Build a Powerful Recommendation Engine Using Image Recognition Technology on AWS Samuel James | 09.09.2019 Community Day 2019 Sponsors
� � � About Architrave Gmbh PropTech Company of the year 2018 Over 3,900 managed assets worth € 80 billion Over € 12 billion in annual transaction volume (including Germany's largest single transactions: Sony Centre 2017 and Frankfurt Omni Tower 2018) Frankfurt Berlin Dresden
What This Talk is Not About Recommender Systems’ Algorithms
Agenda Recommendation system Why recommender systems are important How recommendation engines work Leveraging on AWS Rekognition for product recommendation Demo
What Is Recommendation ? Recommendation is about providing relevant content to the user based on knowledge of the user, content, and interactions between the user and items.
� 6
According to McKinsey & Company, 35% of Amazon.com’s revenue is generated by its recommendation engine � 7
"Netflix saves up to $1 billion a year via its personalised recommendations” – Business Insider � 8
Personalised recommendations drive sales � 9
Recommendation Phases Data Data Storage Data Analysis Data Filtering Collection
Data Filtering Techniques Collaborative Filtering Content-Based Filtering Hybrid Filtering
Collaborative Filtering Interactions of users with products (like movies watched, products viewed, products bought etc.
Collaborative Filtering Technique in a Retail Site
Content-based Filtering Focuses on properties of items. Similarity of items is determined by measuring the similarity in their properties.
Content-Based Filtering Technique in a Retail Site
Hybrid Filtering Combines collaborative filtering and content-based filtering.
Leveraging on AWS Rekognition API for analysis of unstructured data like images � 17
AWS Rekognition at a Glance
� 19
� 20
DEMO � 21
Using S3 Batch Operations
Handling new Uploads
� 24
How the visual search works � 25
� 26
Lessons Learned Building with AWS reduces your development time You don't need to be AI experts to have AI capability in your app Build a repeatable and reusable infrastructure with AWS
Thanks! Questions? Contact Samuel James Architrave GmbH Samuel James Bouchéstraße 12, Email: james@architrave.de Building 1A, 12435 Berlin @samuelabiodunj � 28
Recommend
More recommend