基于文献结构的自动文摘的初探

首先阐述了人们进行自动文摘研究的必要性，并介绍了自动文摘的技术背景。然后着重论述了基于文献结构的自动文摘的特点和实现方法，并对已有的数学模型进行评价，提出了作者的一些改进想法。指出了传统自动文摘方法的缺陷，并根据基于文献结构的自动文摘的特点提出了解决问题的办法。本文在最后还探讨了自动文摘在“知识发现”和文本信息挖掘领域内的初步应用。

The Explore of the Automatic Abstracting Based on Text Structure

In this paper, the author first discusses the necessity of automatic abstracting, and then introduces its technology background. This paper mainly discusses the characters of the text structure -based automatic abstracting and the methods to realize it. The author evaluates several mathematic models and gives some opinions on them,and also points out the flaws of the traditional methods of automatic abstracting and gives some advice on solving the problems. In the end, the author discusses the application of automatic abstracting in the KDD (Knowledge Discovery from the Data) field.