面向多源异构数据的云推送平台的研究与应用

在传统推送中,多源异构数据的推送面临时效性不强、安全性不高、数据难以重用等问题。针对多源异构数据的特征,综合移动互联网的安全性和隐私性等特点,提出了一种多维决策云推送模型来计算分布式环境中多源异构数据的特征值和特征向量,以快速分离数据源中的同构数据和异构数据;并基于此模型设计了云推送平台,其利用云推送技术来实现同构数据和异构数据的自动分离和高效推送。根据云推送平台在实验环境中的运行情况及相关指标分析,说明该平台适用于多源异构数据的推送,是一种高效可行的推送方式。     

基于云数据中心的多源异构数据治理技术研究

目前常规的多源异构数据治理方法主要通过对数据属性进行判断,从而实现分区域数据清洗,由于缺乏对非线性数据的分析,导致治理性能不佳;对此,提出基于云数据中心的多源异构数据治理技术;采用关系型数据库中的ETL功能对数据进行清洗,对数据转换模式以及数据清洗规则进行定义;引入互信息系数对数据相关程度进行判定,并进行非线性数据相关性分析;以云数据中心作为载体,对多源异构数据治理体系进行构建;在实验中,对提出的数据治理技术进行了治理性能的检验;最终的实验结果表明,提出的数据治理技术具备较高的查准率,对云数据中心多源异构数据具备较为理想的数据治理效果。     

基于QoS预测与约束层次模型的云服务推荐方法研究

如何精确地预测云服务的QoS,灵活地处理用户的QoS约束是实现可靠云服务推荐的关键问题。针对上述问题,提出了一种基于事例推理的云服务QoS动态预测方法,用于预测候选云服务的QoS值,提高云服务质量的可靠性;提出了一种基于约束层次（Constraint Hierarchy,CH）的QoS评价模型,该模型依据QoS约束对用户业务的重要程度,将其分为必须满足的硬约束层和允许被偏离的软约束层,该模型能够有效地解决过约束问题。通过实验验证了提出的云服务推荐方法的可行性和有效性。     

基于用户浏览行为的个性化推荐研究综述

研究如何充分利用海量用户浏览行为数据,构建更加精确的推荐算法和模型,以提高推荐系统性能,是目前个性化推荐领域研究的热点.针对这些问题,首先对用户的浏览行为进行了简要概括表述,给出了基于浏览行为推荐系统的总体框架,回顾总结了基于用户浏览行为的推荐系统的发展历程.对其关键技术和单一浏览行为量化方法与混合浏览行为量化方法进行总结、对比和分析.最后讨论了结合多源异构数据的浏览行为推荐的最新成果,总结了该领域未来研究难点和发展趋势.     

基于PDCA模型的云资源管理方法研究

随着云计算技术的发展,企业级的云计算平台在数字化浪潮中占据越来越重要的地位。云资源管理是云平台运行的关键环节,要在确保云平台稳定运行的基础上提升云资源利用效率,目前大多数企业级平台已经历大规模建设阶段,在使用阶段暴露出闲置资源多、资源分配不均、缺少管理要求及统一规范等问题,故亟需构建一套科学的云资源管理方法论,实现云资源精益管理。在PDCA模型的基础上,提出了云资源管理方法,拓展管理者对云资源全过程管理的视野,挖掘痛点问题,为企业级云计算平台长效化管理提供参考。     

基于深度学习的内容推荐算法研究综述

推荐系统是学习用户偏好,实现个性化推荐的系统化应用技术,在商品购买、影音推荐、关联阅读等多领域得到了广泛的应用.近年来,随着多源异构数据的激增和深度学习的兴起,传统推荐算法中的表征学习模式逐步被深度学习代替.梳理推荐算法的背景和发展趋势,并给出内容推荐的算法思路及其优劣评价,分别介绍多层感知机、自动编码器、卷积神经网络...     

