一种基于本体的文本主题提取方法研究

AN ONTOLOGY-BASED TEXT TOPIC EXTRACTION METHOD RESEARCH

在深入分析当前流行的文本主题提取技术和方法的基础上,提出一种将本体技术应用于文本主题提取的方法。使用本体技术用语义向量表示文本句,对文本进行预处理,然后进行语义相似度计算和语义聚类,最后从每类中抽取代表句生成文本主题。实验结果表明,该方法在提取文本主题方面是一个有效的方法。

After thorough exploration with present text topic extraction technology and methods,the paper proposes an approach to apply ontology technology to text topic extraction.The approach counts on ontology technology to utilize semantic vectors to represent text sentences.Then it carries out semantic similarity calculation and semantic clustering.Finally it extracts representative sentences from every class to generate text topics.Experimental results show that the approach is feasible in text topic extraction.