基于语义图结构的中文文本分类研究

发布时间：2018-04-23 21:43

本文选题：图结构 + 文本表示　；参考：《西安电子科技大学》2012年硕士论文

【摘要】：随着互联网技术的迅速发展和普及，人们日常需要处理的信息日益丰富。面对海量的信息资源，人们很难迅速有效地找到真正所需的信息资源。文本分类技术作为信息过滤、信息检索、搜索引擎、数字化图书馆等领域的技术基础，有着广泛的应用。为了解决传统基于统计的文本表示方法中词语间语义信息缺失的问题，本文结合图论中的理论，提出了一种新的基于语义图结构的文本表示模型和中文文本分类方法。论文首先对语义图结构和基于语义图结构的文本之间的相似度计算方法进行了定义，通过将文本表示成语义图结构，有效地解决了文本表示过程中的语义信息流失问题；其次，给出了基于语义图结构的文本分类算法；最后，介绍了一种基于语义图结构的文本分类系统RCSGC。实验表明，与传统的基于统计向量的SVM（支持向量机）等方法相比，，RCSGC方法更有利于文本的语义信息的表示。实验表明：该方法比传统的SVM方法有更好的性能。
[Abstract]:With the rapid development and popularization of Internet technology, people need to deal with more and more information. In the face of massive information resources, it is difficult to find the real information resources quickly and effectively. Text classification technology is widely used as the technical foundation of information filtering, information retrieval, search engine, digital library and so on. In order to solve the problem of the absence of semantic information between words in the traditional statistical text representation, a new text representation model based on semantic graph structure and a Chinese text classification method are proposed in this paper based on the theory of graph theory. Firstly, the similarity calculation method between the semantic graph structure and the text based on semantic graph structure is defined, and the loss of semantic information in the process of text representation is effectively solved by expressing the text into the semantic graph structure. A text classification algorithm based on semantic graph structure is presented. Finally, a text classification system RCSGC based on semantic graph structure is introduced. The experimental results show that compared with the traditional SVM (support vector machine) method, the RCSGC method is more convenient to represent the semantic information of the text. Experiments show that this method has better performance than the traditional SVM method.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP391.1

【参考文献】