基于情绪特征的中文微博用户性别识别

Gender recognition of Chinese micro-blog users based on emotion features

随着互联网的蓬勃发展,微博受到了越来越多用户的青睐,对微博用户性别的研究也逐渐成为学术界研究的热点。目前,对英文微博文本用户的性别识别已有研究,但针对中文微博用户性别识别的研究较少。从两性表达情绪的差异出发,提出了一种基于情绪特征的中文微博用户性别识别方法。本文考虑的情绪特征包括情绪词特征和与情绪相关的语言风格特征。实验结果表明,利用情绪特征提高了用户性别识别的精度。

With the vigorous development of the Internet, micro-blog service is attracting more and more users. Gender recognition of micro-blog users thereby becomes a hot research topic. Tremendous ef forts have been made on gender recognition of Twitter users. However, research on Chinese micro-blogs users is still new. Based on the difference in the emotion expressions between males and females, we propose a gender recognition method of Chinese micro-blog users based on emotion features. The emo- tion features including emotional words and linguistic style features associated with the emotions. Ex- perimental results show that using emotion features can improve the accuracy of gender recognition.