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
When should you use an ALS model compared to SVD?
What are the advantages and disadvantages of hybrid approaches in recommendation systems?
Explain how LightFM works and what it is used for.
What is the cold start problem and how can it be addressed?
Explain the concept of contextual bandits and how they are applied in recommendation systems.
How can graph neural networks be used to enhance recommendation systems?
When and why should A/B testing be used to evaluate recommendations?
What is collaborative filtering and how does it differ from content-based filtering?
Explain how the Singular Value Decomposition (SVD) method works.
Why is it important to balance exploration and exploitation in recommendation systems?
Explain how Graph Neural Networks (GNN) work in the context of recommendations.
What is LightFM and how does it combine collaborative and content-based filtering?
How do two-tower models work in recommendation systems?
Why is A/B testing an important aspect of evaluating recommendation systems?
What is NDCG and how is it used to evaluate recommendation systems?
Compare matrix factorization models such as SVD and NMF in terms of their application in recommendation systems.
When should content-based filtering be used and what are its limitations?
How does Thompson sampling work and in what situations can it be beneficial?
Prepare for your next interview with Vibe Interview.
Download Vibe Interview