Personalized Federated Sequential Recommender

TL;DR

This paper introduces PFSR, a personalized federated sequential recommender system that enhances efficiency and personalization by capturing global user profiles, enabling fine-tuning, and preserving local information.

Contribution

The paper proposes the PFSR framework with novel components like the Associative Mamba Block, Variable Response Mechanism, and Dynamic Magnitude Loss for improved personalized recommendation.

Findings

Improved prediction efficiency through the Associative Mamba Block.

Enhanced personalization with the Variable Response Mechanism.

Better preservation of local user information using Dynamic Magnitude Loss.

Abstract

In the domain of consumer electronics, personalized sequential recommendation has emerged as a central task. Current methodologies in this field are largely centered on modeling user behavior and have achieved notable performance. Nevertheless, the inherent quadratic computational complexity typical of most existing approaches often leads to inefficiencies that hinder real-time recommendation. Moreover, these methods face challenges in being effectively adapted to the personalized requirements of users across diverse scenarios. To tackle these issues, we propose the Personalized Federated Sequential Recommender (PFSR). In this framework, an Associative Mamba Block is introduced to capture user profiles from a global perspective while improving prediction efficiency. In addition, a Variable Response Mechanism is developed to enable fine-tuning of parameters in accordance with individual user needs. A Dynamic Magnitude Loss is further devised to preserve greater amounts of localized personalized information throughout the training process.