自动文摘系统评价方法的研究与实践  

Study and Practice of Evaluation Method for Automatic Summarizing System

作　　者：沈洲[1] 王永成[1] 许一震[1] 方澈[1] 

机构地区：[1]上海交通大学计算机系,上海200030

出　　处：《情报学报》2001年第1期66-72,共7页Journal of the China Society for Scientific and Technical Information

基　　金：到国家 8 63项目资助 !项目编号 :863 30 6 0 4 0 2 0 2

摘　　要：自动文摘系统作为一种信息压缩精选工具 ,越来越引起人们广泛的兴趣。但是 ,目前仍未有一种客观公认的评价方法来对已有的系统进行评价 ,从而极大地阻碍了自动文摘领域的研究。本文首先阐述了自动文摘系统评价所面临的问题 ,随后综述了当前国内外自动文摘的评价方法 。In such an era of information explosion, more and more people are interested in the field of automatic summarization which can serve as a kind of information refining tools. However, there isn't an objective evaluation method that could be accepted by the majority of researchers, so, the lack of evaluation method becomes a big obstacle for study of automatic summarizing. In this paper, the problems facing with the evaluation of automatic summarizing systems were introduced. Then, many evaluation methods used in the field were reviewed. At last, a kind of evaluation method referred to the Turing Testing was introduced and practiced.

关 键 词：自然语言处理 自动文摘系统 评价方法 内部评价方法 外部评价方法 

分 类 号：G354[文化科学—情报学]