中文摘要：

需求文本的分析和建模是需求工程中一个重要环节,其获取建模过程的自动化也渐渐成为了需求工程中一项重要研究内容.本文针对中文自然语言处理和需求分析中的难点,提出了基于面向类图语义框架的中文需求类图半自动建模方法.该建模方法的流程包括：文本分词与词性标注,基于语义框架的类图模型提取,基于问卷的模型改进和手动模型编辑.该方法具有较高精确度,能够显著地提高中文需求建模的效率.本文设计实现了半自动建模系统来完整地支持整个建模方法流程.最后,本文通过对3个实际需求文本的样例进行实验,检验了基于语义框架抽取类图元素的效果.

英文摘要：

Requirement analysis and modeling is an important part of requirements engineering,and how to automate this course has gradually become an important research issue.Facing the difficulties in Chinese natural language processing and requirements analysis,this paper proposes a semi-automated method based on the semantic analysis.The modeling method includes the following steps：word segmentation,class diagram elicitation based on the semantic framework,model improvement with intelligent questionnaire and manual revision.The method can achieve higher accuracy,and it can obviously improve the efficiency of Chinese requirement modeling.This paper designs and implements a modeling tool to provide full support to the proposed process.Finally,we use three practical requirements specifications as show cases to evaluate the result of model elicitation with semantic framework.