Abstract Link prediction aims at predicting the missing links or new links based on known topological or attribute information of networks, which is one of the most significant and challenging tasks in complex network analysis. Recently, many local similarity-based methods have been proposed and they performed well in most cases. However, most of these methods simultaneously ignore the contributions of the local structure information between endpoints and their common neighbors, as well as transmission abilities of different 3-hop paths. To address these issues, in this paper, we propose a novel link prediction method that aims at improving the prediction accuracy of the existing local similarity-based methods by integrating with local structure information and node degree information along 3-hop paths. Extensive experiments have been performed on nine real-world networks and the results demonstrate that our proposed method is superior to the existing state-of-the-art methods.

Abstract The influence maximization (IM) problem has received great attention in the field of social network analysis, and its analysis results can provide reliable basis for decision makers when promoting products or political viewpoints. IM problem aims to select a set of seed users from social networks and maximize the number of users expected to be influenced. Most previous studies on the IM problem focused only on the single-objective problem of maximizing the influence spread of the seed set, ignoring the cost of the seed set, which causes decision makers to be unable to develop effective management strategies. In this work, the IM problem is formulated as a multi-objective IM problem that considers the cost of the seed set. An improved multi-objective particle swarm optimization (IMOPSO) algorithm is proposed to solve this problem. In the IMOPSO algorithm, the initialization strategy of Levy flight based on degree value is used to improve the quality of the initial solution, and the local search strategy based on greedy mechanism is designed to improve the Pareto Frontier distribution and promote algorithm convergence. Experimental results on six real social networks demonstrate that the proposed IMOPSO algorithm is effective, reducing runtime while providing competitive solutions.

High-speed railway (HSR) is developing toward automation and intelligence with the motivation of improved safety. The automatic train control system is deeply committed to HSR safety, and depends on long term evolution for railway (LTE-R). The authentication scheme is the primary contributor among all security measures to secure communication in LTE-R. We found vulnerabilities in the authentication mechanism used in LTE-R, such as key leakage, inability to resist privileged user attacks, and large authentication latency. To address these vulnerabilities, we propose a secure and efficient access authentication scheme for LTE-R. Unlike other studies, Our proposal capitalizes on the utilization of blockchain to establish a decentralized authentication system, effectively eliminating the single point of collapse vulnerability. Recognizing the public nature of the blockchain, we integrate the secgear framework to provide robust privacy protection measures. Furthermore, our scheme augments LTE-R access authentication by introducing a handover authentication phase and a password change phase, introducing additional security attributes for enhanced protection. In addition, we have conducted fleshed-out security analysis and simulation experiments, which show that our scheme is able to achieve privacy protection with acceptable efficiency.

Federated learning (FL) has emerged as a popular machine learning paradigm recently. Compared with traditional distributed learning, its unique challenges mainly lie in communication efficiency and non-IID (heterogeneous data) problem. While the widely adopted framework FedAvg can reduce communication overhead significantly, its effectiveness on non-IID data still lacks exploration. In this paper, we study the non-IID problem of FL from the perspective of domain adaptation. We propose a distribution regularization for FL on non-IID data such that the discrepancy of data distributions between clients is reduced. To further reduce the communication cost, we devise two novel distributed learning algorithms, namely rFedAvg and rFedAvg+, for efficiently learning with the distribution regularization. More importantly, we theoretically establish their convergence for strongly convex objectives. Extensive experiments on 4 datasets with both CNN and LSTM as learning models verify the effectiveness and efficiency of the proposed algorithms.

