As software systems are used widely and deeply in society, trustworthiness, which refers to the ability to operate correctly under any situation, becomes one of the hot topics in practice and research. To develop trustworthy software system needs the elicitation of trustworthiness requirements. But the elicitation of the trustworthiness requirements needs the requirements engineer to possess lots of knowledge about undesired conditions a software system may face and the corresponding countermeasures the software system could take to deal with them. There still lacks a systematic method to guide the elicitation of trustworthiness requirements. This paper introduces a meta-model of trustworthiness, founds a knowledge base according to this meta-mod-el, presents a trustworthiness requirements pattern and a method about how to generate patterns from the knowledge base to help eliciting trustworthiness requirements. Finally a stock trading system is used to show the feasibility of this method.%软件系统的可信性已经成为一个受到广泛关注的焦点问题.开发可信的软件系统的前提是在需求阶段提取恰当的可信性需求.能否提取出足够好的软件可信性需求,不仅依赖于需求工程师对未来软件系统可能面临的威胁的认识,还依赖于其对各种威胁有效的应对措施的了解和掌握.目前缺少系统化的方法指导软件可信性需求的提取.文中提出一个软件可信性需求上层本体作为软件可信性需求的概念框架.在此框架的基础上,开发了一个软件可信性需求知识库,定义了软件可信性需求模式框架以及如何根据知识库的内容进行模式实例化的过程,帮助提取可信需求.最后利用一个股票交易系统作为案例展示了该方法的可行性.
展开▼
