基于HowNet自动文摘的研究

自动文摘是指利用计算机自动对文本编制摘要,是自然语言理解的重要应用领域之一。限于相关领域的已有水平,现阶段的自动文摘系统多数是以词语频率作为依据,以词频高的词语作为文章的关键词语,得到的文摘往往与原文中心思想相差甚远,因此,需要对文章的语法、语义和语境进行分析。本文利用HowNet得到词语概念的方法,建立基于概念的自动文摘系统。

HowNet-based Automatic Abstracting Research

Refers to the use of automaticing are automatic computer preparation of a summary of the text is natural language understanding, one of the important applications. Has been limited to the level of related fields, at this stage, the majority of automatic abstracting system based on word frequency as a basis for high-frequency words, as the article's , the text of the abstracts are often a far cry from the central idea, therefore, necessary for the article syntax, semantic and contextual analysis. In this pap...