017`\n8. [Applying Deep Learning To Airbnb Search](https:\u002F\u002Farxiv.org\u002Fabs\u002F1810.09591) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1810.09591.pdf)) `Airbnb` `2018`\n9. [In-session Personalization for Talent Search](https:\u002F\u002Farxiv.org\u002Fabs\u002F1809.06488) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1809.06488.pdf)) `LinkedIn` `2018`\n10. [Talent Search and Recommendation Systems at LinkedIn](https:\u002F\u002Farxiv.org\u002Fabs\u002F1809.06481) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1809.06481.pdf)) `LinkedIn` `2018`\n11. [Food Discovery with Uber Eats: Building a Query Understanding Engine](https:\u002F\u002Feng.uber.com\u002Fuber-eats-query-understanding\u002F) `Uber` `2018`\n12. [Globally Optimized Mutual Influence Aware Ranking in E-Commerce Search](https:\u002F\u002Farxiv.org\u002Fabs\u002F1805.08524) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1805.08524.pdf)) `Alibaba` `2018`\n13. [Reinforcement Learning to Rank in E-Commerce Search Engine](https:\u002F\u002Farxiv.org\u002Fabs\u002F1803.00710) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1803.00710.pdf)) `Alibaba` `2018`\n14. [Semantic Product Search](https:\u002F\u002Farxiv.org\u002Fabs\u002F1907.00937) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1907.00937.pdf)) `Amazon` `2019`\n15. [Machine Learning-Powered Search Ranking of Airbnb Experiences](https:\u002F\u002Fmedium.com\u002Fairbnb-engineering\u002Fmachine-learning-powered-search-ranking-of-airbnb-experiences-110b4b1a0789) `Airbnb` `2019`\n16. [Entity Personalized Talent Search Models with Tree Interaction Features](https:\u002F\u002Farxiv.org\u002Fabs\u002F1902.09041) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1902.09041.pdf)) `LinkedIn` `2019`\n17. [The AI Behind LinkedIn Recruiter Search and recommendation systems](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2019\u002F04\u002Fai-behind-linkedin-recruiter-search-and-recommendation-systems) `LinkedIn` `2019`\n18. [Learning Hiring Preferences: The AI Behind LinkedIn Jobs](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2019\u002F02\u002Flearning-hiring-preferences--the-ai-behind-linkedin-jobs) `LinkedIn` `2019`\n19. [The Secret Sauce Behind Search Personalisation](https:\u002F\u002Fwww.gojek.io\u002Fblog\u002Fthe-secret-sauce-behind-search-personalisation) `Gojek` `2019`\n20. [Neural Code Search: ML-based Code Search Using Natural Language Queries](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fneural-code-search-ml-based-code-search-using-natural-language-queries\u002F) `Facebook` `2019`\n21. [Aggregating Search Results from Heterogeneous Sources via Reinforcement Learning](https:\u002F\u002Farxiv.org\u002Fabs\u002F1902.08882) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1902.08882.pdf)) `Alibaba` `2019`\n22. [Cross-domain Attention Network with Wasserstein Regularizers for E-commerce Search](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3357384.3357809) `Alibaba` `2019`\n23. [Understanding Searches Better Than Ever Before](https:\u002F\u002Fwww.blog.google\u002Fproducts\u002Fsearch\u002Fsearch-language-understanding-bert\u002F) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1810.04805.pdf)) `Google` `2019`\n24. [How We Used Semantic Search to Make Our Search 10x Smarter](https:\u002F\u002Fmedium.com\u002Ftokopedia-engineering\u002Fhow-we-used-semantic-search-to-make-our-search-10x-smarter-bd9c7f601821) `Tokopedia` `2019`\n25. [Query2vec: Search query expansion with query embeddings](https:\u002F\u002Fbytes.grubhub.com\u002Fsearch-query-embeddings-using-query2vec-f5931df27d79) `GrubHub` `2019`\n26. [MOBIUS: Towards the Next Generation of Query-Ad Matching in Baidu’s Sponsored Search](http:\u002F\u002Fresearch.baidu.com\u002FPublic\u002Fuploads\u002F5d12eca098d40.pdf) `Baidu` `2019`\n27. [Why Do People Buy Seemingly Irrelevant Items in Voice Product Search?](https:\u002F\u002Fwww.amazon.science\u002Fpublications\u002Fwhy-do-people-buy-irrelevant-items-in-voice-product-search) ([Paper](https:\u002F\u002Fassets.amazon.science\u002Ff7\u002F48\u002F0562b2c14338a0b76ccf4f523fa5\u002Fwhy-do-people-buy-irrelevant-items-in-voice-product-search.pdf)) `Amazon` `2020`\n28. [Managing