基于云理论的文本分类方法  

作　　者：姜健[1] 

机构地区：[1]辽宁工程技术大学工商管理学院,辽宁葫芦岛125105

出　　处：《黑龙江科技信息》2009年第36期137-137,共1页Heilongjiang Science and Technology Information

摘　　要：提出一种基于云理论和神经网络构造决策树的文本分类方法。运用云神经网络学习变量间的云映射关系,从中生成云决策树。这种方法结合了神经网络的学习算法和决策树的推理方法,具有神经网络的学习能力,并且应用了云发生器对处理不确定性的能力。更符合人类的思维方式,从而进一步提高了文本分类的效率、准确性和可靠性。This paper proposes a text classification method based on Cloud Theory and neural network structure decision tree. It adopts cloud neural network to study the cloud mapping relationship between variables, so as to generate cloud decision tree. Combining the study calculation of neural network with inference technique of decision tree, this method owns both the study capability of neural network and the cloud generator＇s ability to deal with uncertainty. Therefore it adjusts to human＇s mode of thinking and further improves the efficiency, correctness and reliability of text classification.

关 键 词：文本分类 云模型 决策树 神经网络 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TP273.4[自动化与计算机技术—计算机科学与技术]