Recently, the graph neural network (GNN) has shown great power in matrix completion by formulating a rating matrix as a bipartite graph and then predicting the link between the corresponding user and item nodes. The majority of GNN-based matrix completion methods are based on Graph Autoencoder (GAE), which considers the one-hot index as input, maps a user (or item) index to a learnable embedding, applies a GNN to learn the node-specific representations based on these learnable embeddings and finally aggregates the representations of the target users and its corresponding item nodes to predict missing links. However, without node content (i.e., side information) for training, the user (or item) specific representation can not be learned in the inductive setting, that is, a model trained on one group of users (or items) cannot adapt to new users (or items). To this end, we propose an inductive matrix completion method using GAE (IMC-GAE), which utilizes the GAE to learn both the user-specific (or item-specific) representation for personalized recommendation and local graph patterns for inductive matrix completion. Specifically, we design two informative node features and employ a layer-wise node dropout scheme in GAE to learn local graph patterns which can be generalized to unseen data. The main contribution of our paper is the capability to efficiently learn local graph patterns in GAE, with good scalability and superior expressiveness compared to previous GNN-based matrix completion methods. Furthermore, extensive experiments demonstrate that our model achieves state-of-the-art performance on several matrix completion benchmarks.
This paper considers the problem of practical Heterogeneous wireless charger Placement with Obstacles (HIPO), i.e., given a number of heterogeneous rechargeable devices distributed on a 2D plane where obstacles of arbitrary shapes exist, deploying heterogeneous chargers with a given cardinality of each type, i.e., determining their positions and orientations, the combination of which we name as strategies, on the plane such that the rechargeable devices achieve maximized charging utility. After presenting our practical directional charging model, we first propose to use a piecewise constant function to approximate the nonlinear charging power, and divide the whole area into multi-feasible geometric areas in which a certain type of chargers have constant approximated charging power. Next, we propose the Practical Dominating Coverage Set extraction algorithm to reduce the unlimited solution space to a limited one by exacting a finite set of candidate strategies for all multi-feasible geometric areas. Finally, we prove the problem falls in the realm of maximizing a monotone submodular function subject to a partition matroid constraint, which allows a greedy algorithm to solve with approximation ratio of 1/2 - ε. We conduct experiments to evaluate the performance. Results show that our algorithm outperforms the comparison algorithms by at least 33.49 percent on average.
Internet of Vehicles (IoV) supports multiple traffic services by processing abundant data from sensors and video surveillance devices. With edge computing, video surveillance services can be certainly improved due to the handy resource provision for video storage and processing. Generally, to reduce the hardware and maintenance investment, it is a popular manner to deploy the limited amount of edge nodes along with the surveillance devices. However such edge node layout leads to the unstable service distribution and complicated data transmission across the surveillance devices and edge nodes, which consequently decreases the quality of the surveillance services. In addition, the service trustworthiness is suspected since the privacy information may be revealed to some extent during the data transmission. To combat these challenges, a trust-aware task offloading method (TOM) for video surveillance in edge computing enabled IoV is presented for minimizing the response time of the services, achieving the load balance of the edge nodes and realizing privacy protection. Technically, SPEA2 (improving the strength Pareto evolutionary algorithm) is employed to acquire balanced task offloading solutions. Then, TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) and MCDM (Multiple Criteria Decision Making) are exercised to ascertain the optimal solution. Finally, the experimental simulation demonstrates that TOM performs efficient and trust.
Using Web APIs registered in service sharing communities for mobile APP development can not only reduce development period and cost, but also fully reuse state-of-the-art research outcomes in broad domain so as to ensure up-to-date APP development and applications. However, the big volume of available APIs in Web communities as well as their differences make it difficult for APIs selection considering compatibility, preferred partial APIs and expected APIs functions which are often of high variety. Accordingly, how to recommend a set of functional-satisfactory and compatibility-optimal APIs based on the APP developer's multiple function expectation and pre-chosen partial APIs is on demand as a significant challenge for successful APP development. To address this challenge, we first construct a Web APIs correlation graph that incorporates functional descriptions and compatibility information of Web APIs, and then propose a correlation graph-based approach for personalized and compatible Web APIs recommendation in mobile APP development. Finally, through extensive experiments on a real dataset crawled from Web APIs websites, we prove the feasibility of our proposed recommendation approach.
The Industrial Internet of Things (IIoT) is a subset of the broader IoT concept, which specifically emphasizes the integration of smart devices and automation technologies within industrial settings. IIoT represents a significant shift in how humans approach manufacturing, logistics, and supply chain management by leveraging real-time monitoring, automation, and data analytics. One of the main advantages of IIoT is predictive maintenance, which involves the use of real-time data analytics and other machine learning algorithms to detect potential equipment failures before they occur. This approach can mitigate downtime, save costs, and increase efficiency. In addition, IIoT can facilitate remote monitoring, allowing for real-time tracking of production processes and equipment performance, enabling better decision-making, and reducing waste. IIoT also has implications for supply chain management, by providing end-to-end visibility into supply chain processes, enabling greater tracking capabilities, and facilitating agile manufacturing approaches. This has the potential to significantly reduce inventory costs and improve customer satisfaction. In summary, although IIoT plays a significant role in optimizing multiple aspects for organizations, the success of IIoT depends on thorough consideration of organizational needs, challenges, and goals. In the following sections, the book first introduces the resource scheduling problem in IIoT to achieve better resource utilization and reduce scheduling costs. Then the security enforcement and privacy-preserving in IIoT are discussed, and the final section introduces business process management.
Fog computing is emerging as a powerful and popular computing paradigm, which extends cloud computing paradigm to enable the service execution in the edge network. The mobile and IoT (Internet of Things) applications could choose the computing nodes in both fog and cloud for resource provisioning. Generally, load balancing is one of the key factors to achieve resource efficiency and avoid bottleneck, overloaded and low-load resource usage. However, it is still a challenge to realize the load balancing for the computing nodes in the fog-cloud environment. In view of this challenge, a Virtual Machine (VM) scheduling method for load balancing in fog-cloud computing is proposed in this paper. Technically, a resource model and a load balancing model are analyzed first. Then a heuristic VM scheduling method is designed through VM placement and dynamic VM scheduling by leveraging the VM live migration technique. Consequentially, experimental evaluation and comparison analysis are conducted to validate the efficiency and effectiveness of our proposed method.
The science process is a process to solve a complex scientific problem which involves large numbers of resources and scientists. Organizing science processes in workflow forms and deploying such scientific workflows in grid environments could provide the effective process management and resource sharing so as to achieve the automatic performance of science processes. However, most current workflow models rarely take account of specific characteristics of science processes and are not very suitable for scientific workflow modeling. Therefore, a new model special for science processes, named as the problem-based scientific workflow model, is proposed in this paper. It takes subproblems as basic modeling elements and puts forward three new relationships based on the problem logic analysis. Moreover, based on this model which conforms to characteristics of science processes, the mechanism about workflow performance in grid environments is provided
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