Diversity in Airbnb Search](https:\u002F\u002Farxiv.org\u002Fabs\u002F2004.02621) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2004.02621.pdf)) `Airbnb` `2020`\n29. [Improving Deep Learning for Airbnb Search](https:\u002F\u002Farxiv.org\u002Fabs\u002F2002.05515) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2002.05515.pdf)) `Airbnb` `2020`\n30. [Quality Matches Via Personalized AI for Hirer and Seeker Preferences](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2020\u002Fquality-matches-via-personalized-ai) `LinkedIn` `2020`\n31. [Understanding Dwell Time to Improve LinkedIn Feed Ranking](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2020\u002Funderstanding-feed-dwell-time) `LinkedIn` `2020`\n32. [Ads Allocation in Feed via Constrained Optimization](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fabs\u002F10.1145\u002F3394486.3403391) ([Paper](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fpdf\u002F10.1145\u002F3394486.3403391), [Video](https:\u002F\u002Fcrossminds.ai\u002Fvideo\u002F5f33697a0576dd25aef288ea\u002F)) `LinkedIn` `2020`\n33. [Understanding Dwell Time to Improve LinkedIn Feed Ranking](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2020\u002Funderstanding-feed-dwell-time) `LinkedIn` `2020`\n34. [AI at Scale in Bing](https:\u002F\u002Fblogs.bing.com\u002Fsearch\u002F2020_05\u002FAI-at-Scale-in-Bing) `Microsoft` `2020`\n35. [Query Understanding Engine in Traveloka Universal Search](https:\u002F\u002Fmedium.com\u002Ftraveloka-engineering\u002Fquery-understanding-engine-in-traveloka-universal-search-410ad3895db7) `Traveloka` `2020`\n36. [Bayesian Product Ranking at Wayfair](https:\u002F\u002Ftech.wayfair.com\u002Fdata-science\u002F2020\u002F01\u002Fbayesian-product-ranking-at-wayfair) `Wayfair` `2020`\n37. [COLD: Towards the Next Generation of Pre-Ranking System](https:\u002F\u002Farxiv.org\u002Fabs\u002F2007.16122) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2007.16122.pdf)) `Alibaba` `2020`\n38. [Shop The Look: Building a Large Scale Visual Shopping System at Pinterest](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fabs\u002F10.1145\u002F3394486.3403372) ([Paper](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fpdf\u002F10.1145\u002F3394486.3403372), [Video](https:\u002F\u002Fcrossminds.ai\u002Fvideo\u002F5f3369790576dd25aef288d7\u002F)) `Pinterest` `2020`\n39. [Driving Shopping Upsells from Pinterest Search](https:\u002F\u002Fmedium.com\u002Fpinterest-engineering\u002Fdriving-shopping-upsells-from-pinterest-search-d06329255402) `Pinterest` `2020`\n40. [GDMix: A Deep Ranking Personalization Framework](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2020\u002Fgdmix--a-deep-ranking-personalization-framework) ([Code](https:\u002F\u002Fgithub.com\u002Flinkedin\u002Fgdmix)) `LinkedIn` `2020`\n41. [Bringing Personalized Search to Etsy](https:\u002F\u002Fcodeascraft.com\u002F2020\u002F10\u002F29\u002Fbringing-personalized-search-to-etsy\u002F) `Etsy` `2020`\n42. [Building a Better Search Engine for Semantic Scholar](https:\u002F\u002Fmedium.com\u002Fai2-blog\u002Fbuilding-a-better-search-engine-for-semantic-scholar-ea23a0b661e7) `Allen Institute for AI` `2020`\n43. [Query Understanding for Natural Language Enterprise Search](https:\u002F\u002Farxiv.org\u002Fabs\u002F2012.06238) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2012.06238.pdf)) `Salesforce` `2020`\n44. [Things Not Strings: Understanding Search Intent with Better Recall](https:\u002F\u002Fdoordash.engineering\u002F2020\u002F12\u002F15\u002Funderstanding-search-intent-with-better-recall\u002F) `DoorDash` `2020`\n45. [Query Understanding for Surfacing Under-served Music Content](https:\u002F\u002Fresearch.atspotify.com\u002Fpublications\u002Fquery-understanding-for-surfacing-under-served-music-content\u002F) ([Paper](https:\u002F\u002Flabtomarket.files.wordpress.com\u002F2020\u002F08\u002Fcikm2020.pdf)) `Spotify` `2020`\n46. [Embedding-based Retrieval in Facebook