基于SELC模型的新闻文本分类方法

A method of news text classification based on SELC model

中文长新闻文本的分类问题主要采用截断式提取方法,但是这种方法会大幅降低特征提取的准确性。针对以上问题,文中提出了在文本处理阶段对长文本新闻进行摘要抽取处理,这样既能保留源文本的特征信息,也能降低时间复杂度。使用ERNIE进行词向量表示,经过自注意力和卷积网络后提高分类任务的准确性。

In the classification of Chinese long news texts,the truncation extraction method is often used,but this method will greatly reduce the accuracy of feature extraction.In view of the above problems,this paper proposes to extract the abstract of long text news in the text processing stage,which can not only retain the feature information of the source text,but also reduce the time complexity.Use ERNIE for word vector representation,improve the accuracy of classification tasks after self-attention and convolutional networks.