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.

1635
questions
16
categories
50
companies

Showing 20 of 139 questions

MediumRecSys

Explain the concept of cold start in recommendation systems and how it can be mitigated.

cold startmitigation strategies
Practice in bot
HardRecSys

Why can the position of recommendations affect user engagement?

position biasuser engagement
Practice in bot
HardRecSys

What are the key principles of using graphs in recommendation systems?

graph-basedprinciples
Practice in bot
HardRecSys

Compare nearest neighbor methods such as FAISS and Annoy in terms of their performance and application.

approximate nearest neighborsperformance
Practice in bot
MediumRecSysQuantum One

Describe a recommendation project in detail: problem definition, testing, and challenges.

ml-recsys
Practice in bot
MediumRecSysQuantum One

How did you determine which product categories are compatible for recommendations?

ml-recsys
Practice in bot
MediumRecSysQuantum One

What features did you use in the recommendation ranking model?

ml-recsys
Practice in bot
MediumRecSysQuantum One

How did you build product embeddings for the recommendation system?

ml-recsys
Practice in bot
MediumRecSysWildberries

Explain ranking metrics: Precision@K, Recall@K, NDCG, and MRR.

ml-recsys
Practice in bot
MediumRecSysVK

Which business metric would you optimize in a short-video recommendation system: average session time or total session time?

ml-recsys
Practice in bot
MediumRecSysVK

Why not simply rank by the probability output of a binary classifier? What advantage does pairwise ranking offer?

ml-recsys
Practice in bot
MediumRecSysVK

Why can't NDCG be optimized directly as a loss function? What alternatives are available?

ml-recsys
Practice in bot
MediumRecSysVK

Explain singular value decomposition (SVD) and how it is used in recommendation systems.

ml-recsys
Practice in bot
MediumRecSysVK

How would you use content-based and collaborative embeddings for candidate generation in a recommendation system?

ml-recsys
Practice in bot
MediumRecSys

What recommendation models and approaches do you know?

ml-recsys
Practice in bot
MediumRecSys

How would you address the cold-start problem in recommendation systems?

ml-recsys
Practice in bot
MediumRecSysСамокат

Which ranking metrics do you know, and how do they differ?

ml-recsys
Practice in bot
MediumRecSysСамокат

How would you improve a recommendation algorithm as more data becomes available?

ml-recsys
Practice in bot
MediumRecSysСамокат

Compare pointwise and pairwise ranking approaches in terms of quality, speed, and metrics.

ml-recsys
Practice in bot
MediumRecSysСамокат

What is the difference between MAP and NDCG?

ml-recsys
Practice in bot

Prepare for your next interview with Vibe Interview.

Download Vibe Interview
RecSys interview questions — page 3 | Vibe Interview