Recommendation System beyond traditional Collaborative filtering
summary
Describes a custom recommendation system built for Snapdeal’s 12M-item, multi-category catalog that extends beyond traditional collaborative filtering. Covers specific techniques and mathematical models for user category affinity at fine granularity, content-based product similarity, complementary-product (paired) recommendations, repurchase prediction, trending-product suggestions based on user affinity, using implicit feedback to improve relevance, and a weighted use of collaborative filtering; also addresses technical challenges of scalability and real-time serving.
AI-written summary · about
chapters (12)
- 0:00Start
- 0:52Introduction
- 1:29Collaborative Filtering
- 6:25Issues with this approach
- 8:59Capturing Data in Near Real Time
- 12:49Key terms
- 14:27Clean Seed Generation
- 16:46Exploiter
- 20:46Explorer
- 23:05Merger
- 25:12Technical Challenges
- 29:19What's Next?
related talks
41mRecommender Systems: Crafting Effective Personalised Experiences
32mLet's talk about Recommendation Systems!
34mRecommender Systems
55mImproving product discovery via relevance and ranking optimization
37mA primer on recommendation systems.
29mThe Challenges of Building a Personalized Recommendation System in E-Commerce
28mThe Power of Personalized Customer Experiences Through AI Innovations
22mExploiting Auxiliary Information for Better Recommendation