In recommender systems, recent advances highlight the critical role of alignment and uniformity (AU) in representation learning. Specifically, AU-based methods pull positive user-item pairs closer (alignment) and spread the overall representation distribution (uniformity), typically relying on observed positive samples. Despite their effectiveness, exist methods face two limitations: (1) noise issues have a more severe impact on AU-based methods in the absence of negative samples, leading to the capture of spurious signals such as misclicks or non-preferential behaviors; (2) data sparsity weakens the alignment of user-item representations, hindering reliable representation learning and harming recommendations for sparse users. To tackle these issues, we propose a novel recommendation framework named Constrained Alignment and Filtered Uniformity (CAFU). CAFU enhances robustness through Filtered Uniformity (FU) and improves performance under data sparsity via Constrained Alignment (CA). Specifically, FU adopts a threshold-based strategy to eliminate unreliable samples that degrade embedding quality, thereby strengthening robustness. In parallel, CA mitigates the impact of sparsity by masking low-confidence user-item pairs based on angular distance, leading to better recommendation for sparse users. Extensive experiments on three datasets and three backbones demonstrate the effectiveness and generalization of the proposed framework.

Graph Contrastive Learning (GCL) has recently emerged as a powerful paradigm for modeling user–item interactions and learning high-quality representations in recommender systems. While existing GCL-based methods benefit from data augmentation and sampling strategies, they often overlook the inherent limitations of the contrastive objectives: 1) Stacking multiple Graph Convolutional Network layers to capture high-order information often causes the over-smoothing phenomenon, where node representations become overly similar. 2) Structurally similar negative sample pairs may exhibit high cosine similarity, causing gradient saturation during representation optimization. To address the above challenges, we revisit matrix factorization in recommendation models and uncover its implicit connection to a parallel graph filter bank. This perspective reveals how overly aggressive low-pass or high-pass filtering distorts feature distributions, contributing to gradient saturation. Building on this insight, we propose Light Cosine Similarity Collaborative Filtering (LightCSCF), a margin-constrained method that improves gradient optimization in contrastive learning by focusing on structurally hard examples, alleviating both gradient saturation and boundary over-smoothing. Extensive experiments on three real-world datasets demonstrate that LightCSCF consistently outperforms state-of-the-art baselines in recommendation accuracy and robustness to data sparsity.

Hourglass: Enabling Efficient Split Federated Learning with Data Parallelism

In an edge-assisted federated learning (FL) system, edge servers aggregate the local models from the clients within their coverage areas to produce intermediate models for the production of the global model. This significantly reduces the communication overhead incurred during the FL process. To accelerate model convergence, FedEdge, the state-of-the-art edge-assisted FL system, trains clients' models in local federations when they wait for the global model in each training round. However, our investigation reveals that it drives the global model towards clients with excessive local training, causing model drifts that undermine model performance for other clients. To tackle this problem, this paper presents Maverick, a new edge-assisted FL system that mitigates model drifts by training personalized local models for clients through contrastive local training. It introduces a model-contrastive loss to facilitate personalized local federated training by driving clients' local models away from the global model and close to their corresponding intermediate models. In addition, Maverick includes anomalous models in contrastive local training as negative samples to accelerate the convergence of clients' local models. Extensive experiments are conducted on three widely-used models trained on three datasets to comprehensively evaluate the performance of Maverick. Compared to state-of-the-art edge-assisted FL systems, Maverick accelerates model convergence by up to 16.2x and improves model accuracy by up to 12.7%.

The traditional cloud computing paradigm faces challenges with the increasing number of Artificial Intelligence of Things (AIoT) devices generated at the network edge. Edge computing provides a novel approach to overcoming the limitations of cloud computing by delivering lower service latency and higher quality of service (QoS) to AIoT devices. However, edge computing encounters constraints due to scarce resources on edge servers and the long distance between AIoT devices and remote cloud servers. To ensure high QoS for AIoT devices, a balance is required between limited computing resources on the network edge and high latency caused by the geographic distance on the cloud side. In this paper, we propose an edge-cloud architecture that achieves optimal QoS for AIoT devices. We employ a federated learning-based architecture to train AIoT devices’ data locally, thereby ensuring privacy. We evaluate the effectiveness and efficiency of our proposed approach by comparing it with the centralized approach from the state-of-the-art using two widely used datasets. The experimental results demonstrate that our architecture achieves higher effectiveness and efficiency in improving AIoT devices’ QoS. Overall, our proposed edge-cloud architecture overcomes the limitations of traditional cloud computing, enhances user privacy, and delivers high QoS to AIoT devices.

