E-commerce evaluation text classification algorithm

Abstract

Based on the research results of the false evaluation and identification of e-commerce shopping, this paper analyzes the characteristics of the evaluation content of the merchant’s brush list. In order to solve the problem that the convolutional neural network (CNN) model is difficult to capture the feature information of the whole evaluation text in the false evaluation identification task, a convolutional belief network (CBN) model based on keyword weighting is proposed. Firstly, the TF-IDF algorithm is used to construct the keyword set, then the word vector is weighted by the keyword vector. Then use restricted boltzmann machine instead of the pooling layer of CNN model to reduce the dimension; finally, the weighted word vector is classified by this algorithm model to complete the recognition task of false evaluation text.