Federated Learning for Personalized User Experiences in the Metaverse

Muhammad Asad, Zayed University
Safa Otoum, Zayed University

Document Type

Conference Proceeding

Source of Publication

2024 2nd International Conference on Intelligent Metaverse Technologies & Applications (iMETA)

Publication Date

11-29-2024

Abstract

The metaverse presents significant challenges in managing vast amounts of user data while ensuring personalized experiences. Traditional centralized approaches to data management risk privacy breaches and inefficiencies. This paper proposes a novel application of Federated Learning (FL) to address these issues by enabling decentralized model training across user devices. Our FL framework ensures data privacy and security while delivering personalized interactions and content. Through simulations, we demonstrate the framework's effectiveness in enhancing user experiences in the metaverse. We also address key technical challenges, including scalability, latency, and model accuracy, and discuss future research directions to further improve FL in this context. Our findings highlight the potential of FL to revolutionize data management and personalization in the rapidly evolving landscape of the metaverse.

ISBN

979-8-3503-5151-4

Publisher

IEEE

Volume

00

First Page

050

Last Page

055

Disciplines

Computer Sciences

Keywords

Federated Learning, Metaverse, Data Privacy, User Personalization, Decentralized Training

Indexed in Scopus

no

Open Access

no

Link to Full Text

Share

COinS