基于云的数据库的研究

随着云计算的逐步发展,云数据库的也越来越重要,在未来数据库领域占有重要地位。文章介绍了云数据的现状、发展、特性、影响。最后讨论了云数据库的未来的发展方向。     

云计算下的多源信息资源云体系及云服务模型研究

针对传统信息服务模式存在信息资源规模单一、虚拟化程度低、自适应程度较差及知识泛化等问题, 同时考虑到目前云服务体系中忽略多源信息资源这一重要因素, 提出了云计算环境下的多源信息资源云体系及服务模型。首先构建了基于云计算的多源信息资源云体系, 对信息资源搜索、查询及匹配等技术进行阐述, 进而提出了新的云服务模型——基于云计算的多源信息资源即服务（multi-source information resource as a service, MIRaaS）, 在此基础上搭建了该模型在商务领域下的应用平台。分析表明, 该模型能对多源信息资源进行重构、分析、整合及匹配, 并充分考虑用户体验, 很好地扩展了云服务的研究范围及应用领域。     

基于云模型的协同过滤推荐算法

协同过滤系统是电子商务系统中最重要的技术之一,用户相似性度量方法是影响推荐算法准确率高低的关键因素.针对传统相似性度量方法存在的不足,利用云模型在定性知识表示以及定性、定量知识转换时的桥梁作用,提出一种在知识层面比较用户相似度的方法,克服了传统基于向量的相似度比较方法严格匹配对象属性的不足.以该方法为核心,在全面分析传统方法的基础上,提出一种新的协同过滤推荐算法.实验结果表明,算法在用户评分数据极端稀疏的情况下,仍能取得较理想的推荐质量.     

基于双维度云模型的协同过滤推荐算法

随着网络的迅速普及,网络应用多样化趋势加快,大数据时代已经到来,电商业对于在线推荐系统的要求也越来越高。为了满足人们的需求,传统的推荐算法必须得到改进和发展。本文主要考虑数据比较稀疏及数据规模较大的情况下如何更好地实现协同过滤推荐算法,针对这两个方面的不足,结合Hadoop平台的优势,基于双维度云模型的协同过滤推荐算法由此应运而生了。通过实验,利用云模型和双维度数据,有关数据稀疏性问题得到了合理的解决,预测准确率方面也得到了一定的改进。并且通过MapReduce处理流程,使大数据环境下运行困难效率低下的缺陷得到了弥补。     

云加端的嵌套滑动窗口故障信号在线检测方法研究*

针对当前制造业生产线设备的故障检测效率以及检测方法通用性不高的问题,本文提出了一种云端融合的动态嵌套滑动窗口故障信号在线检测算法。该算法采用云服务跟智能终端在线检测相结合的架构,利用云服务大存储量和高计算速度、精度的优势,解决了终端设备对故障信号处理能力不足以及仅能对线上数据进行单次扫描的问题。云计算中心根据数据流的波动情况初步确定滑动窗口大小,再根据对异常信号的判断,向智能终端反馈故障信号的大小和相对位置,通过动态嵌套滑动窗口对其进行定位。理论分析和实验结果表明,该方法对周期信号有较好的通用性,而且有效提高了故障检测的效率。     

A cloud supported model for efficient community health awareness

The needs for efficient and scalable community health awareness model become a crucial issue in today’s health care applications. Many health care service providers need to provide their services for long terms, in real time and interactively. Many of these applications are based on the emerging Wireless Body Area networks (WBANs) technology. WBANs have developed as an effective solution for a wide range of healthcare, military, sports, general health and social applications. On the other hand, handling data in a large scale (currently known as Big Data) requires an efficient collection and processing model with scalable computing and storage capacity. Therefore, a new computing paradigm is needed such as Cloud Computing and Internet of Things (IoT). In this paper we present a novel cloud supported model for efficient community health awareness in the presence of a large scale WBANs data generation. The objective is to process this big data in order to detect the abnormal data using MapReduce infrastructure and user defined functions with minimum processing delay. The goal is to have a large monitored data of WBANs to be available to the end user or to the decision maker in reliable manner. While reducing data packet processing energy, the proposed work is minimizing the data processing delay by choosing cloudlet or local cloud model and MapReduce infrastructure. So, the overall delay is minimized, thus leading to detect the abnormal data in the cloud in real time mode. In this paper we present a multi-layer computing model composed of Local Cloud (LC) layer and Enterprise Cloud (EP) layer that aim to process the collected data from Monitored Subjects (MSs) in a large scale to generate useful facts, observations or to find abnormal phenomena within the monitored data. Performance results show that integrating the MapReduce capabilities with cloud computing model will reduce the processing delay. The proposed MapReduce infrastructure has also been applied in lower layer, such as LC in order to reduce the amount of communications and processing delay. Performance results show that applying MapReduce infrastructure in lower tire will significantly decrease the overall processing delay.     

