Network computing and applications for Big Data analytics

Network computing and applications for Big Data analytics

Journal of Network and Computer Applications 59 (2016) 361 Contents lists available at ScienceDirect Journal of Network and Computer Applications jo...

164KB Sizes 2 Downloads 159 Views

Journal of Network and Computer Applications 59 (2016) 361

Contents lists available at ScienceDirect

Journal of Network and Computer Applications journal homepage: www.elsevier.com/locate/jnca

Editorial

Network computing and applications for Big Data analytics Big Data presents myriad possibilities for transforming science, engineering, medicine, healthcare, finance, business and ultimately society itself. It offers significant opportunities for organizations to obtain critical intelligence to drive decisions and obtain insights as never before (Bhattacharya et al., 2015). However, the volume and variety of data mandates development of new approaches and applications grounded in emerging analytics techniques and emerging processing technologies within the Big Data paradigm (Abawajy, 2015). This special issue presents original contributions in the area of network computing and applications for Big Data. With the exponential increase in an online information, how to find information of interest has become a serious challenge. This problem necessitated the development of personalized online information services. Xu et al. (2015) proposed a novel two-stage approach for R&D project opportunity recommendation to interested researchers and practitioners. At the top level, a filtering approach is used to identity appropriate R&D projects as a candidate set for researchers and practitioners. The second level deploys an information aggregation method with various constraints to recommend suitable R&D projects to interested researchers and practitioners. Amini et al. (2015) addressed the problem of clustering data streams. They propose a method called the MuDi-Stream which combines an online phase and an offline phase with four main components. In the online phase, summary information about the evolving multi-density data stream is maintained in the form of core mini-clusters. In the offline phase, the final clusters are generated using an adapted density-based clustering algorithm. MuDiStream employs the grid-based method as an outlier buffer to deal with both noises and multi-density data as well as to reduce the time complexity of combining the clusters. Empirical evaluation of MuDi-Stream on various synthetic and real-world datasets using different quality metrics shows that MuDi-Stream improves clustering quality in multi-density environments. Villaça et al. (2015) discuss a system called HCube that can be used for routing and similarity search in Data Centers. In contrast to Data Centres (DCs) based data storage/retrieval, HCube data storage/retrieval is based on the similarity search where similar content is collected on servers physically close within the HCube. A three-dimensional structure is used to arrange the HCube network. HCube uses the Gray Space Filling Curve (SFC) in combination with the Random Hyperplane Hashing (RHH) function and the XOR-based flat routing mechanism for the similarity search. Mershad et al. (2015) conducted analysis and performance evaluation of the Reconfigurable Active SSD (RASSD) Middleware Server (MWS), which combines active solid state drives and reconfigurable FPGAs into a storage-compute node that can be used by a cloud data center to achieve accelerated computations while running data intensive applications. They studied the utilization of three important elements of the MWS: CPU, memory, and network bandwidth. For each, we derive the parameters that govern and affect its operations, and propose formulas for its utilization factor. We use the analysis results, http://dx.doi.org/10.1016/j.jnca.2015.11.007 1084-8045/& 2015 Published by Elsevier Ltd.

while applying different values to the system parameters, to illustrate important benefits and limitations of the system. Big Data pose new computational challenges including very high dimensionality and sparseness of data. Bhattacharya et al. (2015) addresses this challenge and presents an evolutionary algorithm with enhanced ability to deal with the problems of high dimensionality and sparseness of data. In addition to an informed exploration of the solution space, the proposed approach uses informed genetic operators in order to balance exploration and exploitation using a hierarchical multi-population approach. In order to deal with the problem of high dimensionality, a multi-tier hierarchical architecture is employed. Empirical study of the proposed technique on some benchmark problems shows the superior performance of the proposed approach as compared to the baseline approaches. The algorithm has also been successfully applied to a real world problem of financial portfolio management.

References Abawajy Jemal. Comprehensive analysis of big data variety landscape. Int J Parallel Emerg Distrib Syst 2015;30:5–14 Taylor & Francis. Amini Amineh, Saboohi Hadi, Herawan Tutut, Ying Wah Teh. MuDi-Stream: a multi density clustering algorithm for evolving data stream. J Netw Comput Appl 2015;59:365–80. Bhattacharya Maumita, Islam Rafiqul, Abawajy Jemal. Evolutionary optimization: a big data perspective. J Netw Comput Appl 2015;59:411–21. Mershad Khaleel, Artail Hassan, Saghir Mazen, Hajj Hazem, Awad Mariette. A mathematical model to analyze the utilization of a cloud datacenter middleware. J Netw Comput Appl 2015;59:394–410. Villaça RS, Pasquini R, de Paula LB, Magalhães MF. HCube: routing and similarity search in Data Centers. J Netw Comput Appl 2015;59:381–93. Xu Wei, Sun Jianshan, Ma Jian, Du Wei. A personalized information recommendation system for R&D project opportunity finding in big data contexts. J Netw Comput Appl 2015;59:357–64.

Jemal H. Abawajy Deakin University, School of Information Technology, Faculty of Science, Engineering and Built Environment, Waurn Ponds Campus, Locked Bag 20000, Geelong, Vic. 3220, Australia

Albert Y. Zomaya School of Information Technologies, University of Sydney, Sydney, NSW 2006, Australia

Ivan Stojmenovic Deakin University, School of Information Technology, Melbourne, Vic. 3125, Australia