中文摘要：

微博客是近年来自然语言处理领域研究的热点。主要针对中文微博客中的情感分类展开研究。结合网络新词和基础情感词，同时考虑了情感词的极性情感强弱，构建四个词典，分别是基础情感词典、表情符号词典、否定词词典和双重否定词词典；在情感词典的基础上’，融合汉语语言学特征和微博情感表达特征，提出一种新的基于极性词典的情感分类方法。实验准确率达到82．2％。实验结果表明，提出的方法可以对中文微博进行较好的情感分类，有一定的应用价值。

英文摘要：

Microblogging is the focus in research field of natural language processing recently. Our study in this paper is mainly in regard to the sentiment classification of Chinese microblog. In combination with new Internet words and basic emotional words and taking into account the strength of the polarity of emotions, we construct four lexicons, they are ： the basic sentiments lexicon, emotional signs lexicon, negative words lexicon and double negative words lexicon respectively. On the basis of sentiments lexicon and fused in Chinese linguistic features and the sentiment expression features in microblogging, we propose a new sentiment classification method based on polarity lexicons. The precision in the experiments reaches 82. 2%. Experimental result indicates that the method proposed in the paper can conduct the sentiment classification on Chinese microblog well, and has certain applied value.