Search](https:\u002F\u002Farxiv.org\u002Fabs\u002F2006.11632) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2006.11632.pdf)) `Facebook` `2020`\n47. [Towards Personalized and Semantic Retrieval for E-commerce Search via Embedding Learning](https:\u002F\u002Farxiv.org\u002Fabs\u002F2006.02282) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2006.02282.pdf)) `JD` `2020`\n48. [QUEEN: Neural query rewriting in e-commerce](https:\u002F\u002Fwww.amazon.science\u002Fpublications\u002Fqueen-neural-query-rewriting-in-e-commerce) ([Paper](https:\u002F\u002Fassets.amazon.science\u002Ff9\u002F78\u002Fdda8f1e143dba8ca96e43ec487c6\u002Fqueen-neural-query-rewriting-in-ecommerce.pdf)) `Amazon` `2021`\n49. [Using Learning-to-rank to Precisely Locate Where to Deliver Packages](https:\u002F\u002Fwww.amazon.science\u002Fblog\u002Fusing-learning-to-rank-to-precisely-locate-where-to-deliver-packages) ([Paper](https:\u002F\u002Fwww.amazon.science\u002Fpublications\u002Fgetting-your-package-to-the-right-place-supervised-machine-learning-for-geolocation)) `Amazon` `2021`\n50. [Seasonal relevance in e-commerce search](https:\u002F\u002Fwww.amazon.science\u002Fpublications\u002Fseasonal-relevance-in-e-commerce-search) ([Paper](https:\u002F\u002Fassets.amazon.science\u002Fac\u002F5e\u002Fd47612a846d6bec15738d7c8ab40\u002Fseasonal-relevance-in-ecommerce-search.pdf)) `Amazon` `2021`\n51. [Graph Intention Network for Click-through Rate Prediction in Sponsored Search](https:\u002F\u002Farxiv.org\u002Fabs\u002F2103.16164) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2103.16164.pdf)) `Alibaba` `2021`\n52. [How We Built A Context-Specific Bidding System for Etsy Ads](https:\u002F\u002Fcodeascraft.com\u002F2021\u002F03\u002F23\u002Fhow-we-built-a-context-specific-bidding-system-for-etsy-ads\u002F) `Etsy` `2021`\n53. [Pre-trained Language Model based Ranking in Baidu Search](https:\u002F\u002Farxiv.org\u002Fabs\u002F2105.11108) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2105.11108.pdf)) `Baidu` `2021`\n54. [Stitching together spaces for query-based recommendations](https:\u002F\u002Fmultithreaded.stitchfix.com\u002Fblog\u002F2021\u002F08\u002F13\u002Fstitching-together-spaces-for-query-based-recommendations\u002F) `Stitch Fix` `2021`\n55. [Deep Natural Language Processing for LinkedIn Search Systems](https:\u002F\u002Farxiv.org\u002Fabs\u002F2108.08252) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2108.08252.pdf)) `LinkedIn` `2021`\n56. [Siamese BERT-based Model for Web Search Relevance Ranking](https:\u002F\u002Farxiv.org\u002Fabs\u002F2112.01810) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2112.01810.pdf), [Code](https:\u002F\u002Fgithub.com\u002Fseznam\u002FDaReCzech)) `Seznam` `2021`\n57. [SearchSage: Learning Search Query Representations at Pinterest](https:\u002F\u002Fmedium.com\u002Fpinterest-engineering\u002Fsearchsage-learning-search-query-representations-at-pinterest-654f2bb887fc) `Pinterest` `2021`\n58. [Query2Prod2Vec: Grounded Word Embeddings for eCommerce](https:\u002F\u002Faclanthology.org\u002F2021.naacl-industry.20\u002F) `Coveo` `2021`\n59. [3 Changes to Expand DoorDash’s Product Search Beyond Delivery](https:\u002F\u002Fdoordash.engineering\u002F2022\u002F05\u002F10\u002F3-changes-to-expand-doordashs-product-search\u002F) `DoorDash` `2022`\n60. [Learning To Rank Diversely](https:\u002F\u002Fmedium.com\u002Fairbnb-engineering\u002Flearning-to-rank-diversely-add6b1929621) `Airbnb` `2022`\n61. [How to Optimise Rankings with Cascade Bandits](https:\u002F\u002Fmedium.com\u002Fexpedia-group-tech\u002Fhow-to-optimise-rankings-with-cascade-bandits-5d92dfa0f16b) `Expedia` `2022`\n62. [A Guide to Google Search Ranking Systems](https:\u002F\u002Fdevelopers.google.com\u002Fsearch\u002Fdocs\u002Fappearance\u002Franking-systems-guide) `Google` `2022` \n63. [Deep Learning for Search Ranking at Etsy](https:\u002F\u002Fwww.etsy.com\u002Fcodeascraft\u002Fdeep-learning-for-search-ranking-at-etsy) `Etsy` `2022`\n64. [Search at Calm](https:\u002F\u002Feng.calm.com\u002Fposts\u002Fsearch-at-calm) `Calm` `2022`\n\n## Embeddings\n1. [Vector