Question Bank
RecSys interview questions
Recommendations must find relevant items while respecting product constraints. Review candidate retrieval, ranking and evaluation. Explain how cold starts, user feedback and the difference between offline metrics and online experiments affect your choices.
Showing 20 of 139 questions
Explain what collaborative filtering is and how it works.
What is matrix factorization and what is it used for in recommendation systems?
Compare explicit and implicit feedback in the context of recommendation systems.
Explain how Alternating Least Squares (ALS) works in matrix factorization.
When and why should hybrid approaches be used in recommendation systems?
What is GRU4Rec and how is it applied in session-based recommendations?
Explain how LightFM works and what its advantages are.
Why are metrics like NDCG and MAP important when evaluating recommendations?
What are the main differences between candidate retrieval and ranking approaches in recommendation systems?
Explain how methods like epsilon-greedy and Thompson sampling work in the context of exploration and exploitation.
When should you use matrix factorization methods like SVD compared to ALS?
What are the main metrics for evaluating the performance of a recommendation system and how should they be interpreted?
Explain the principle of two-stage ranking models in recommendation systems.
How does Thompson Sampling work and when should it be applied?
Explain how GRU4Rec is suitable for session-based recommendations.
Compare and analyze approximate nearest neighbor search methods like FAISS and Annoy.
What are the main reasons for popularity bias in recommendation systems and how can it affect the outcomes?
Why is A/B testing important in recommendation systems?
Explain the difference between explicit and implicit feedback in recommendation systems.
Prepare for your next interview with Vibe Interview.
Download Vibe Interview