中文摘要：

由于中文情绪表达的多样性，以及微博情绪的丰富性和敏感性，情绪词在表达情绪时存在强弱差别，相同的情绪词在不同的语料语境中也可能表达不同的情绪强度。因此文中提出了基于情绪强度的中文微博情绪分析，并根据语料上下文计算出情绪词的情绪相似度，基于情绪相似度自动标注了情绪强度，利用情绪强度进行微博文本的情绪分析。实验结果表明，对情绪词进行情绪强度的标注可以更细致地识别出微博中的主要情绪，进一步提高微博情绪分析的准确率。

英文摘要：

Due to the diversity of expressions of emotion in Chinese,and the richness and sensitivity of micro-blogs emotion,emotional words may express emotions at different levels of strength,an emotional word may express different levels of emotional strength in different corpora. The Chinese micro-blogs emotion analysis is proposed based on emotional strength,calculate the emotional similarity of emotional words based on corpus context,and use it to annotate the emotional strength of words. Finally,analyze the emotion of Chinese micro-blogs based on emotional strength. The experimental results show that emotional strength is effective in identifying the major emo-tion of micro-blogs,and improves the accuracy of micro-blog emotional analysis further.