Representation Of Items, Customer And Cart To Build A Recommendation System](https:\u002F\u002Farxiv.org\u002Fabs\u002F1705.06338) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1705.06338.pdf)) `Sears` `2017`\n2. [Billion-scale Commodity Embedding for E-commerce Recommendation in Alibaba](https:\u002F\u002Farxiv.org\u002Fabs\u002F1803.02349) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1803.02349.pdf)) `Alibaba` `2018`\n3. [Embeddings@Twitter](https:\u002F\u002Fblog.twitter.com\u002Fengineering\u002Fen_us\u002Ftopics\u002Finsights\u002F2018\u002Fembeddingsattwitter.html) `Twitter` `2018`\n4. [Listing Embeddings in Search Ranking](https:\u002F\u002Fmedium.com\u002Fairbnb-engineering\u002Flisting-embeddings-for-similar-listing-recommendations-and-real-time-personalization-in-search-601172f7603e) ([Paper](https:\u002F\u002Fwww.kdd.org\u002Fkdd2018\u002Faccepted-papers\u002Fview\u002Freal-time-personalization-using-embeddings-for-search-ranking-at-airbnb)) `Airbnb` `2018`\n5. [Understanding Latent Style](https:\u002F\u002Fmultithreaded.stitchfix.com\u002Fblog\u002F2018\u002F06\u002F28\u002Flatent-style\u002F) `Stitch Fix` `2018`\n6. [Towards Deep and Representation Learning for Talent Search at LinkedIn](https:\u002F\u002Farxiv.org\u002Fabs\u002F1809.06473) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1809.06473.pdf)) `LinkedIn` `2018`\n7. [Personalized Store Feed with Vector Embeddings](https:\u002F\u002Fdoordash.engineering\u002F2018\u002F04\u002F02\u002Fpersonalized-store-feed-with-vector-embeddings\u002F) `DoorDash` `2018`\n8. [Should we Embed? A Study on Performance of Embeddings for Real-Time Recommendations](https:\u002F\u002Farxiv.org\u002Fabs\u002F1907.06556)([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1907.06556.pdf)) `Moshbit` `2019`\n9. [Machine Learning for a Better Developer Experience](https:\u002F\u002Fnetflixtechblog.com\u002Fmachine-learning-for-a-better-developer-experience-1e600c69f36c) `Netflix` `2020`\n10. [Announcing ScaNN: Efficient Vector Similarity Search](https:\u002F\u002Fai.googleblog.com\u002F2020\u002F07\u002Fannouncing-scann-efficient-vector.html) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1908.10396.pdf), [Code](https:\u002F\u002Fgithub.com\u002Fgoogle-research\u002Fgoogle-research\u002Ftree\u002Fmaster\u002Fscann)) `Google` `2020`\n11. [BERT Goes Shopping: Comparing Distributional Models for Product Representations](https:\u002F\u002Faclanthology.org\u002F2021.ecnlp-1.1\u002F) `Coveo` `2021`\n12. [The Embeddings That Came in From the Cold: Improving Vectors for New and Rare Products with Content-Based Inference](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3383313.3411477) `Coveo` `2022`\n13. [Embedding-based Retrieval at Scribd](https:\u002F\u002Ftech.scribd.com\u002Fblog\u002F2021\u002Fembedding-based-retrieval-scribd.html) `Scribd` `2021`\n14. [Multi-objective Hyper-parameter Optimization of Behavioral Song Embeddings](https:\u002F\u002Farxiv.org\u002Fabs\u002F2208.12724) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2208.12724.pdf)) `Apple` `2022`\n15. [Embeddings at Spotify's Scale - How Hard Could It Be?](https:\u002F\u002Farize.com\u002Fresource\u002Fembeddings-at-scale-spotify-recsys\u002F) `Spotify` `2023`\n\n## Natural Language Processing\n1. [Abusive Language Detection in Online User Content](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F2872427.2883062) ([Paper](http:\u002F\u002Fwww.yichang-cs.com\u002Fyahoo\u002FWWW16_Abusivedetection.pdf)) `Yahoo` `2016`\n2. [Smart Reply: Automated Response Suggestion for Email](https:\u002F\u002Fresearch.google\u002Fpubs\u002Fpub45189\u002F) ([Paper](https:\u002F\u002Fstorage.googleapis.com\u002Fpub-tools-public-publication-data\u002Fpdf\u002F45189.pdf)) `Google` `2016` \n3. [Building Smart Replies for Member Messages](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2017\u002F10\u002Fbuilding-smart-replies-for-member-messages) `LinkedIn` `2017`\n4. [How Natural Language Processing Helps LinkedIn Members Get Support