Federated recommender systems enable collaborative model training while keeping user interaction data local and sharing only essential model parameters, thereby mitigating privacy risks. However, existing methods overlook a critical issue, i.e., the stable learning of a generalized item embedding throughout the federated recommender system training process. Item embedding plays a central role in facilitating knowledge sharing across clients. Yet, under the cross-device setting, local data distributions exhibit significant heterogeneity and sparsity, exacerbating the difficulty of learning generalized embeddings. These factors make the stable learning of generalized item embeddings both indispensable for effective federated recommendation and inherently difficult to achieve. To fill this gap, we propose a new federated recommendation framework, named Federated Recommendation with Generalized Embedding Learning (FedRecGEL). We reformulate the federated recommendation problem from an item-centered perspective and cast it as a multi-task learning problem, aiming to learn generalized embeddings throughout the training procedure. Based on theoretical analysis, we employ sharpness-aware minimization to address the generalization problem, thereby stabilizing the training process and enhancing recommendation performance. Extensive experiments on four datasets demonstrate the effectiveness of FedRecGEL in significantly improving federated recommendation performance. Our code is available at https://github.com/anonymifish/FedRecGEL.

Recommender systems have become increasingly influential in shaping user behavior and decision-making, highlighting their growing impact in various domains. Meanwhile, the widespread adoption of machine learning models in recommender systems has raised significant concerns regarding user privacy and security. As compliance with privacy regulations becomes more critical, there is a pressing need to address the issue of recommendation unlearning, i.e., eliminating the memory of specific training data from the learned recommendation models. Despite its importance, traditional machine unlearning methods are ill-suited for recommendation unlearning due to the unique challenges posed by collaborative interactions and model parameters. This survey offers a comprehensive review of the latest advancements in recommendation unlearning, exploring the design principles, challenges, and methodologies associated with this emerging field. We provide a unified taxonomy that categorizes different recommendation unlearning approaches, followed by a summary of widely used benchmarks and metrics for evaluation. By reviewing the current state of research, this survey aims to guide the development of more efficient, scalable, and robust recommendation unlearning techniques. Furthermore, we identify open research questions in this field, which could pave the way for future innovations not only in recommendation unlearning but also in a broader range of unlearning tasks across different machine learning applications.

With the escalating privacy concerns in recommender systems, attribute unlearning has drawn widespread attention as an effective approach against attribute inference attacks. This approach focuses on unlearning users' privacy attributes to reduce the performance of attackers while preserving the overall effectiveness of recommendation. Current research attempts to achieve attribute unlearning through adversarial training and distribution alignment in the statistic setting. However, these methods often struggle in dynamic real-world environments, particularly when considering scenarios where unlearning requests are frequently updated. In this paper, we first identify three main challenges of current methods in dynamic environments, i.e., irreversible operation, low efficiency, and unsatisfied recommendation preservation. To overcome these challenges, we propose a Pluggable Attribute Unlearning framework, PAU. Upon receiving an unlearning request, PAU plugs an additional erasure module into the original model to achieve unlearning. This module can perform a reverse operation if the request is later withdrawn. To enhance the efficiency of unlearning, we introduce rate distortion theory and reduce the attack performance by maximizing the encoded bits required for users' embedding within the same class of the unlearned attribute and minimizing those for different classes, which eliminates the need to calculate the centroid distribution for alignment. We further preserve recommendation performance by constraining the compactness of the user embedding space around a reasonable flood level. Extensive experiments conducted on four real-world datasets and three mainstream recommendation models demonstrate the effectiveness of our proposed framework.

With the growing privacy concerns in recommender systems, recommendation unlearning is getting increasing attention. Existing studies predominantly use training data, i.e., model inputs, as unlearning target. However, attackers can extract private information from the model even if it has not been explicitly encountered during training. We name this unseen information as attribute and treat it as unlearning target. To protect the sensitive attribute of users, Attribute Unlearning (AU) aims to make target attributes indistinguishable. In this article, we focus on a strict but practical setting of AU, namely Post-Training Attribute Unlearning (PoT-AU), where unlearning can only be performed after the training of the recommendation model is completed. To address the PoT-AU problem in recommender systems, we propose a two-component loss function. The first component is distinguishability loss, where we design a distribution-based measurement to make attribute labels indistinguishable from attackers. We further extend this measurement to handle multi-class attribute cases with efficient computational overhead. The second component is regularization loss, where we explore a function-space measurement that effectively maintains recommendation performance compared to parameter-space regularization. We use stochastic gradient descent algorithm to optimize our proposed loss. Extensive experiments on four real-world datasets demonstrate the effectiveness of our proposed methods.

Federated Recommendation (FedRec) has been widely applied recently for realizing privacy preservation in recommender systems. However, due to direct uploads of model gradients from all clients, FedRec is vulnerable to potential poisoning attacks. In this article, we focus on the targeted model poisoning attacks in FedRec, which aims to raise the exposure ratio of specific target items by generating poisoned gradients to influence global training. Challenges emerge when implementing this kind of attack. On the one hand, simulating authentic users on the malicious clients for downstream poisoning is hard when access to prior knowledge is limited. On the other hand, distinguished item attributes and personalized user preferences require the attack to be adaptive to complex distributions. To this end, we propose a novel attack DuAda with two modules, i.e., dummy user simulator and adaptive distribution attacker . The dummy user simulator is designed to generate malicious users with characteristics similar to real users, which exploits authentic user representations and preference labels simultaneously through two-stage inversion optimization. The attacker first extracts heterogeneous distributions by a special multi-prototype clustering method, and then conducts adaptive attacks from both explicit and implicit promotion perspectives. The explicit promotion raises the prediction scores of target items based on the inherent characteristics, while the implicit promotion imbues them with the features of popular items. Targeted at our proposed attack method, we also design a merged adaptive defense mechanism to fight against DuAda and conduct defensive experiments. Empirical studies on four real-world datasets demonstrate the effectiveness and interpretability of DuAda.

The service system involves multiple stakeholders, making it crucial to ensure fairness. In this paper, we take the example of a typical service system, the recommender system, to investigate how to identify and tackle fairness issues within the service system. Recommender systems often exhibit bias towards a small user group, resulting in pronounced unfairness in recommendation performance, specifically the User-Oriented Fairness (UOF) issue. Existing research on UOF faces limitations in addressing two pivotal challenges: CH1: Current methods fall short in addressing the root cause of the UOF issue, stemming from an unfair training process between advantaged and disadvantaged users. CH2: Current methods struggle to unveil compelling correlations among users in sparse datasets. In this paper, we propose a novel Hypergraph Convolutional Network for User-Oriented Fairness, namely HyperUOF, to address the aforementioned challenges. HyperUOF serves as a versatile framework applicable to various backbone recommendation models for achieving UOF. To address CH1, HyperUOF employs an in-processing method that enhances the training process of disadvantaged users during model training. To addressCH2, HyperUOF incorporates a hypergraph-based approach, proven effective in sparse datasets, to explore high-order correlations among users. We conduct extensive experiments on three real-world datasets based on four backbone recommendation models to prove the effectiveness of our proposed HyperUOF.