The Explore of the Automatic Abstracting Based on Text Structure
基于文献结构的自动文摘的初探

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.