基于神经网络的文本聚类促进本体的构建

Improvement of Creating and Maintaining of Ontology based on Neural Network Document Clustering

通过使用SOM(Self-Orgahizing Map)进行聚类分析,得到自组织语义图,同时针对其中某些情况使用传统聚类算法k-medoids算法,在输出模块中确定每一类中的内容,以获得更为精确的结果并结合到本体内,用于解决由支持本体的构建和维护而引发的自动化问题。

Using SOM （Self-Organizing Map） implements clustering analysis to obtain self-organizing semantic map, and the same time, by using traditional clustering algorithm, K-medoids algorithm for some condition, the contents in each kind will be decided in the output model to obtain more accurate resuh and integrate the result into ontology to support creating and maintaining ontology.