File classification labeling method and system based on sequence coding

The invention belongs to the technical field of text classification, and provides a file classification labeling method and system based on sequence coding, and the method comprises the following steps: obtaining the position features of a to-be-classified file; according to the obtained position features, word embedding is carried out on syntactic and semantic information of word levels in the to-be-classified files, sequence coding is carried out on relation and structure information of sentence levels in the to-be-classified files, and conversion of the to-be-classified files from a document space to a vector space is completed; extracting vector space features of the files to be classified, and performing sequence coding on the extracted vector space features; and classifying the files based on the sequence codes of the vector space features and a preset file classification model. 本公开属于文本分类技术领域，提供了一种基于序列编码的文件分类标注方法与系统，包括以下步骤：获取待分类文件的位置特征；根据所获取的位置特征，对待分类文件中单词级的句法和语义信息进行词嵌入，对待分类文件中的句子级的关系和结构信息进行序列编码，完成待分类文件