Relevance Tuning Is Not Magic — It's Machine Learning!
summary
Patricia Kaufmann explains how relevance tuning for search is not magic but uses machine learning techniques. She discusses practical approaches to improving search relevance and analytics as applied in e-commerce search systems, drawing on her work at SHI GmbH.
AI-written summary · about
chapters (10)
- 0:00Relevance tuning basics
- 1:31Streaming platform example
- 3:41Solr relevance tuning
- 8:15Manual boosting challenges
- 12:32Machine learning solution
- 15:00Learning to Rank
- 20:24Feature definition challenges
- 23:10Using models in Solr
- 25:19Model visualization
- 26:45Model testing
related talks
40mLevel Up: Everything you need to know about search, ...
40mAgentic Relevance Tuning With OpenSearch: Faster, Easier, and Higher Quality Search at Scale
21mAmazon Search: The Joy of Ranking Products
39mMachine Learning Engineering
39mIntroducing OpenSearch’s Search Relevance Workbench!
22mMachine Learning: The Engine Behind Modern E-commerce Search
39mMachine Learning for Relevance and Serendipity
38mLearning to Rank: where search meets machine learning #bbuzz