摘要： 在自然语言处理，尤其是在基于语法和语义规则的信息检索、机器翻译系统中，对于句群的处理显得格外重要。它是计算机从理解孤立的词义和句义上升到理解篇章整体中心内容的一个重要的跃变步骤。作为句群理解的关键一步，句群的识别显得尤为重要。该文从句群本身的构成特点出发，对句群进行了内部语义组合方式的划分，这一划分适宜计算机进行处理。根据已经取得的“HNC语言概念空间表示”的研究成果，制定了识别具有以上构成特点句群的相关规则。实验表明，划分方法具有很高的句群覆盖率，同时切分规则具有很高的准确度。

关键词: 句群, 切分策略, 计算语言学

Abstract: In nature language processing, the capacity of processing the sentence group becomes more and more important. It is the key to obtain the meaning of a paragraph, a chapter, and even the full text. This paper proposes a method for marking off the Chinese sentence groups based on the semantic relationship between the sentences within the sentence group. The paper also presents some formalized rules for detecting the Chinese sentence group. They are based on the division method and take advantage of the symbolic system of language concept space which is used to express the meaning of a word or a sentence aforetime. The experiment indicates that the method and the rules are good at detecting the Chinese sentence group.

Key words: Chinese sentences group, Strategies for detecting, Computational linguistics