FeDecider: An LLM-Based Framework for Federated Cross-Domain Recommendation
Open Access
- 12 April 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 6206-6217
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.Keywords
All Related Versions
Funding Information
- National Science Foundation under award (No. IIS-2316233)
- National Science Foundation under award (No. IIS-2117902)
This publication has 33 references indexed in Scilit:
- Federated Contrastive Learning for Cross-Domain RecommendationIEEE Transactions on Services Computing, 2025
- FedDCSR: Federated Cross-domain Sequential Recommendation via Disentangled Representation Learning2024
- Parameter-efficient fine-tuning of large-scale pre-trained language modelsNature Machine Intelligence, 2023
- FedCDR: Privacy-preserving federated cross-domain recommendationDigital Communications and Networks, 2022
- Beyond Localized Graph Neural Networks: An Attributed Motif Regularization Framework2020
- Guaranteeing Local Differential Privacy on Ultra-Low-Power Systems2018
- Cross Domain Recommender SystemsACM Computing Surveys, 2017
- Neural Collaborative Filtering2017
- Ups and Downs2016
- Introduction to Recommender Systems Handbook2010