Easily](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2019\u002F04\u002Fhow-natural-language-processing-help-support) `LinkedIn` `2019`\n5. [Gmail Smart Compose: Real-Time Assisted Writing](https:\u002F\u002Farxiv.org\u002Fabs\u002F1906.00080) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1906.00080.pdf)) `Google` `2019`\n6. [Goal-Oriented End-to-End Conversational Models with Profile Features in a Real-World Setting](https:\u002F\u002Fwww.amazon.science\u002Fpublications\u002Fgoal-oriented-end-to-end-chatbots-with-profile-features-in-a-real-world-setting) ([Paper](https:\u002F\u002Fassets.amazon.science\u002F47\u002F03\u002Fe0d14dc34d3eb6e0d4ec282067bd\u002Fgoal-oriented-end-to-end-chatbots-with-profile-features-in-a-real-world-setting.pdf)) `Amazon` `2019`\n7. [Give Me Jeans not Shoes: How BERT Helps Us Deliver What Clients Want](https:\u002F\u002Fmultithreaded.stitchfix.com\u002Fblog\u002F2019\u002F07\u002F15\u002Fgive-me-jeans\u002F) `Stitch Fix` `2019`\n8. [DeText: A deep NLP Framework for Intelligent Text Understanding](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2020\u002Fopen-sourcing-detext) ([Code](https:\u002F\u002Fgithub.com\u002Flinkedin\u002Fdetext)) `LinkedIn` `2020`\n9. [SmartReply for YouTube Creators](https:\u002F\u002Fai.googleblog.com\u002F2020\u002F07\u002Fsmartreply-for-youtube-creators.html) `Google` `2020`\n10. [Using Neural Networks to Find Answers in Tables](https:\u002F\u002Fai.googleblog.com\u002F2020\u002F04\u002Fusing-neural-networks-to-find-answers.html) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2004.02349.pdf)) `Google` `2020`\n11. [A Scalable Approach to Reducing Gender Bias in Google Translate](https:\u002F\u002Fai.googleblog.com\u002F2020\u002F04\u002Fa-scalable-approach-to-reducing-gender.html) `Google` `2020`\n12. [Assistive AI Makes Replying Easier](https:\u002F\u002Fwww.microsoft.com\u002Fen-us\u002Fresearch\u002Fgroup\u002Fmsai\u002Farticles\u002Fassistive-ai-makes-replying-easier-2\u002F) `Microsoft` `2020`\n13. [AI Advances to Better Detect Hate Speech](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fai-advances-to-better-detect-hate-speech\u002F) `Facebook` `2020`\n14. [A State-of-the-Art Open Source Chatbot](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fstate-of-the-art-open-source-chatbot) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2004.13637.pdf)) `Facebook` `2020`\n15. [A Highly Efficient, Real-Time Text-to-Speech System Deployed on CPUs](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fa-highly-efficient-real-time-text-to-speech-system-deployed-on-cpus\u002F) `Facebook` `2020`\n16. [Deep Learning to Translate Between Programming Languages](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fdeep-learning-to-translate-between-programming-languages\u002F) ([Paper](https:\u002F\u002Farxiv.org\u002Fabs\u002F2006.03511), [Code](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FTransCoder)) `Facebook` `2020`\n17. [Deploying Lifelong Open-Domain Dialogue Learning](https:\u002F\u002Farxiv.org\u002Fabs\u002F2008.08076) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2008.08076.pdf)) `Facebook` `2020`\n18. [Introducing Dynabench: Rethinking the way we benchmark AI](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fdynabench-rethinking-ai-benchmarking\u002F) `Facebook` `2020`\n19. [How Gojek Uses NLP to Name Pickup Locations at Scale](https:\u002F\u002Fwww.gojek.io\u002Fblog\u002Fnlp-cartobert) `Gojek` `2020`\n20. [The State-of-the-art Open-Domain Chatbot in Chinese and English](http:\u002F\u002Fresearch.baidu.com\u002FBlog\u002Findex-view?id=142) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2006.16779.pdf)) `Baidu` `2020`\n21. [PEGASUS: A State-of-the-Art Model for Abstractive Text Summarization](https:\u002F\u002Fai.googleblog.com\u002F2020\u002F06\u002Fpegasus-state-of-art-model-for.html) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1912.08777.pdf), [Code](https:\u002F\u002Fgithub.com\u002Fgoogle-research\u002Fpegasus)) `Google` `2020`\n22. [Photon: A Robust Cross-Domain Text-to-SQL