基于语义特征的多视图情感分类方法

情感分析也称为意见挖掘，是对文本中所包含的情感倾向进行分析的技术。目前很多情感分析工作都是基于纯文本的。而在微博上，除了文本，大量的图片信息也蕴含了丰富的情感信息。本文提出了一种基于文本和图像的多模态分类算法，通过使用潜在语义分析，将文本特征和图像特征分别映射到同维度下的语义空间，得到各自的语义特征，并用SVM-2K进行分类。利用新浪微博热门微博栏目下爬取的文字和配图的微博数据进行了实验。实验结果表明，通过融合文本和图像的语义特征，情感分类的效果好于单独使用文本特征或者图像特征。

Multi-view sentiment classification of microblogs based on semantic features

The objective in sentiment analysis is to analyze the sentiment tendency contained in subjective text. Most sentiment analysis methods deal with text only and ignore the information provided in the corresponding pictures. In this paper, we propose a multi-view microblog analysis method based on semantic features. Using latent semantic analysis, we map both the text and image features to the semantic space in the same dimensionality, and use SVM-2K to obtain and classify the respective semantic features. We conducted experiments by crawling text and pictures from popular microblogs. The results show that, by combining the semantic features of text and pictures, the sentiment classification result is better than that obtained using text or image features alone.