Using Machine Learning to Generate Structured Snippets
summary
Describes using machine learning to generate structured snippets for search results to improve relevance despite language ambiguity and user preference variability. Covers problems where relevant items can appear anywhere in results and outlines ML approaches to detect and surface structured, high-relevance snippets for users.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 5:51Job search snippets problem
- 11:51User research insights
- 17:07Automated data labeling
- 24:37Model training
- 30:20Generating snippets
- 33:10A/B test results
- 37:26Summary
- 40:18Q&A
related talks
42mThe Machine Understands Text — Learning and Search at Micro and Macro Scale
39mMachine Learning for Relevance and Serendipity
35mIndustrial Applications of Machine Learning in Search
22mMachine Learning: The Engine Behind Modern E-commerce Search
30mSemantic Search and Neural Information Retrieval
40mLevel Up: Everything you need to know about search, ...
41mResearch in the Age of AI
23mML-Powered Search & Recommendation 101