基于卷积神经网络的英文邮件分类

English Mail Classification Based on Convolution Neural Network

文本信息量呈指数次方急剧增长,比如邮件的数目,信息的容量等,如何高效的处理信息成为关注的焦点,邮件分类可以将垃圾邮件进行过滤,提高工作效率。利用卷积神经网络进行英文邮件中垃圾邮件分类的同时,采用的数据集是Enron邮件数据集,对该数据集进行了数据预处理、卷积神经算法以及训练,最终在英文邮件中垃圾邮件分类的准确率以及分类速度都有明显的提高。

The amount of text information increases exponentially,such as the number of mail and the capacity of information.How to deal with information efficiently is the focus of attention.Mail classification can filter spam and improve work efficiency.This paper uses convolutional neural network to classify English spam.At the same time,the data set is Enron mail data set.The data set is preprocessed,together with convolutional neural algorithm and training.Finally,the accuracy and classification speed of English spam classification are improved obviously.