微博文本的情感分类，是自然语言处理在短文本分类上的一个应用，其具有的上下文信息少，文本长度短，口语化及网络化语言成分高等特点使传统分类方法难以实现，本文应用卷积神经网络实现微博的情感分类，首先对卷积神经网络的基本概念及原理进行了阐述，对核心模块的实现过程进行详细的说明和公式推导，在6000条微博文本上进行实验，并基于tensorflow构建框架，最后使用图形化的技术实现一个机器学习的简易系统，实验结果表明：模型在在测试集上运行的准确率为75%。

关键词  机器学习 文本分类 卷积神经网络 自然语言处理

Title       Text Classification

Based on Convolution Neural Network

Abstract

The emotional classification of weibo text is an application of natural language processing in short text classification. The text from weibo has the characteristics of less context information, short text length, colloquial and networked language composition, which makes the traditional classification method difficult to realize, This paper applies the convolution neural network on the essential classification of weibo text; first，It introduces the basic concepts and principles of the convolution  neural network, after that it offers the detailed implementation of the core module and the formula derivation. The framework is built based on tensorflow, with 6000 weibo text on the experiment. Meanwhile, I apply the graphical technology to build a simple machine learning system, the experimental results show that the accuracy rate of the model can reach 75% on the test set.

Keywords  Convolution Neural Network, Machine learning, Text Classification, Natural Language Processing