Recommendation Systems

• Recommendation System
• Personalisation
• Recommendation Scenario
• User (RecSys)
• Item
• Catalog
• User-Item Interaction
• Explicit Feedback
• Implicit Feedback
• User-Item Matrix
• Relevance (RecSys)
• Long Tail

• Rating Prediction
• Matrix Completion
• Top-N Recommendation
• Slate Recommendation
• Next-Item Prediction
• Recommendation as Retrieval
• Recommendation as Ranking

• Memory-Based CF
• Model-Based CF
• Neighborhood Method
• Significance Weighting

• Content-Based Recommender
• Item Profile
• User Profile (Content)
• Content Vector
• Categorical Features (RecSys)
• Profile Matching
• Tag-Based Recommendation

• Context-Aware Recommendation
• Contextual Pre-Filtering
• Contextual Post-Filtering
• Contextual Modeling
• Tensor Factorization
• Time-Aware Recommendation

• Hybrid Recommender
• Weighted Hybrid
• Switching Hybrid
• Mixed Hybrid
• Feature Combination Hybrid
• Cascade Hybrid
• Meta-Level Hybrid
• Demographic Filtering

• Cold Start Problem
• New-User Cold Start
• New-Item Cold Start
• Side Information
• Sparsity (RecSys)
• Gray Sheep Problem

• Popularity Bias
• Selection Bias (RecSys)
• Position Bias
• Implicit Feedback Bias
• Filter Bubble
• Echo Chamber
• Algorithmic Fairness (RecSys)
• Ranking Calibration
• Privacy-Preserving Recommendation
• Federated Recommendation
• Differential Privacy in RecSys

• Shilling Attack
• Profile Injection Attack
• Attack Detection (RecSys)
• Robustness (RecSys)