A Study on Process Diagnostic-Check Procedures and Reliability Assessment Technologies for Software Process Quality Improvement

在本研究中，基于最终产品质量严重依赖于开发过程质量的认识，我们讨论了提高软件过程质量的过程诊断检查程序和相关的软件质量/可靠性评估技术，并考虑了开发技术和工具、人员、项目、平台等过程要素的特点。我们假设可以通过改善这些特性来提高工艺质量，例如，开发技术和工具的使用频率、开发人员的技能和经验、项目的能力、平台的稳定性等。特别是考虑到开发过程中人为因素对过程质量的影响，我们得到了以下研究结果：（1）基于质量工程方法进行了设计评审实验，即：田口方法，我们已经量化的中间产品的质量/可靠性和最终的之间的关系。（2）基于过程能力成熟度模型（CMM），通过对成功项目的风险管理，建立了过程诊断检查程序框架，以持续改进过程质量。(3)We提出了量化开发过程中上游产品质量与下游产品质量之间关系的软件可靠性评估方法，以及与分布式开发环境和开源系统开发等现代软件开发范式相对应的软件可靠性评估方法。

In this study, based on the understanding that final product quality is heavily dependent on development process quality, we have discussed process diagnostic-check procedures for improving software process quality and related software quality/reliability assessment technologies by considering the characteristics of process elements such as development techniques and tools, personnel, projects, platforms and so on. Here, we assume that the process quality can be increased by improving these characteristics, e.g., the use frequency of development techniques and tools, the skill and experience of development personnel, the projects' capability, the stability of platforms etc. In particular, considering that the process quality is affected by human factors in the development process, we have obtained the following research results :(1)Conducting the design-review experiments based on a quality engineering approach, i.e., Taguchi method, we have quantified the relationship between the intermediate product's quality/reliability and the final one.(2)Based on the process capability maturity model generally called CMM, we have developed a framework of the process diagnostic-check procedures by using risk management for successful projects in order to improve the process quality sustainably.(3)We have proposed software reliability assessment methods quantifying the relationship between the upper-stream product's quality and lower-stream one in the development process, and those corresponding to modern software development paradigms such as distributed development environment and open source system development.