New FedHUR framework improves personalized recommendations with hierarchical client relations

Researchers have developed FedHUR, a new federated recommendation framework designed to improve personalized recommendations by learning hierarchical client relationships. Unlike previous methods that rely on predefined assumptions for client relations, FedHUR constructs these relations based on item-item filters and utility signals. This approach allows clients to identify and aggregate useful collaborative information, leading to better prediction performance. Experiments on five real-world datasets demonstrate that FedHUR significantly outperforms existing federated recommendation baselines. AI

IMPACT This research could lead to more accurate and personalized recommendation systems by improving how data is shared and utilized across decentralized clients.

RANK_REASON The item is a research paper published on arXiv detailing a new framework for federated recommendation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.IR (Information Retrieval) →

• alphaXiv
• arXiv
• CatalyzeX
• CORE Recommender
• DagsHub
• FedHUR
• Gotit.pub
• Hugging Face
• Influence Flower
• ScienceCast

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →