云计算环境下分布式语义文本自适应分类方法研究

传统分布式语义文本分类方法难以高效地在云计算环境下实现文本快速准确分类，为此，提出一种新的云计算环境下分布式语义文本自适应分类方法。通过期望交叉熵对分布式语义文本特征进行选择，针对任意类别中的词，按照权重值从大到小的顺序对其进行排列，将排在前面的若干词看作特征词，针对分布式语义文本集中的所有类别进行同样的操作，将获取的所有类别特征词结合在一起，建立特征词典。针对主题引入加权策略，通过权重值对不同主题针对不同类别的判断能力进行描述，以获取最佳主题，给出新文本特征产生过程。依据提取的分布式语义文本特征，通过朴素贝叶斯分类器实现分布式语义文本的自适应分类。实验结果表明，所提方法分类精度和效率高。

Research on distributed semantic text adaptive classification method in cloud computing environment

The traditional distributed semantic text classification method is efficient in cloud computing to achieve fast and accurate text classification, this environment, put forward a new classification method of cloud computing distributed semantic text adaptive environment. The expected cross entropy on the distributed semantic text feature selection, for any category of words, according to the weight value from big to small order of the arrangement, the row in front of the number of words as feature words, are the same for all categories of distributed semantic text of all categories of feature words will get together, the establishment of the feature dictionary. The weighted strategy is introduced into the topic, and the ability to judge the different categories according to the different values of the weight is used to obtain the best topic, and the process of generating new text features is given. According to the extracted distributed semantic text features, the adaptive classification of distributed semantic text is realized by the simple Bias classifier. Experimental results show that the proposed method has high classification accuracy and efficiency.