Federated News Recommendation with Fine-grained Interpolation and Dynamic Clustering

By Sanshi Lei et al

Published on Oct. 21, 2023

Read the original document by opening this link in a new tab.

Table of Contents

1. Introduction
2. Related Work
3. Preliminaries
4. Methodology

Summary

Researchers have adapted privacy-preserving Federated Learning (FL) to news recommendation to protect users' privacy. Personalized Federated Learning (PFL) and model interpolation have emerged to address data heterogeneity issues. The proposed framework, FINDING, focuses on fine-grained model interpolation and group-level personalization to enhance news recommendation performance. Extensive experiments on real-world datasets demonstrate the effectiveness of the approach.