Multi-level classification model training method and device and multi-level classification method and device

The embodiment of the invention discloses a multi-level classification model training method, a multi-level classification method and a multi-level classification device, which convert a first-level label and a second-level label into feature vectors and utilize the semanteme of the labels so as to improve the accuracy of model training and classification. 本发明实施例公开了一种多层次分类模型训练方法、多层次分类方法和装置，将一级标签和二级标签转换成特征向量，利用了标签的语义性，从而提高了模型训练和分类的准确性。