Design a search relevance prediction approach
Company: Amazon
Role: Machine Learning Engineer
Category: Machine Learning
Difficulty: medium
Interview Round: Onsite
## Search relevance prediction
You are asked to predict relevance for an e-commerce search engine (given a user query and a product/document).
### Prompt
1. How would you model relevance (classification vs regression vs learning-to-rank)?
2. What features would you engineer (lexical, semantic, behavioral, context)?
3. How would you obtain labels (human judgments vs clicks/purchases) and handle bias in behavioral data?
4. What offline metrics would you use, and how would you validate online?
Quick Answer: This question evaluates competency in machine learning for search relevance, including relevance modeling, feature engineering across lexical, semantic, behavioral and contextual signals, label sourcing and bias mitigation, and offline/online evaluation metrics.
