基于过程挖掘的SOA场景发现与研究

发布时间：2018-02-25 07:08

本文关键词： 过程挖掘 SOA 业务流程管理场景 ProM 云平台　出处：《上海交通大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着云计算理念的深入人心,企业在设计和建设信息系统时愈来愈趋向于采用面向服务的体系结构(Service-Oriented Architecture,SOA)模式。但对于一个企业,尤其是业务流程和功能需求时刻变化的企业在构建IT系统时仍存在很大的挑战。企业构建信息系统来支撑企业业务处理的重心主要集中在业务流程管理(Business Process Management)和SOA的有机结合基础之上。目前作为信息系统设计关键环节的传统业务建模方法依赖于人工经验进行业务分析和建模,难以全面把握业务实际执行过程中动态变化的场景信息,同时制约了系统应用开发的质量和效率,因此需要一个可以反映实际业务处理需求的建模方法来帮助业务建模人员分析和设计业务流程,并为服务应用开发提供良好支持。随着业务规模的日趋庞大,由于业务需求的不断更新导致实际业务处理逐渐偏离初始设计模型,使得系统应用严重制约业务执行效率。针对该问题,论文提出了基于过程挖掘技术的场景发现方法框架,重点研究如何利用过程挖掘的三种视图层面挖掘技术发现业务场景中任务活动控制流、组织结构和业务对象三要素及其相互作用关系,进而得到包含了角色、任务和业务对象的应用场景信息,从任务节点执行的粒度分析业务处理,并在此基础利用业务模型帮助开发人员进行服务编制管理,从而高效利用SOA服务来满足实际业务处理需求。主要研究内容如下:(1)提出基于过程挖掘的场景模型应用框架描述基于过程挖掘的场景模型的应用环境,并说明场景模型对云服务平台的支持作用和应用模式。设计基于过程挖掘的场景模型应用框架,包括数据持久层、挖掘层、以及应用层。最后阐述应用框架在云平台流程服务生命周期的运行、接收反馈到分析改进三个阶段的数据流动情况。(2)构建基于过程挖掘的场景元模型根据对过程挖掘技术以及场景概念相关内容的研究,建立基于过程挖掘的场景元模型。元模型定义了日志记录子模型和场景挖掘处理子模型,覆盖了过程挖掘三种视图、服务运行时信息和企业业务相关数据。然后基于场景元模型分析业务流程的任务控制结构和组织结构两大要素,为实现基于应用场景的业务流程模型构建、分析和改进提供理论基础,进而指导服务组合优化。(3)建立基于过程挖掘的场景发现方法框架在过程挖掘技术的控制视图、组织视图和实例视图三种层面的挖掘技术基础上设计场景发现方法的实现步骤。根据场景元模型从服务调用日志中挖掘业务场景与应用场景。对BPEL描述机制基于场景元模型进行扩展,使得场景模型可以指导服务编制,并快速转化为可执行服务文件。(4)基于过程挖掘的场景发现方法框架的原型构建与验证基于ProM开源框架构建原型系统。结合某交通物流云服务平台背景对服务日志记录进行挖掘分析和验证,最后,定性分析与比较基于场景的建模方法和典型的业务建模方法。原型系统的实现和验证结果表明,本文提出的基于过程挖掘的SOA场景发现方法框架为监控分析和改进业务流程起到了有效的支持作用,具有一定的理论价值和应用前景。
[Abstract]:With the popular concept of cloud computing, enterprise in the design and construction of information system is more and more tend to be in the service oriented architecture (Service-Oriented, Architecture, SOA). But for an enterprise, especially the business process and functional requirements of the constantly changing enterprises still a big challenge in the construction enterprises to build IT system. The information system to support the focus of enterprise business processing focuses on business process management (Business Process Management) based on organic combination and SOA. At present as the traditional business modeling method of key information system design on business analysis and modeling on the artificial experience, it is difficult to fully grasp the business actual implementation of dynamic scene information. At the same time restricts the application of the system of quality and efficiency, so we need a can reflect the actual business needs The modeling method to help the business modeling analysis and design of business processes, and provide good support for service application development. With the increasingly large scale of business, because business needs updating leads to the actual business processes gradually deviate from the initial design model, the system should perform efficiency restrict business. Aiming at this problem, this paper puts forward the methods of framework the discovery process mining technology based on the scene, the three view focuses on how to use the process of mining technology flow control tasks in the business scenario, organization structure and business object three factors and their interactions, and contains the role, application of scene information tasks and business objects, business processing tasks from the size analysis node implementation, and on this basis the use of business model to help developers of service management, and Gao Xiaoli Use the SOA service to meet the actual needs of business processes. The main contents are as follows: (1) the application framework of scene model mining process description application environment scene model based on process mining, and explain the scene model support for cloud service platform and application mode. Application framework design scene model based on process mining. Including the data persistence layer, mining layer, and application layer. Finally, the operation framework in the cloud platform process service life cycle, receiving feedback to the analysis and improvement of three stages of data flow. (2) the construction process of mining scene element model according to the process of mining technology and related concepts based on the content of the scene. Set up the scene model based on process mining. Element model defines a sub model of logging model and scene mining, covering the mining process of three kinds of visual map service Business information and data bank. Then the scene element model based on the analysis of business process control task structure and organization structure of the two elements, in order to realize the construction of business process model based on scenario analysis and improvement, provide a theoretical basis and guidance service optimization. (3) to establish the control framework of mining technology in the process of view method the discovery of the mining process based on the scene, the organization view and instance view three aspects of mining technology based on the design scene found implementation steps method. According to dig business scenarios and application scenarios of Chang Jingyuan model dug from the service call log. On the BPEL scene description mechanism element model based on the extended model, makes the scene can guide service system and quickly converted to executable file service. (4) the mining process based on the scene found that construction and verification of ProM based on open framework prototype Construction of the prototype system source framework. Combined with a traffic logistics service platform for service log mining background analysis and verification, finally, qualitative analysis and comparative modeling method based on scene and typical business modeling method. The implementation of the prototype and verification results show that the proposed process mining framework of service discovery method based on SOA scene the process to play a supporting role for the effective monitoring and improvement, has a certain theoretical value and application prospect.

【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP311.52

【相似文献】