计算机自动文摘方法的分类探讨  

Exploration & Discussion on the Classification of Automatic Summarization Method of Computer

作　　者：孙秋月[1] 焦伟婷[2] 蔡宁[1] SUN Qiuyue;JIAO Weiting;CAI Ning(Hebei Normal University of Science and Technology,Qinhuangdao 066004)

机构地区：[1]河北科技师范学院外国语学院,秦皇岛066004 [2]河北科技师范学院教育学院,秦皇岛066004

出　　处：《办公自动化》2023年第3期56-58,38,共4页Office Informatization

摘　　要：信息化时代,自动文摘是语言信息处理领域的重要研究课题。根据计算机判别语篇主题所采取的理论基础和方法的不同,本文将自动文摘技术分为三类:基于语篇表层特征的统计自动文摘法、基于语篇理解的自动文摘法、语篇表层特征和语篇理解相结合的自动文摘法,探讨了每种自动文摘方法的实施路径、特点和不足,提出基于语篇深层次理解的自动文摘的必要性。In the age of informatization, the automatic summarization is an important research topic in language information processing. According to the theoretical basis and methods, the paper classifies the automatic summarization into three categories: the statistical automatic summarization based on the textual superficial features, the automatic summarization based on the textual understanding and the automatic summarization based on the combination with textual superficial features and textual understanding. The paper illustrates the method of each category, the characteristics and the weaknesses. Finally, the paper proposes the possibility of automatic summarization based on deeper understanding of textual meaning.

关 键 词：自动文摘 统计法 语篇表层特征 语篇理解 

分 类 号：TP301.2[自动化与计算机技术—计算机系统结构] H085.2[自动化与计算机技术—计算机科学与技术] C81[语言文字—语言学]