MOTIVATION: Extracting useful molecular features is essential for molecular property prediction. Atom-level representation is a common representation of molecules, ignoring the sub-structure or branch information of molecules to some extent; however, it is vice versa for the substring-level representation. Both atom-level and substring-level representations may lose the neighborhood or spatial information of molecules. While molecular graph representation aggregating the neighborhood information of a molecule has a weak ability in expressing the chiral molecules or symmetrical structure. In this article, we aim to make use of the advantages of representations in different granularities simultaneously for molecular property prediction. To this end, we propose a fusion model named MultiGran-SMILES, which integrates the molecular features of atoms, sub-structures and graphs from the input. Compared with the single granularity representation of molecules, our method leverages the advantages of various granularity representations simultaneously and adjusts the contribution of each type of representation adaptively for molecular property prediction. RESULTS: The experimental results show that our MultiGran-SMILES method achieves state-of-the-art performance on BBBP, LogP, HIV and ClinTox datasets. For the BACE, FDA and Tox21 datasets, the results are comparable with the state-of-the-art models. Moreover, the experimental results show that the gains of our proposed method are bigger for the molecules with obvious functional groups or branches. AVAILABILITY AND IMPLEMENTATION: The code and data underlying this work are available on GitHub at https://github. com/Jiangjing0122/MultiGran. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Minimax similarity stresses the connectedness of points via mediating elements rather than favoring high mutual similarity. The grouping principle yields superior clustering results when mining arbitrarily-shaped clusters in data. However, it is not robust against noises and outliers in the data. There are two main problems with the grouping principle: first, a single object that is far away from all other objects defines a separate cluster, and second, two connected clusters would be regarded as two parts of one cluster. In order to solve such problems, we propose robust minimum spanning tree (MST)-based clustering algorithm in this letter. First, we separate the connected objects by applying a density-based coarsening phase, resulting in a low-rank matrix in which the element denotes the supernode by combining a set of nodes. Then a greedy method is presented to partition those supernodes through working on the low-rank matrix. Instead of removing the longest edges from MST, our algorithm groups the data set based on the minimax similarity. Finally, the assignment of all data points can be achieved through their corresponding supernodes. Experimental results on many synthetic and real-world data sets show that our algorithm consistently outperforms compared clustering algorithms.

Context-Aware Recommender System (CARS) aims to not only recommend services similar to those already rated with the highest score, but also provide opportunities for exploring the important role of temporal, spatial and social contexts for personalized web services recommendation. A key step for temporal-based CARS methods is to explore the time decay process of past invocation records to make the Quality of Services (QoS) prediction. However, it is a nontrivial task to model the temporal effects on web services recommendation, due to the dynamic features of contextual information in view of temporal spatial correlations. For instance, in location-aware services recommendation, the user's geographical position would change very frequently as time goes on. In this paper, we propose a Context-Aware Services Recommendation based on Temporal Effectiveness (CASR-TE) method. Inspired by existing time decay approaches, we first present an enhanced temporal decay model combining the time decay function with traditional similarity measurement methods. Then, we model temporal spatial correlations as well as their impacts on the user preference expansion. Finally, we evaluate the CASR-TE method on WS-Dream dataset by evaluation matrices of both RMSE and MAE. Experimental results show that our approach outperforms several benchmark methods with a significant margin.

Context-Aware Recommender System aims to recommend items not only similar to those already rated with the highest score, but also that could combine the contextual information with the recommendation process. Existing context-aware Web services recommendation methods directly use context as a "filter" to discard services that may conflict with the current user's preference. However, the discarded services may be valuable for another user or for the same user under a new context, as one man's trash may be another's treasure. We assume that failing to handle the contextual reasons behind the user preference may introduce inaccurate recommendation, and even significant biases in recommendation. In this work, we propose a novel method dubbed CASR-UP, which aims to exploit the contextual factors of the user preference to improve Quality of Service (QoS) prediction and services recommendation accuracy. Our method consists of three stages: 1) context-aware similarity mining to get the set of users having similar context with the current user, 2) data filtering based on user preference in current context so as to get the invocation records of the services corresponding to the current user's preference, 3) Web services QoS prediction, recommendation and evaluation by Bayesian Inference. Experimental results on WS-Dream dataset is evaluated by both RMSE and MAE. The results show the proposed method improves prediction accuracy and outperforms the compared methods.

