Personalized Federated Recommender Systems with Private and Partially Federated AutoEncoders

Keyphrases

• Autoencoder 100%
• Federated Recommender System 100%
• Recommender Systems 66%
• User Data 66%
• Federated Learning 50%
• Server-side 33%
• Main Components 16%
• Global Model 16%
• Application Domain 16%
• Decoder 16%
• Encoder 16%
• Personal Model 16%
• Computation Overhead 16%
• Privacy Concerns 16%
• Communication Overhead 16%
• Local Data 16%
• Personalization 16%
• Recommendation Model 16%
• System Guarantee 16%
• Partial Compression 16%
• User-item Interaction 16%
• Digital Marketing 16%
• Data Embedding 16%
• Activity-based Model 16%
• Less Computation 16%
• Local Datasets 16%
• Reliable Recommendations 16%

Computer Science

• Recommender Systems 100%
• Autoencoder 100%
• User Data 40%
• Federated Learning 30%
• Communication Overhead 10%
• Privacy Problem 10%
• Embedding Data 10%
• Application Domain 10%
• Individual User 10%
• Main Component 10%