首页 | 本学科首页   官方微博 | 高级检索  
     

基于篇章理解的自动文摘研究
引用本文:王建波,杜春玲.基于篇章理解的自动文摘研究[J].中文信息学报,1995,9(3):33-42.
作者姓名:王建波  杜春玲
作者单位:1.深圳市EDI服务中心2.哈尔滨工业大学计算机系
摘    要:本文在自然语言理解的基础上, 对自动文摘系统进行研究。研究在两级上进行, 一级是基于中心名词, 中心动词同其修饰成分间关系研究, 实现句子成分过滤, 完成句子主干提取。一级是上下文级, 基于篇章文体结构, 句子语义关联, 分别实现摘要基集产生算法, 摘要基集扩充算法及摘要集合生成算法。摘要基集的产生依赖于中心段, 中心句选取, 摘要基集扩充和摘要集合产生依赖于模糊语义距离的测试。

关 键 词:自然语言理解  自动文摘  信息处理  自然语言语义  模糊理论  

Study of Automatic Abstraction System Based on Natural Language Understanding
Wang Jianbo, Du Chunling, Wang Kaizhu.Study of Automatic Abstraction System Based on Natural Language Understanding[J].Journal of Chinese Information Processing,1995,9(3):33-42.
Authors:Wang Jianbo  Du Chunling  Wang Kaizhu
Abstract:The study of automatic abstraction based on natural language understanding is a practical but a difficult branch in information processing. It consists of study of natural language understanding and automatic abstracting. Text meaning representation (TMR) is designed on the requirement of automatic abstraction system. TMR is an attributed tree structure. Rule based technique and semantics directed ATN technique are employed in sentence meaning analysing. A conjunction matching algorithem has been implemented to analyse complex sentence. Lexical coherences of action words are used to analyse relations between sentences. Study of abstraction works on sentence level and context level. On sentence level,a sentence consititute filtering algorithem has been implemented to extract sentence frame. On context level, primary set of abstract sentence producing algorithem, primary set expanding algorithem and abstract producing algorithem have been.implemented.
Keywords:Natural Language Understanding  Automatic Abstraction  Information Processing  Natural Langguage Semantics  Fuzzy Set Theory  
本文献已被 CNKI 维普 等数据库收录!
点击此处可从《中文信息学报》浏览原始摘要信息
点击此处可从《中文信息学报》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号