面向协同任务优化的序列模式分析方法研究

发布时间：2018-02-11 09:40

本文关键词： 协同任务资源服务序列工作流跨组织时序依赖关系　出处：《华侨大学》2017年硕士论文　论文类型：学位论文

【摘要】：在互联网、物联网、云计算、大数据及其他新兴信息技术的支持下,多组织协同完成一项任务的模式-协同任务模式在越来越多的领域得到应用,提高协同任务中业务过程的整体效率至关重要。相应地,资源服务也应该以“服务流”的方式服务于业务过程,促使二者之间更佳地匹配。然而,协同任务通常是一种跨组织的业务过程,资源服务的选取和使用对于每个组织而言相对独立,因此,这种自治性和全局最优化之间的矛盾制约了资源服务流的选取。为此,资源服务流的挖掘已成为协同任务系统中一个重要的研究主题。在协同工作流下,分析资源服务之间的时序关系,挖掘资源服务的序列模式,着重考虑资源服务序列的跨组织特性,并提出相应的分析和挖掘方法。分别展开如下研究:(1)资源服务时序关系模型。为了准确地描述资源服务序列,基于工作流模型,把资源服务的时序关系从工作流模型中分离出来,提出一种资源服务时序关系模型。(2)资源服务序列模式挖掘方法。为了分析并获得资源服务序列模式,通过分析资源服务之间的时序依赖关系,求解时序关系强度;同时,针对大规模业务数据,提出一种基于迭代的挖掘结果合并方法,其中,借助了移动平均法合并各分析结果为最终的资源服务时序关系模型,最后通过实验验证方法的有效性并对实验结果进行分析。(3)基于相似度的跨组织资源服务序列模式获取方法。考虑资源服务序列的组织特性,同时也为提高资源服务序列模式的可重用性,基于资源服务序列模式的分析和获取,对已有的资源服务序列进行扩展,首先定义同构序列和原始序列,从资源服务时序关系模型中得到原始序列,再对组织信息的相似度进行量化定义,把相似度高的同构序列作为扩展序列,并通过实验验证获取方法的有效性。最后,详细介绍了一个典型的协同任务系统-中小企业云制造服务平台,并通过该平台验证理论研究结果的有效性。
[Abstract]:With the support of the Internet, Internet of things, cloud computing, big data and other emerging information technologies, the multi-organization collaborative task model has been applied in more and more fields. It is essential to improve the overall efficiency of business processes in collaborative tasks. Accordingly, resource services should also serve business processes in a "service flow" manner, enabling a better match between the two. Collaborative tasks are usually a cross-organizational business process, and the selection and use of resource services are relatively independent for each organization. Therefore, this contradiction between autonomy and global optimization restricts the selection of resource service flows. Resource service flow mining has become an important research topic in collaborative task system. Under collaborative work flow, the temporal relationship between resources and services is analyzed, and the sequential patterns of resource service are mined. The cross-organizational characteristics of resource service sequence are considered, and the corresponding analysis and mining methods are proposed. The following researches are carried out respectively on the model of resource service temporal relationship. In order to accurately describe the resource service sequence, the workflow model is used to describe the resource service sequence. In order to analyze and obtain the sequential pattern of resource service, a method of mining resource service sequence pattern is proposed by separating the temporal relation of resource service from workflow model. By analyzing the temporal dependence between resources and services, the intensity of temporal relationship is solved. At the same time, for large-scale business data, an iterative mining result merging method is proposed, in which, With the help of the moving average method, the results of the analysis are combined to form the final model of the time series of resources and services. Finally, the validity of the method is verified by experiments and the experimental results are analyzed. At the same time, in order to improve the reusability of the resource service sequence pattern, based on the analysis and acquisition of the resource service sequence pattern, the existing resource service sequence is extended. First, the isomorphic sequence and the original sequence are defined. The original sequence is obtained from the model of resource service time series, and then the similarity of organizational information is quantified. The isomorphic sequence with high similarity is regarded as the extended sequence, and the validity of the method is verified by experiments. A typical collaborative task system, cloud manufacturing service platform for small and medium-sized enterprises, is introduced in detail, and the validity of the theoretical research results is verified by the platform.
【学位授予单位】：华侨大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.09

【相似文献】