Online Trends

Design and implementation of user behavior analysis system based on Hadoop Beijing Jiaotong University Shang This subject under the background of large data, aiming at the problem of not fully and accurately analyzing network user behavior, using network security to develop package libnids and distributed platform Hadoop key technology, This paper focuses on the design and development of user behavior Analysis system based on Hadoop. The system realizes the functions of mass data packet grabbing and distributed storage, TCP recombination and Application layer HTTP behavior analysis, which not only helps the service provider to provide better recommendation according to the user's behavior characteristics.

Design and implementation of Water conservancy Census data mining system based on Hadoop cloud platform Fanlong Wandingshen Gu Xinzhen aiming at the characteristics of large amount of data and attribute dimension of Water Conservancy census, this paper presents a data mining system of Water Conservancy survey based on Hadoop cloud platform by discussing and analyzing the composition and operation mechanism of Hadoop. This paper introduces the design idea and structure of the system, and expounds the function of each module in detail. The system utilizes the Hadoop cloud platform to implement an improved parallel association rule algorithm, which makes full use of the Hadoop platform to process data in parallel.

Cloud distributed BPEL engine placement mechanism based on K Lin Rongheng Wu Budan Zhaoyao Yang Spring Release the research on the placement of distributed BPEL engine in the cloud, a distributed BPEL engine placement mechanism based on K is proposed, This mechanism puts the BPEL engine placement problem into the relevant optimization mathematical model, and maps the model to the K algorithm for solving. This mechanism also discusses the application of the algorithm in different network topology random graph and tree network topology. Finally, the related experiment simulation is carried out by the statistic software R, and the simulation knot ...

The research of BP Neural network genetic algorithm based on MapReduce in nonlinear system identification Chen Chunping Chaya champion Shandandan often encounter complex s nonlinear systems in engineering applications, which are complex and difficult to model accurately by mathematical methods. The three-layer BP neural network can approximate continuous function with arbitrary precision. But the BP network has the disadvantage of falling into the local optimal value, and then the genetic algorithm is added. The traditional training method of the serial BP neural network has great problem when processing the massive data.

Research on mapreduce sequence rule in recommendation system the personalized recommendation system model of two chrysanthemum Guo Jinwei Pijan is usually based on collaborative filtering or content-based and partly based on association rules. These algorithms do not take into account the order of transactions, but in many applications such sequences are important. In this paper, a simple recommendation model based on sequential pattern is proposed, and the MapReduce programming model is combined with the consideration of large-scale data processing. This simple recommendation model can be used to assist in the general ...

Cloud computing Virtual machine Migration system based on Mda_fhmipv6 Chen Zhang Weiyun jianjun Microelectronics and computers migration system of cloud computing virtual machine in the period of 2014 03 based on Mda_fhmipv6

Design and realization of college SNS system based on MongoDB Hu Heng SNS (Social Networking services) social network service, also known as social networking Services, is an Internet application service formed through interpersonal network. Among them, specially for the campus student service SNS website is called the campus SNS website.  This paper introduces the key technology, architecture design scheme and realization of the SNS system in colleges and universities using SPRINGMVC+MONGODB architecture. Design and implementation of SNS system in colleges and universities based on MongoDB

Design and implementation of cloud computing platform based on OpenStack Li Yu, Wang Xiaobo, Fan Jianhua, Wang Mingjiang cloud computing can effectively organize various kinds of computing resources, improve computing performance. The core technology and construction process of the existing mature cloud computing platform is not available for commercial reasons, and based on open source project OpenStack, a cloud computing platform is designed and implemented, and the platform supports multiple types of clients dynamically connected through administrator and user interface to request various configuration virtual machines and obtain computing, storage, Software resources and other services. Cloud virtualization management, scheduling control, storage management 、...

Constructing handwritten recognition data--service cloud platform based on OpenStack Li Desheng the paper proposes the handwritten recognition data, namely the service cloud platform based on the open source cloud computing platform OpenStack IaaS layer (handwriting recognition data Platform as a Service, abbreviation: Hrdpaas. The purpose and significance of this is to provide a reasonable solution to the problems currently encountered by handwriting recognition: 1 Low configuration end (such as embedded, web-side) handwriting recognition ...

Based on OpenStack resource monitoring System Liang Yuyang Li Hongbin Lan Resource monitoring is an important means to improve the reliability of cloud platform.  Based on the characteristics of OpenStack cloud Platform, this paper designs and implements a comprehensive, intelligent and efficient resource monitoring system, completes the design of the overall structure of the resource monitoring system and the function division of each module, and gives the concrete realization methods of the example monitoring mode and the data storage model. Finally, through ...

