Recommendation System: Past, Present and Future

Recommendation systems power 35% of Amazon's sales, account for 80% of viewing hours on Netflix, and the domain is projected to generate more than $15 billion worth of value by 2030. Being such an in-demand domain, this live session aims to provide an introduction to the field, and underscore some of the key research papers that have led the field to its current state. The key takeaways from this session will be

Metrics for Recommendation Systems

• 1st Generation RecSys - Content Based, Collaborative Filtering, Hybrid
• 2nd Generation RecSys - Utilizing Implicit Data, Mulit-Faceted CF and more
• 3rd Generation RecSys - Wide and Deep Networks, Two-Tower Network, Graph Network and more
• Current challenges and how you can get started

Karun Thankachan is a Senior Data scientist specializing in NLP and Recommender systems with nearly 8 of experience across data science/software development and research.

He completed his Master's in Data Science at Carnegie Mellon University (2020) and has worked for several e-commerce giants such as Amazon, Walmart, and Dell. You can reach him on LinkedIn.