System](https:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002F2020.acl-demos.24\u002F) ([Paper](https:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002F2020.acl-demos.24.pdf)) ([Demo](http:\u002F\u002Fnaturalsql.com)) `Salesforce`\t`2020`\n23. [GeDi: A Powerful New Method for Controlling Language Models](https:\u002F\u002Fblog.einstein.ai\u002Fgedi\u002F) ([Paper](https:\u002F\u002Farxiv.org\u002Fabs\u002F2009.06367), [Code](https:\u002F\u002Fgithub.com\u002Fsalesforce\u002FGeDi)) `Salesforce` `2020`\n24. [Applying Topic Modeling to Improve Call Center Operations](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=kzRR8OjF_eI&t=2s) `RICOH` `2020`\n25. [WIDeText: A Multimodal Deep Learning Framework](https:\u002F\u002Fmedium.com\u002Fairbnb-engineering\u002Fwidetext-a-multimodal-deep-learning-framework-31ce2565880c) `Airbnb` `2020`\n26. [Dynaboard: Moving Beyond Accuracy to Holistic Model Evaluation in NLP](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fdynaboard-moving-beyond-accuracy-to-holistic-model-evaluation-in-nlp) ([Code](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fdynalab?fbclid=IwAR3qcV7QK2uXm4s4M0XUoQQo4i2DEsDy0LZFKxSQCHhP-3hF6fr2-NDFWX8)) `Facebook`  `2021`\n27. [How we reduced our text similarity runtime by 99.96%](https:\u002F\u002Fmedium.com\u002Fdata-science-at-microsoft\u002Fhow-we-reduced-our-text-similarity-runtime-by-99-96-e8e4b4426b35) `Microsoft` `2021`\n28. [Textless NLP: Generating expressive speech from raw audio](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Ftextless-nlp-generating-expressive-speech-from-raw-audio\u002F) [(Part 1)](https:\u002F\u002Farxiv.org\u002Fabs\u002F2102.01192) [(Part 2)](https:\u002F\u002Farxiv.org\u002Fabs\u002F2104.00355) [(Part 3)](https:\u002F\u002Farxiv.org\u002Fabs\u002F2109.03264) [(Code and Pretrained Models)](https:\u002F\u002Fgithub.com\u002Fpytorch\u002Ffairseq\u002Ftree\u002Fmaster\u002Fexamples\u002Ftextless_nlp) `Facebook` `2021`\n29. [Grammar Correction as You Type, on Pixel 6](https:\u002F\u002Fai.googleblog.com\u002F2021\u002F10\u002Fgrammar-correction-as-you-type-on-pixel.html) `Google` `2021`\n30. [Auto-generated Summaries in Google Docs](https:\u002F\u002Fai.googleblog.com\u002F2022\u002F03\u002Fauto-generated-summaries-in-google-docs.html) `Google` `2022`\n31. [ML-Enhanced Code Completion Improves Developer Productivity](https:\u002F\u002Fai.googleblog.com\u002F2022\u002F07\u002Fml-enhanced-code-completion-improves.html) `Google` `2022`\n32. [Words All the Way Down — Conversational Sentiment Analysis](https:\u002F\u002Fmedium.com\u002Fpaypal-tech\u002Fwords-all-the-way-down-conversational-sentiment-analysis-afe0165b84db) `PayPal` `2022`\n\n## Sequence Modelling\n1. [Doctor AI: Predicting Clinical Events via Recurrent Neural Networks](https:\u002F\u002Farxiv.org\u002Fabs\u002F1511.05942) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1511.05942.pdf)) `Sutter Health` `2015`\n2. [Deep Learning for Understanding Consumer Histories](https:\u002F\u002Fengineering.zalando.com\u002Fposts\u002F2016\u002F10\u002Fdeep-learning-for-understanding-consumer-histories.html) ([Paper](https:\u002F\u002Fdoogkong.github.io\u002F2017\u002Fpapers\u002Fpaper2.pdf)) `Zalando` `2016`\n3. [Using Recurrent Neural Network Models for Early Detection of Heart Failure Onset](https:\u002F\u002Fwww.ncbi.nlm.nih.gov\u002Fpmc\u002Farticles\u002FPMC5391725\u002F) ([Paper](https:\u002F\u002Fwww.ncbi.nlm.nih.gov\u002Fpmc\u002Farticles\u002FPMC5391725\u002Fpdf\u002Focw112.pdf)) `Sutter Health` `2016`\n4. [Continual Prediction of Notification Attendance with Classical and Deep Networks](https:\u002F\u002Farxiv.org\u002Fabs\u002F1712.07120) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1712.07120.pdf)) `Telefonica` `2017` \n5. [Deep Learning for Electronic Health Records](https:\u002F\u002Fai.googleblog.com\u002F2018\u002F05\u002Fdeep-learning-for-electronic-health.html) ([Paper](https:\u002F\u002Fwww.nature.com\u002Farticles\u002Fs41746-018-0029-1.pdf)) `Google` `2018`\n6. [Practice on Long Sequential User Behavior Modeling for Click-Through Rate Prediction](https:\u002F\u002Farxiv.org\u002Fabs\u002F1905.09248) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1905.09248.pdf))`Alibaba` `2019`\n7. [Search-based