With the further integration of the informatization and industrialization, it is a key factor for enterprise to develop them by enterprise information construction. At present, with the development of computer, enterprise informatization generally has formed three layers architecture, which contains: ERP (Enterprise Resource Planning), MES (Manufacturing Execution System) and PCS (Process Control System). MES plays an irreplaceable position in the enterprise informatization. However, different needs of the design for these systems. Then, they lead system not to share information, business process effectively and work together with others. Then finally lead to "Information Island". It has made people to pay more attention to the topic, which is how to implement the exchange of information real-time and accurately between ERP, PCS and MES. So, we analyze some major exchange ways of the present information between MES and other systems in this paper. Then, we propose a flexible and configurable EAI architecture-AiOintegration (All-in-One integration), which has integrated the current methods of exchanging between system information. And AiOintegration implemented information exchange of cross-language and cross-platform between with MES and other systems.

Traditional virtual screening in the grid needs chemists to upload small molecule files and collect the results manually, which cannot implement docking and collection of results automatically. This caused heavy workload to chemists. In this paper, we took advantage of Hadoop platform in the massive data storage. We stored and managed small molecule files and docking results files using HDFS. In addition, MapReduce programming framework is used for parallel molecular docking to preliminarily process results files, in order to achieve the automation of the virtual screening molecular docking. The research of this thesis will be helpful to drug researcher by offering a massive data storage management system for large-scale virtual screening, and will also provide a reference for drug discovery in the cloud environment to promote the development of computational chemistry e-science.

Scientific workflows (SWFs) aim to automate cooperative design through compilation of known sequences of actions for routine procedures. However, current SWF systems lack the ability, in the one hand, to capture the context in which a SWF is designed and developed, and, in the other hand, to deliver a real-time assistance to help designers to make effective decisions in the selection of relevant SWFs for their problem. We propose a context-oriented framework for improving cooperative SWF design. This framework allows making context explicit and realizing the contextualization of SWFs in a SWF repository (and thus sharable with other scientists). Context is made explicit thanks to Contextual Graphs (CxGs), a formalism for representing uniformly all the ingredients of a cooperative SWF design process. Thus, scientists can customize information and formalize their research and strategies in a shared context. We use the context-oriented framework in an application in virtual screening. Finally, context-based formalisms, such as CxGs, appear as the key element to enhance users' cooperation during the SWF design phase.

E-Research aims to facilitate collaboration across time and distance. Researchers need techniques and tools to support their collaborative work. Groupware is one technique that supports groups of people engaging in a common task over the network. Besides, it is also one of the most effective means to solve the collaboration problem. Existing groupware projects provide fixed functions such as messaging, conferencing, electronic meeting, document management, document collaboration and so on. However, they put limited emphasis on scientistspsila research work in their specific fields. This paper proposes a groupware environment, and tries to give a domain-specific group editor to facilitate researcherspsila collaboration. The groupware implements a visual molecule group editor for chemists to co-edit molecular structures over network. And it has a plug-in extensible architecture intending to easily integrate other tools which is useful for chemistspsila collaboration. The idea given in this paper could be a possible solution to facilitate chemistspsila collaborative research work.

SWF, short for scientific workflow, has recently emerged as a paradigm for orchestrating large-scale e-Science applications. SWF specification connects workflow designer with workflow engine, which makes it act ass one of the key components in SWF systems. We adopted BPEL as the Gird-based SWF specification because of its potential benefits to promote SWF sharing and reproducibility. However, grid-based SWF has unique features, which pure BPEL is not capable to handle. Appropriate ways to extend BPEL to fulfill the special needs should be found. We concluded three most necessary requirements and three kinds of alternative methods to extend BPEL within China Grid Support Platform, and implemented the method of adding additional abstractions upon BPEL to enhance user experience. Our work and findings suggest that our extension approach is feasible and would manifest potential advantage to SWF systems and finally innovate in the e-Science research and applications.

Studying in how to develop a Gaussian syntax-analyzer is significant for there are many researches on chemistry soft wares. The paper develops a Gaussian analyzer using ANTLR, implementing of Gaussian syntax-analyzer using ANTLR is good at analyzing other chemistry software using ANTLR, as well as the ANTLR is discussed on how to implement syntax-analyzer. Test the Gaussian syntax-analyzer using an example as well as build a AST. Whilst the paper improve one kind of syntactical parser methods.