云计算环境下DaaS生命周期若干关键技术研究

发布时间：2018-10-26 07:28

【摘要】：随着“互联网+”理念的涌现,带来的是各行业的互联网化与现实世界的数据化,使得数据的种类和规模正在以前所未有的速度不断增长和累积,并以其蕴含的丰富价值对社会生产产生巨大推动作用。为了更好地获取数据中潜在的价值,越来越多的企业或个人正在通过云交付的模式,将自身对于大数据的存储、计算与分析能力开放给第三方,通过构建数据服务的形式来对数据进行更好的管理和利用,数据即服务正在成为影响产业格局的新一代业务模式。尽管很多公司、组织、研究人员在数据即服务领域的研究中已取得了丰富的研究成果,但云计算环境下数据资源描述的异构性、数据服务数量的爆炸式增长等因素为数据服务带来了新的挑战:1)随着数据服务描述的多样性,缺乏一种易于描述的数据服务描述模型来实现数据服务的协同工作;2)随着数据服务数量的急速增长,数据服务质量参差不齐,缺乏一种高效的数据服务发现方法;3)随着对数据进行可视分析的需求的增加,可视化开发所需的成本不断提高,缺乏一种对用户体验友好的、提高可视化开发效率的数据可视化方法。为解决以上问题,需要对数据服务的集成方式、描述模型等几个关键方面进行研究。为此,本文对云计算环境下DaaS生命周期相关关键技术展开研究,主要创新工作包括:(1)提出一种基于DaaS的数据服务柔性描述模型。在分析现有数据服务描述语言不足的基础上,对数据服务进行集成,接着给出原子数据服务的柔性描述模型以及复合数据服务的柔性描述模型,并对根据所给出的柔性描述模型进行数据服务的自动化调用。(2)提出一种面向全局社交服务网的数据服务聚类方法。将聚类与全局社交服务网进行结合,给出社交属性、历史社交域、未来社交域等概念,利用服务的社交属性对服务进行聚类,提高服务在全局环境下的发现效率。(3)提出一种基于可视模板的数据服务分析方法。基于软件工程领域的复用思想,给出可视模板等相关定义,利用可视模板高效制作可视化图表以及利用精灵线程、物化缓存、副本迭代等技术实现对用户友好的可视化图表访问体验。本论文的研究内容作为浙江省科技厅公益性技术应用研究计划项目“服务计算环境下的网络大数据平台软件关键技术研究”(编号:2014C33071)和浙江省重大科技专项计划项目“面向产业联盟的特种设备云设计服务平台关键技术研究及其应用”(编号:2014C11SA1B0006)的部分成果,从构建云制造服务平台原型的角度,讨论云计算环境下数据服务生命周期及其关键技术,结合所提出的数据服务柔性描述模型,对数据服务进行集成、管理、组合、调用、数据可视化。
[Abstract]:With the emergence of the concept of "Internet", what brings about is the Internetization of various industries and the digitization of the real world, which makes the variety and scale of data growing and accumulating at an unprecedented speed. And with its rich value of social production has a huge role. In order to better obtain the potential value of data, more and more enterprises or individuals are opening their storage, calculation and analysis capabilities for big data to third parties through the cloud delivery model. By constructing the form of data service to better manage and utilize the data, data as a service is becoming a new generation of business model that affects the industrial pattern. Although many companies, organizations and researchers have made rich achievements in the field of data as a service, the heterogeneity of data resource descriptions in cloud computing environments, The explosive growth of the number of data services brings new challenges to data services: 1) with the diversity of data services description, there is a lack of an easy-to-describe data service description model to achieve the collaborative work of data services; 2) with the rapid growth of the quantity of data service, the quality of data service is not uniform, and there is a lack of an efficient method of data service discovery; 3) with the increasing demand for visual analysis of data, the cost of visual development is increasing, and there is a lack of a data visualization method which is friendly to user experience and improves the efficiency of visual development. In order to solve the above problems, several key aspects such as data service integration, description model and so on need to be studied. Therefore, the key technologies related to DaaS lifecycle in cloud computing environment are studied in this paper. The main innovations are as follows: (1) A flexible description model of data services based on DaaS is proposed. On the basis of analyzing the insufficiency of the existing data service description language, this paper integrates the data service, then gives the flexible description model of atomic data service and the flexible description model of composite data service. According to the given flexible description model, the data service can be automatically invoked. (2) A data service clustering method for global social service network is proposed. By combining clustering with global social service network, the concepts of social attribute, historical social domain, future social domain and so on are given, and the services are clustered by using the social attributes of services. The efficiency of service discovery in the global environment is improved. (3) A data service analysis method based on visual template is proposed. Based on the reuse idea in the field of software engineering, the definition of visual template and other related definitions are given. The visual template is used to efficiently make visual charts, and to use wizard thread, materialized cache, etc. Replica iteration and other techniques to achieve user-friendly visual chart access experience. The research content of this thesis is regarded as "Research on the key Technology of Network big data platform Software in the Environment of Service Computing" (No. 2014C33071) and Zhejiang Province's Major Science and Technology Project as the Public Welfare Technology Application Research Project of Zhejiang Science and Technology Department. Part of the planned project "Research and application of key technologies for cloud design service platform for special equipment for industrial alliances" (number: 2014C11SA1B0006), From the point of view of constructing the prototype of cloud manufacturing service platform, this paper discusses the life cycle of data service and its key technology in cloud computing environment. Combined with the proposed flexible description model of data service, the data services are integrated, managed, combined and called. Data visualization.
【学位授予单位】：浙江工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.09

【参考文献】