Time efficient secure DNA based access control model for cloud computing environment

The uses of Big Data (BD) are gradually increasing in many new emerging applications, such as Facebook, eBay, Snapdeal, etc. BD is a term, which is used for describing a large volume of data. The data security is always a big concern of BD. Besides the data security, other issues of BD are data storage, high data accessing time, high data searching time, high system overhead, server demand, etc. In this paper, a new access control model has been proposed for BD to solve all these issues, where fast accessing of the large volume of data are provided based on the data size Here, a long 512-bit Deoxyribonucleic Acid (DNA) based key sequence has been used for improving the data security, and it is secured against the collision attack, man-in-the-middle attack, internal attack, etc. The proposed scheme is evaluated in terms of both theoretical and experimental results, which show the proficiency of the proposed scheme over the existing schemes.     

A cloud based health insurance plan recommendation system: A user centered approach

The recent concept of “Health Insurance Marketplace” introduced to facilitate the purchase of health insurance by comparing different insurance plans in terms of price, coverage benefits, and quality designates a key role to the health insurance providers. Currently, the web based tools available to search for health insurance plans are deficient in offering personalized recommendations based on the coverage benefits and cost. Therefore, anticipating the users’ needs we propose a cloud based framework that offers personalized recommendations about the health insurance plans. We use the Multi-attribute Utility Theory (MAUT) to help users compare different health insurance plans based on coverage and cost criteria, such as: (a) premium, (b) co-pay, (c) deductibles, (d) co-insurance, and (e) maximum benefit offered by a plan. To overcome the issues arising possibly due to the heterogeneous data formats and different plan representations across the providers, we present a standardized representation for the health insurance plans. The plan information of each of the providers is retrieved using the Data as a Service (DaaS). The framework is implemented as Software as a Service (SaaS) to offer customized recommendations by applying a ranking technique for the identified plans according to the user specified criteria.     

基于云的计算机等级考试架构

传统的计算机等级考试安装系统和服务器花费大量人力物力和时间,部署及更改系统相对困难,并且每个考场均需安装一台服务器,为防止系统崩溃和软件故障还需备份一台服务器,服务器利用率较低;目前国内数据中心及云计算平台的兴起,采用虚拟化的技术使得在物理服务器上构架虚拟服务器和桌面成为了现实,采用云计算平台的计算机等级考试系统可以改变传统考试系统的多种弊端,节约资源,提高硬件的使用率和考试系统的安全性、稳定性.     

基于云计算的图书馆云服务模型研究

通过借鉴OCLC在云计算图书馆的设计思路以及基于云计算整合图书馆资源与日常服务的办法,提出了在云计算图书馆云服务的构建思路及其模型。     

基于云计算的网络安全的探析

随着云计算的快速的发展,越来越多的用户将应用和数据托管到了云端。但是由于云计算具有的虚拟化、多用户、可伸缩等新特性,使得传统的安全技术并不能保证云计算的安全,云计算中的安全问题成了阻碍云计算发展的最主要的问题之一。因此,云计算的安全性成为了当前研究云计算的重点。     

基于云模型的用户双重聚类推荐算法

协同过滤是一种应用广泛的推荐算法,但存在着效率低和数据稀疏等问题。为解决这些问题,提出了一种改进的聚类推荐算法。该算法引用云模型,先从项目属性和用户属性两方面计算云模型期望、熵和超熵,并考虑到评分时间、评分高低和评分习惯等因素的影响,建立用户兴趣模型;接着,采用基于云模型的修正相似度量方法进行用户兴趣相似度比较,并使用K-means算法进行聚类;最后,利用参与预测人数的比例对公共项目进行推荐结果合并。在MovieLens上的实验结果表明,该算法不仅可以解决效率低和数据稀疏等问题,还提高了推荐的准确性。     

A hybrid approach for scalable sub-tree anonymization over big data using MapReduce on cloud

In big data applications, data privacy is one of the most concerned issues because processing large-scale privacy-sensitive data sets often requires computation resources provisioned by public cloud services. Sub-tree data anonymization is a widely adopted scheme to anonymize data sets for privacy preservation. Top–Down Specialization (TDS) and Bottom–Up Generalization (BUG) are two ways to fulfill sub-tree anonymization. However, existing approaches for sub-tree anonymization fall short of parallelization capability, thereby lacking scalability in handling big data in cloud. Still, either TDS or BUG individually suffers from poor performance for certain valuing of k-anonymity parameter. In this paper, we propose a hybrid approach that combines TDS and BUG together for efficient sub-tree anonymization over big data. Further, we design MapReduce algorithms for the two components (TDS and BUG) to gain high scalability. Experiment evaluation demonstrates that the hybrid approach significantly improves the scalability and efficiency of sub-tree anonymization scheme over existing approaches.