FeDecider: An LLM-Based Framework for Federated Cross-Domain Recommendation

TL;DR

FeDecider introduces a novel LLM-based federated cross-domain recommendation framework that effectively addresses heterogeneity and privacy challenges by disentangling updates and learning personalized weights, demonstrating superior performance across datasets.

Contribution

The paper presents FeDecider, a new framework that enables federated cross-domain recommendation using LLMs, with innovative update disentanglement and personalized integration strategies.

Findings

FeDecider outperforms baseline methods on multiple datasets.

Disentangling low-rank updates improves model robustness.

Personalized weights enhance cross-domain recommendation accuracy.

Abstract

Federated cross-domain recommendation (Federated CDR) aims to collaboratively learn personalized recommendation models across heterogeneous domains while preserving data privacy. Recently, large language model (LLM)-based recommendation models have demonstrated impressive performance by leveraging LLMs' strong reasoning capabilities and broad knowledge. However, adopting LLM-based recommendation models in Federated CDR scenarios introduces new challenges. First, there exists a risk of overfitting with domain-specific local adapters. The magnitudes of locally optimized parameter updates often vary across domains, causing biased aggregation and overfitting toward domain-specific distributions. Second, unlike traditional recommendation models (e.g., collaborative filtering, bipartite graph-based methods) that learn explicit and comparable user/item representations, LLMs encode knowledge implicitly through autoregressive text generation training. This poses additional challenges for effectively measuring the cross-domain similarities under heterogeneity. To address these challenges, we propose an LLM-based framework for federated cross-domain recommendation, FeDecider. Specifically, FeDecider tackles the challenge of scale-specific noise by disentangling each client's low-rank updates and sharing only their directional components. To handle the need for flexible and effective integration, each client further learns personalized weights that achieve the data-aware integration of updates from other domains. Extensive experiments across diverse datasets validate the effectiveness of our proposed FeDecider.