运用TextCNN的零售平台商品分类

TextCNN Model for Retail Platform Product Classification

笔者采用基于卷积神经网络的TextCNN模型,利用多尺寸卷积核的卷积神经网络进行数据特征提取,并对其进行优化来提高零售商品分类准确率。通过爬取京东网站零售商品信息进行实验,并对比基于TF-IDF的传统机器学习模型和使用广泛的LSTM模型,证明了TextCNN模型在商品信息分类上的优势。

In this paper,the textcnn model based on convolution neural network is used to extract data features and optimize it to improve the accuracy of retail product classification.Through the experiment of crawling the retail product information of Jingdong website,and comparing the traditional machine learning model based on TF-IDF with the widely used LSTM model,it proves the advantage of textcnn model in the classification of product information.