Short Text Classification based on Feature Extension usingInformation in Images

Zhao Shengjie,Jiang Qianyun

Table 1 The comparation of three models: classifier only using text or images and the classifier using both (not add the similarity module)

Method ResNet (Images Only) SVM (Text Only) LDA (Text Only) CNN (Text Only) Images +Text Precision Rate 0.8496 0.8412 0.8210 0.8651 0.9550