User Interest Modeling with Sequential Behavior Data for CTR Prediction](https:\u002F\u002Farxiv.org\u002Fabs\u002F2006.05639) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2006.05639.pdf)) `Alibaba` `2020`\n8. [How Duolingo uses AI in every part of its app](https:\u002F\u002Fventurebeat.com\u002F2020\u002F08\u002F18\u002Fhow-duolingo-uses-ai-in-every-part-of-its-app\u002F) `Duolingo` `2020`\n9. [Leveraging Online Social Interactions For Enhancing Integrity at Facebook](https:\u002F\u002Fresearch.fb.com\u002Fblog\u002F2020\u002F08\u002Fleveraging-online-social-interactions-for-enhancing-integrity-at-facebook\u002F) ([Paper](https:\u002F\u002Fresearch.fb.com\u002Fwp-content\u002Fuploads\u002F2020\u002F08\u002FTIES-Temporal-Interaction-Embeddings-For-Enhancing-Social-Media-Integrity-At-Facebook.pdf), [Video](https:\u002F\u002Fcrossminds.ai\u002Fvideo\u002F5f3369780576dd25aef288cf\u002F)) `Facebook` `2020`\n10. [Using deep learning to detect abusive sequences of member activity](https:\u002F\u002Fengineering.linkedin.com\u002Fblog\u002F2021\u002Fusing-deep-learning-to-detect-abusive-sequences-of-member-activi) ([Video](https:\u002F\u002Fexchange.scale.com\u002Fpublic\u002Fvideos\u002Fusing-deep-learning-to-detect-abusive-sequences-of-member-activity-on-linkedin)) `LinkedIn` `2021`\n\n## Computer Vision\n1. [Creating a Modern OCR Pipeline Using Computer Vision and Deep Learning](https:\u002F\u002Fdropbox.tech\u002Fmachine-learning\u002Fcreating-a-modern-ocr-pipeline-using-computer-vision-and-deep-learning) `Dropbox` `2017`\n2. [Categorizing Listing Photos at Airbnb](https:\u002F\u002Fmedium.com\u002Fairbnb-engineering\u002Fcategorizing-listing-photos-at-airbnb-f9483f3ab7e3) `Airbnb` `2018`\n3. [Amenity Detection and Beyond — New Frontiers of Computer Vision at Airbnb](https:\u002F\u002Fmedium.com\u002Fairbnb-engineering\u002Famenity-detection-and-beyond-new-frontiers-of-computer-vision-at-airbnb-144a4441b72e) `Airbnb` `2019`\n4. [How we Improved Computer Vision Metrics by More Than 5% Only by Cleaning Labelling Errors](https:\u002F\u002Fdeepomatic.com\u002Fen\u002Fhow-we-improved-computer-vision-metrics-by-more-than-5-percent-only-by-cleaning-labelling-errors\u002F) `Deepomatic`\n5. [Making machines recognize and transcribe conversations in meetings using audio and video](https:\u002F\u002Fwww.microsoft.com\u002Fen-us\u002Fresearch\u002Fblog\u002Fmaking-machines-recognize-and-transcribe-conversations-in-meetings-using-audio-and-video\u002F) `Microsoft` `2019`\n6. [Powered by AI: Advancing product understanding and building new shopping experiences](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fpowered-by-ai-advancing-product-understanding-and-building-new-shopping-experiences\u002F) `Facebook` `2020`\n7. [A Neural Weather Model for Eight-Hour Precipitation Forecasting](https:\u002F\u002Fai.googleblog.com\u002F2020\u002F03\u002Fa-neural-weather-model-for-eight-hour.html) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2003.12140.pdf)) `Google` `2020`\n8. [Machine Learning-based Damage Assessment for Disaster Relief](https:\u002F\u002Fai.googleblog.com\u002F2020\u002F06\u002Fmachine-learning-based-damage.html) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1910.06444.pdf)) `Google` `2020`\n9. [RepNet: Counting Repetitions in Videos](https:\u002F\u002Fai.googleblog.com\u002F2020\u002F06\u002Frepnet-counting-repetitions-in-videos.html) ([Paper](https:\u002F\u002Fopenaccess.thecvf.com\u002Fcontent_CVPR_2020\u002Fpapers\u002FDwibedi_Counting_Out_Time_Class_Agnostic_Video_Repetition_Counting_in_the_CVPR_2020_paper.pdf)) `Google` `2020`\n10. [Converting Text to Images for Product Discovery](https:\u002F\u002Fwww.amazon.science\u002Fblog\u002Fconverting-text-to-images-for-product-discovery) ([Paper](https:\u002F\u002Fassets.amazon.science\u002F4c\u002F76\u002F5830542547b7a11089ce3af943b4\u002Fscipub-972.pdf)) `Amazon` `2020`\n11. [How Disney Uses PyTorch for Animated Character Recognition](https:\u002F\u002Fmedium.com\u002Fpytorch\u002Fhow-disney-uses-pytorch-for-animated-character-recognition-a1722a182627) `Disney` `2020`\n12. [Image