juniorRecommendation Systems
What is collaborative filtering?
Updated May 16, 2026
Short answer
Collaborative filtering recommends items based on similar users or similar item interactions.
Deep explanation
It assumes that users who agreed in the past will agree in the future. It builds a user-item interaction matrix and finds similarities using cosine similarity or Pearson correlation.
Real-world example
Amazon recommending products bought by similar users.
Common mistakes
- Assuming it requires item metadata.
Follow-up questions
- What is user-based vs item-based filtering?
- What is sparsity problem?