From Business Process Models to Service-Oriented Architecture Models

组织开发软件应用程序以支持其业务流程(Bp)。高质量软件的开发是一个非常复杂的过程，需要大量的软件开发本身和应用程序领域的知识和技能。尽管软件工程在不断发展，但软件开发过程仍然很复杂。实践表明，由于额外的成本、延长的交付时间和糟糕的质量，软件行业继续面临许多问题，后者使美国公司每年损失约1000亿美元的生产力。*就软件应用程序的开发来支持组织的BPS而言，软件模型应该在某种程度上反映它们所支持的BPS。由于BP建模和模型驱动工程(MDE)的进步，我们相信我们可以利用BP和支持它们的软件应用程序之间的对应关系来获得软件模型。这是当前研究计划的前提，该计划旨在利用MDE方法弥合BPS和软件应用程序之间的差距。拟议研究计划的长期目标是为组织提供一种MDE方法，该方法表征了从BP模型到以面向服务的体系结构(SOA)风格表示的服务的软件模型的转换。更具体地说，我们感兴趣的是两个转换：*转换1.从通用流程模型到精确的特定于组织的流程模型的转换。*我们从应该支持的流程的模型指定SOA模型的方法依赖于详细的特定于组织的BP模型的可用性。然而，以足够的细节对特定于组织的BP进行建模需要分析员和流程设计人员的大量工作。我的博士和博士后研究集中在阐述从通用BP模型中自动获取特定于组织的BP模型的方法、技术和工具，结果表明前景光明。在这项研究计划中，我将提出解决该方法提出的基本问题。*转换2.从精确的业务流程模型到SOA模型的转换。*这种转换旨在阐述一种新的方法、技术和工具，以表征从BP模型到详细的SOA软件模型的转换。*建议的研究计划将在软件工程领域产生新的科学贡献。从实践的角度来看，它将详细阐述一种新的方法，该方法由工具和框架支持，帮助组织从通用的BP生成基于SOA的软件。通过自动化这一过程，我们将降低软件开发的复杂性，提高软件质量，这是软件业实际关注的问题。最后，本研究将有助于新HQP(博士生和硕士研究生)的培养。

Organizations develop software applications to support their business processes (BPs). The development of high quality software is a very complex process that requires vast amounts of knowledge and skills both in software development itself and in the application domain. Despite the continuous evolution in the software engineering, software development process remains complex. Practice has shown that software industry continues to face many problems due to extra cost, extended delivery time and bad quality, the latter is costing US companies about $100 billion per year in lost productivity.***To the extent that software applications are developed to support organizations' BPs, software models should somehow reflect the BPs that they support. Thanks to advances in BP modelling and model-driven engineering (MDE), we believe that we can use the correspondence between BPs and the software applications that support them to obtain software models. This is the premise behind the present research program that aims to bridge the gap between BPs and software applications using an MDE approach. The long-term objective of the proposed research program is to provide organizations with an MDE method that characterizes the transformations from BP models to software models expressed as services within the service-oriented architecture (SOA) style. More specifically, we are interested in exploring two transformations:******Transformation 1. Transition from generic process models to precise organization-specific process models.***Our approach to specify SOA models from models of the processes that they are supposed to support relies on the availability of detailed organization-specific BP models. However, modelling an organization-specific BP with enough detail requires an inordinate amount of effort of analysts and process designers. My doctoral and post-doctoral research focused on elaborating methods, techniques and tools to automatically obtain organization-specific BP models from generic BP models, and the results showed promise. In this research program, I will propose to address the fundamental issues raised by the approach.******Transformation 2. Transition from precise business process models to SOA models.***This transition aims at elaborating a new method, techniques and tools that characterizes the transformations from BP models to detailed SOA software models.******The proposed research program will produce new scientific contributions in the software engineering field. From a practical perspective, it will elaborate a novel method supported by tools and a framework that help organizations to generate SOA-based software from generic BPs. By automating this process, we will reduce the software development complexity and enhance software quality which is an actual concern of the software industry. Finally, this research will contribute to the training of new HQP (PhD and MSc students).