Captioning as an Assistive Technology](https:\u002F\u002Fwww.ibm.com\u002Fblogs\u002Fresearch\u002F2020\u002F07\u002Fimage-captioning-assistive-technology\u002F) ([Video](https:\u002F\u002Fivc.ischool.utexas.edu\u002F~yz9244\u002FVizWiz_workshop\u002Fvideos\u002FMMTeam-oral.mp4)) `IBM` `2020`\n13. [AI for AG: Production machine learning for agriculture](https:\u002F\u002Fmedium.com\u002Fpytorch\u002Fai-for-ag-production-machine-learning-for-agriculture-e8cfdb9849a1) `Blue River` `2020`\n14. [AI for Full-Self Driving at Tesla](https:\u002F\u002Fyoutu.be\u002Fhx7BXih7zx8?t=513) `Tesla` `2020`\n15. [On-device Supermarket Product Recognition](https:\u002F\u002Fai.googleblog.com\u002F2020\u002F07\u002Fon-device-supermarket-product.html) `Google` `2020`\n16. [Using Machine Learning to Detect Deficient Coverage in Colonoscopy Screenings](https:\u002F\u002Fai.googleblog.com\u002F2020\u002F08\u002Fusing-machine-learning-to-detect.html) ([Paper](https:\u002F\u002Fieeexplore.ieee.org\u002Fstamp\u002Fstamp.jsp?tp=&arnumber=9097918)) `Google` `2020`\n17. [Shop The Look: Building a Large Scale Visual Shopping System at Pinterest](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fabs\u002F10.1145\u002F3394486.3403372) ([Paper](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fpdf\u002F10.1145\u002F3394486.3403372), [Video](https:\u002F\u002Fcrossminds.ai\u002Fvideo\u002F5f3369790576dd25aef288d7\u002F)) `Pinterest` `2020`\n18. [Developing Real-Time, Automatic Sign Language Detection for Video Conferencing](https:\u002F\u002Fai.googleblog.com\u002F2020\u002F10\u002Fdeveloping-real-time-automatic-sign.html) ([Paper](https:\u002F\u002Fstorage.googleapis.com\u002Fpub-tools-public-publication-data\u002Fpdf\u002F2eaf0d18ec6bef00d7dd88f39dd4f9ff13eeeeb2.pdf)) `Google` `2020`\n19. [Vision-based Price Suggestion for Online Second-hand Items](https:\u002F\u002Farxiv.org\u002Fabs\u002F2012.06009) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2012.06009.pdf)) `Alibaba` `2020`\n20. [New AI Research to Help Predict COVID-19 Resource Needs From X-rays](https:\u002F\u002Fai.facebook.com\u002Fblog\u002Fnew-ai-research-to-help-predict-covid-19-resource-needs-from-a-series-of-x-rays\u002F) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2101.04909.pdf), [Model](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FCovidPrognosis)) `Facebook` `2021`\n21. [An Efficient Training Approach for Very Large Scale Face Recognition](https:\u002F\u002Farxiv.org\u002Fabs\u002F2105.10375) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2105.10375)) `Alibaba` `2021`\n22. [Identifying Document Types at Scribd](https:\u002F\u002Ftech.scribd.com\u002Fblog\u002F2021\u002Fidentifying-document-types.html) `Scribd` `2021`\n23. [Semi-Supervised Visual Representation Learning for Fashion Compatibility](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2109.08052.pdf) ([Paper](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2109.08052.pdf)) `Walmart` `2021`\n24. [Recognizing People in Photos Through Private On-Device Machine Learning](https:\u002F\u002Fmachinelearning.apple.com\u002Fresearch\u002Frecognizing-people-photos) `Apple` `2021`\n25. [DeepFusion: Lidar-Camera Deep Fusion for Multi-Modal 3D Object Detection](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2203.08195.pdf) `Google` `2022`\n26. [Contrastive language and vision learning of general fashion concepts](https:\u002F\u002Fwww.nature.com\u002Farticles\u002Fs41598-022-23052-9) ([Paper](https:\u002F\u002Fwww.nature.com\u002Farticles\u002Fs41598-022-23052-9.pdf))`Coveo` `2022`\n27. [Leveraging Computer Vision for Search Ranking](https:\u002F\u002Farize.com\u002Fresource\u002Fbazaarvoice-leveraging-computer-vision-models-for-search-ranking\u002F) `BazaarVoice` `2023`\n\n## Reinforcement Learning\n1. [Deep Reinfo","该项目是一个汇集了众多公司关于数据科学和机器学习在实际生产环境中应用的论文和技术博客的资源库。它涵盖了从数据质量、工程到具体应用场景如推荐系统、自然语言处理等广泛的主题，旨在通过分享成功的案例与失败的经验教训来帮助开发者更好地理解如何将机器学习技术应用于实际问题中。特别适合正在规划或实施机器学习项目的团队参考，以了解行业内的最佳实践以及潜在挑战。",2,"2026-06-11 03:23:25","top_topic"]