A Million Products, One Experience: Using Collaborative and Multimodal Embeddings
summary
Describes building a product clustering system for a marketplace to create thematic collections and improve user experience for millions of items. Explains using collaborative and multimodal embeddings (likely combining user-behavior/collaborative signals and multimodal item representations) to scale clustering across a massive catalog.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 3:42Cold product problem
- 8:55Clustering technical challenges
- 16:24HDP Scan experiments
- 23:59Dynamic catalog handling
- 29:38Production solution and results
- 32:37Q&A
related talks
31mEmbeddings... Embeddings Everywhere
27mData with Purpose: The Power of Embeddings
41mSearching for Products by Photo
56mApplications of Multimodal Learning in Media Search Engines
57mUsing Vector Databases for Multimodal Embeddings and Search
01h 08mGoing multi-modal
38mEmbeddings! Embeddings everywhere!
52mThe ML wedding between millions of products, or how to withstand load in a stream