Mining of massive frequent itemsets based on pfp-growth algorithm Jiangyuan, Li Ping with the development of Internet technology, network data becomes more and more huge, how to excavate effective information has become the focus of people's research. In recent years, frequent itemsets mining has been paid more and more attention because of its important role in the task of Mining Association rules and related mining. In this paper, the algorithm of frequent itemsets mining in distributed computing environment is studied, the PFP-GROWTH algorithm is improved, and the improved pfp-growth algorithm is carried out by MapReduce programming model.

Research on mobile cloud Computing Network Security Service and resource optimization management based on SMDP Southwest Jiaotong University Liang Hongbin Mobile Cloud computing service domains are generally distributed according to geographic location. On the basis of the existing mobile cloud computing resources, we can optimize and distribute the cloud computing resources among the mobile cloud computing service domains, so as to make full use of the existing mobile cloud computing resources, improve the revenue of the whole mobile cloud computing network and the service quality of the mobile cloud computing security Service. ...

Research and improvement of desktop virtualization technology based on Spice protocol Xu Haolan Rain on the basis of discussing the principle of Virtual Desktop transport protocol and the factors that affect the performance of virtual desktops, this paper analyzes the architecture and core technology of Spice protocol, and points out the deficiencies of the present Spice Protocol, The improvement scheme under the condition of QoE is put forward, which will play a guiding role for the development of desktop virtualization technology at present. Research and improvement scheme of desktop virtualization technology based on Spice protocol

Research on virtual machine dynamic migration strategy based on Xen Zhang Pu Liaoning University in this paper, a dynamic migration strategy based on Xen virtual machine is proposed for the Xen platform environment, combining the actual application scenario and the Xen bottom migration. The specific work is as follows: first, in order to carry out virtual machine migration in the cloud computing environment, Xen Migration uses a comparison between passing the page bitmap and skipping the page bitmap to determine whether the memory page is retransmission. For page bitmap comparison brings multiple retransmission increase network transmission overhead, put forward the memory optimization algorithm based on AR model, according to the memory page of the record all modified ...

Virtual machine migration time memory optimization algorithm based on Xen Chen Tingwei Zhang Pu Zhang Zhongqing for virtual machine migration in a cloud computing environment, Xen Migration uses a comparison between passing a page bitmap and skipping a page bitmap to determine whether the memory page is retransmission. The problem of multiple retransmission to increase network transport overhead for page bitmap comparisons A memory optimization algorithm based on AR model is proposed, which predicts the next time of memory page modification according to the memory page modification time interval of all records, and retransmission when the next modification time is greater than a certain threshold. The experimental results show that the base ...

Text classification technology based on parallel computing Zhao Zhe Xiangyang Wang Jisheng in view of the traditional text classification method for the mass data classification speed slow precision problem, parallel computing applied to the field of text categorization, design a set of parallel text categorization framework based on MapReduce, In this paper, a parallel training method of support vector machine based on bagging algorithm is proposed, and experiments on the Hadoop cloud computing platform show that the classification method has fast classification speed and high classification accuracy. Text categorization technology based on parallel computing

The design and development of hospital financial management and decision system based on large data four military medical University Tang This thesis mainly completes the following work: 1. Analysis of the system requirement of hospital financial management and decision making based on large data the data from the existing hospital financial management information system and related subsystems, Through the export of financial and related data and analysis, and in accordance with China's financial software data Interim standard interface for standardization, standardization and customization. Through the establishment of modern hospital financial management framework, to meet the current hospital financial management digital ...

Research on heavy-weight model of sci-tech project based on large data mining Li Shanqing Sean Song Lirong Science and Technology Project check weight is one of the important measures to avoid duplication and repetition construction, and there is no effective method at present. This paper puts forward a project checking method based on large data mining and multi-source information integration, based on the basic information of science and Technology project, publishing thesis information, key words, principals information and undertaking organization, the large data network is the research object, and the similarity degree discriminant model of science and Technology project is constructed by using Multi-Source information integration method, and using Hadoop framework to realize the rapid mining of massive data. ...

Based on distributed k-prototypes algorithm design Li Clustering algorithm has been widely used in many fields, for most datasets, the properties of which are not exactly numerical type, which brings difficulty to clustering. The emergence of K-prototypes algorithm solves the difficulty of hybrid attribute clustering, but its calculation is tedious and brings a lot of difficulties to programmers. The emergence of the Hadoop distributed system provides the possibility for compiling the parallel k-prototypes algorithm, which can improve the program parallelism and greatly improve the program ...

Cloud computing cluster resource scheduling based on improved clonal selection algorithm Zhu Lihua Li Chunhua Wu Quanjen in order to realize cluster resource scheduling and resource load balancing under cloud computing environment, a cloud computing cluster resource scheduling method based on clonal selection algorithm is proposed. Firstly, a cloud computing resource scheduling model is defined, which aims at minimizing execution time span and load balancing factor. Based on this, a clonal selection algorithm is proposed to dispatch the cluster resources under the cloud computing environment, and the antibody coding method, antibody and antibody, and the ...