Session 14: Robust RecSys 2

Date: Thursday October 17, 12:00 PM – 12:30 PM (GMT+2)
Room: Petruzzelli Theater
Session Chair: Abhishek Srivastava

• RES 🕓15FedLoCA: Low-Rank Coordinated Adaptation with Knowledge Decoupling for Federated Recommendations
  by Yuchen Ding (University of Science and Technology of China), Siqing Zhang (University of Science and Technology of China), Boyu Fan (University of Helsinki), Wei Sun (University of Science and Technology of China), Yong Liao (University of Science and Technology of China) and Peng Yuan Zhou (Aarhus University)

  Privacy protection in recommendation systems is gaining increasing attention, for which federated learning has emerged as a promising solution. Current federated recommendation systems grapple with high communication overhead due to sharing dense global embeddings, and also poorly reflect user preferences due to data heterogeneity. To overcome these challenges, we propose a two-stage Federated Low-rank Coordinated Adaptation (FedLoCA) framework to decouple global and client-specific knowledge into low-rank embeddings, which significantly reduces communication overhead while enhancing the system’s ability to capture individual user preferences amidst data heterogeneity. Further, to tackle gradient estimation inaccuracies stemming from data sparsity in federated recommendation systems, we introduce an adversarial gradient projected descent approach in low-rank spaces, which significantly boosts model performance while maintaining robustness. Remarkably, FedLoCA also alleviates performance loss even under the stringent constraints of differential privacy. Extensive experiments on various real-world datasets demonstrate that FedLoCA significantly outperforms existing methods in both recommendation accuracy and communication efficiency.

  Full text in ACM Digital Library

• RES 🕓15Accelerating the Surrogate Retraining for Poisoning Attacks against Recommender Systems
  by Yunfan Wu (Chinese Academy of Sciences), Qi Cao (Chinese Academy of Sciences), Shuchang Tao (Chinese Academy of Sciences), Kaike Zhang (Chinese Academy of Sciences), Fei Sun (Chinese Academy of Sciences) and Huawei Shen (Chinese Academy of Sciences)

  Recent studies have demonstrated the vulnerability of recommender systems to data poisoning attacks, where adversaries inject carefully crafted fake user interactions into the training data of recommenders to promote target items. Current attack methods involve iteratively retraining a surrogate recommender on the poisoned data with the latest fake users to optimize the attack. However, this repetitive retraining is highly time-consuming, hindering the efficient assessment and optimization of fake users. To mitigate this computational bottleneck and develop a more effective attack in an affordable time, we analyze the retraining process and find that a change in the representation of one user/item will cause a cascading effect through the user-item interaction graph. Under theoretical guidance, we introduce Gradient Passing (GP), a novel technique that explicitly passes gradients between interacted user-item pairs during backpropagation, thereby approximating the cascading effect and accelerating retraining. With just a single update, GP can achieve effects comparable to multiple original training iterations. Under the same number of retraining epochs, GP enables a closer approximation of the surrogate recommender to the victim. This more accurate approximation provides better guidance for optimizing fake users, ultimately leading to enhanced data poisoning attacks. Extensive experiments on real-world datasets demonstrate the efficiency and effectiveness of our proposed GP.

  Full text in ACM Digital Library

Back to program