基于自编码神经网络的文本表示应用研究

Research indicates neural network based on self-encoding text

本文主要研究高校招生领域Web短文本的表示方法及处理步骤。基于自编码神经网络对输入数据的要求,研究文本预处理的主要技术:中文分词、停用词处理、降维处理,文本向量化表示等技术,并分析了文本表示的处理流程。

This paper studies the field of representation College Enrollment Web and short text processing steps.Based on self-encoding neural network input data requirements,the main technical pretreatment study text：Chinese word,stop word processing,reduce the dimension text representation techniques to quantify and analyze the process flow text representation.