Mobile and Web-of-Things (WoT) devices at the network edge account for more than half of the world’s web traffic, making a great data source for various machine learning (ML) applications, particularly federated learning (FL) which offers a promising solution to privacy-preserving ML feeding on these data. FL allows edge mobile and WoT devices to train a shared global ML model under the orchestration of a central parameter server. In the real world, due to resource heterogeneity, these edge devices often train different versions of models (e.g., VGG-16 and VGG-19) or different ML models (e.g., VGG and ResNet) for the same ML task (e.g., computer vision and speech recognition). Existing FL schemes have assumed that participating edge devices share a common model architecture, and thus cannot facilitate FL across edge devices with heterogeneous ML model architectures. We explored this architecture heterogeneity challenge and found that FL can and should accommodate these edge devices to improve model accuracy and accelerate model training. This paper presents our findings and FlexiFed, a novel scheme for FL across edge devices with heterogeneous model architectures, and three model aggregation strategies for accommodating architecture heterogeneity under FlexiFed. Experiments with four widely-used ML models on four public datasets demonstrate 1) the usefulness of FlexiFed; and 2) that compared with the state-of-the-art FL scheme, FlexiFed improves model accuracy by 2.6%-9.7% and accelerates model convergence by 1.24 × -4.04 ×.

Federated learning (FL) has been widely acknowledged as a promising solution to training machine learning (ML) model training with privacy preservation. To reduce the traffic overheads incurred by FL systems, edge servers have been included between clients and the parameter server to aggregate clients’ local models. Recent studies on this edge-assisted hierarchical FL scheme have focused on ensuring or accelerating model convergence by coping with various factors, e.g., uncertain network conditions, unreliable clients, heterogeneous compute resources, etc. This paper presents our three new discoveries of the edge-assisted hierarchical FL scheme: 1) it wastes significant time during its two-phase training rounds; 2) it does not recognize or utilize model diversity when producing a global model; and 3) it is vulnerable to model poisoning attacks. To overcome these drawbacks, we propose FedEdge, a novel edge-assisted hierarchical FL scheme that accelerates model training with asynchronous local federated training and adaptive model aggregation. Extensive experiments are conducted on two widely-used public datasets. The results demonstrate that, compared with state-of-the-art FL schemes, FedEdge accelerates model convergence by 1.14 × −3.20 ×, and improves model accuracy by 2.14% - 6.63%.

The number of Web services on the Internet has been growing rapidly. This has made it increasingly difficult for users to find the right services from a large number of functionally equivalent candidate services. Inspecting every Web service for their quality value is impractical because it is very resource consuming. Therefore, the problem of quality prediction for Web services has attracted a lot of attention in the past several years, with a focus on the application of the Matrix Factorization (MF) technique. Recently, researchers have started to employ user similarity to improve MF-based prediction methods for Web services. However, none of the existing methods has properly and systematically addressed two of the major issues: 1) retrieving appropriate neighborhood information, i.e., similar users and services; 2) utilizing full neighborhood information, i.e., both users’ and services’ neighborhood information. In this paper, we propose CNMF, acovering-based quality prediction method for Web services vianeighborhood-awarematrixfactorization. The novelty of CNMF is twofold. First, it employs a covering-based clustering method to find similar users and services, which does not require the number of clusters and cluster centroids to be prespecified. Second, it utilizes neighborhood information on both users and services to improve the prediction accuracy. The results of experiments conducted on a real-world dataset containing 1,974,675 Web service invocation records demonstrate that CNMF significantly outperforms eight existing quality prediction methods, including two state-of-the-art methods that also utilize neighborhood information with MF.

The applications of mobile devices are increasingly becoming computationally intensive while the computing capability of the user’s mobile device is limited. Traditional approaches offload the tasks of mobile applications to the remote cloud. However, the rapid growth of mobile devices has made it a challenge for the remote cloud to provide computing and storage capacities with low communication delays due to the fact that the remote cloud is geographically far away from mobile devices. Reducing the completion time of applications in mobile devices through the technical expending mobile cloudlets which are moving collocated with Access Points (APs) is necessary. To address the above issues, this paper proposes EACP-CA (Enhanced Adaptive Cloudlets Placement approach based on Covering Algorithm), an enhanced adaptive cloudlet placement approach for mobile applications in a given network area. We apply the CA (Covering Algorithm) to adaptively cluster the mobile devices based on their geographical locations, the aggregation regions of the mobile devices are identified, and the cloudlet destination locations are also confirmed according to the clustering centers. In addition, we can also obtain the traces between the original and destination locations of these mobile cloudlets. To increase the efficiency, we parallelize CA on Spark. Extensive experiments show that the proposed approach outperforms the existing approach in both effectiveness and efficiency.