Researchers have developed AutoRelAnnotator, a novel system designed to generate high-quality relevance annotations for sponsored search at scale and with reduced costs. The system employs a calibrated model cascade, routing queries through progressively larger fine-tuned classifiers. This approach optimizes for accuracy through domain-specific fine-tuning and cost-efficiency via cascading, with per-class isotonic calibration further enhancing performance. AutoRelAnnotator has been validated in production, processing over 150 million annotations and accelerating experimentation cycles for search and advertising systems. AI
IMPACT Enables more efficient and scalable annotation pipelines for search and advertising systems.
RANK_REASON The item is a research paper detailing a new method for relevance evaluation in sponsored search. [lever_c_demoted from research: ic=1 ai=1.0]
Read on arXiv cs.IR (Information Retrieval) →
- alphaXiv
- arXiv
- AutoRelAnnotator
- CatalyzeX
- Connected Papers
- DagsHub
- Gotit.pub
- Hugging Face
- Litmaps
- ScienceCast
- SciTE
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
