050919_How_To_Build_A_Recommender_System

Transcript

1. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark How to build a Recommender System with Amazon Personalize Marc Cabocel – Startup Solutions Architect 05/09/19
2. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark FRAMEWORKS INTERFACES INFRASTRUCTURE AI Services Broadest and deepest set of capabilities T H E A W S M L ST A C K VISION SPEECH LANGUAGE CHATBOTS FORECASTING RECOMMENDATIONS ML Services ML Frameworks + Infrastructure P O L L Y T R A N S C R I B E T R A N S L A T E C O M P R E H E N D L E X F O R E C A S T R E K O G N I T I O N I M A G E R E K O G N I T I O N V I D E O T E X T R A C T P E R S O N A L I Z E Ground Truth Notebooks Algorithms + Marketplace Reinforcement Learning Training Optimization Deployment Hosting Amazon SageMaker F P G A S E C 2 P 3 & P 3 D N E C 2 G 4 E C 2 C 5 I N F E R E N T I A G R E E N G R A S S E L A S T I C I N F E R E N C E
3. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark AI Services Pre-trained AI services that require no ML skills or training Easily add intelligence to your existing apps and workflows Quality and accuracy from continuously-learning APIs VISION SPEECH LANGUAGE CHATBOTS FORECASTING RECOMMENDATIONS P O L L Y T R A N S C R I B E T R A N S L A T E C O M P R E H E N D L E X F O R E C A S T R E K O G N I T I O N I M A G E R E K O G N I T I O N V I D E O T E X T R A C T P E R S O N A L I Z E
4. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Personalized recommendations Personalized reranking Notifications and emails Related Items
5. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Engagement (up to 15% increase) Product Discovery (up to 80% clicks on tail items) Revenue (up to 5% increase) Conversion (up to 30% increase)
6. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved. S U M M I T Personalizing user experience is proven to increase discoverability, engagement, user satisfaction, and revenue 30% of page views on Amazon are from recommendations
7. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Popularity Trap Naïve models give recommendations similar to popular items Cold Starts New users should get relevant recommendations, new items should show in recommendations Scale Recommendations should scale across millions of users and items Real-Time Personalization must be responsive to the changing user intent Custom models Personalization models must accurately reflect business context and user behavior
8. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Articles, products, videos, etc. Age, location, etc. Amazon Personalize Customized personalization & recommendation API Views, signups, conversion, etc.
9. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Customized personalization & recommendation API F u l l y m a n a g e d b y A m a z o n P e r s o n a l i z e Amazon Personalize INSPECT DATA IDENTIFY FEATURES SELECT ALGORITHMS SELECT HYPERPARAMETERS TRAIN MODELS OPTIMIZE MODELS HOST MODELS BUILD FEATURE STORE CREATE REAL-TIME CACHES User data: Age , location (optional) Item data: Articles, products, videos, etc.(optional) Interactions/Event: Click, Views, signups, conversion, etc.
10. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Amazon Personalize Concepts
11. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark
12. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Use the Console/API to train and experiment with models Use AutoML or pick a pre-defined algorithm • Choose a pre-existing algorithm (packaged as Personalize Recipes) or use AutoML and Personalize will pick the right recipe for you • You can train custom deep learning models on your data and compare accuracy metrics with 2 API calls
13. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Deploy the best models (solutions) by launching a campaign Amazon Personalize Campaigns • Launching a campaign will deploy all the infrastructure needed to create a personalize endpoint • Personalize will automatically scale to serve you real time traffic • Use the getRecommendatons() or getPersonalizeRanking() API for inference
14. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Hands-on https://github.com/aws-samples/amazon-personalize-samples
15. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Suggested Architectures
16. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Your web server Browse r Inference query Recommendations Recommendation request Recommendations Real-time event data
17. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Your web server Browser Inference query(userId, search results Reranked results Search query Reranked results Real-time event data Your search service Search Query Search Results
18. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark S3 S3 Pinpoint
19. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark • Personalize features • https://github.com/aws-samples/amazon-personalize-samples • Build a movie recommendation engine • https://ai-services.go-aws.com/50_personalize.html • Blog post • https://aws.amazon.com/blogs/machine-learning/creating-a- recommendation-engine-using-amazon-personalize/ • https://aws.amazon.com/blogs/aws/amazon-personalize-real- time-personalization-and-recommendation-for-everyone/
20. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark
21. © 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark © 2019, Amazon Web Services, Inc. or its Affiliates. All rights reserved. Amazon Confidential and Trademark Thank You!