Chinese Text Sentiment Classification Based on LSTM Word-Tag Relational Network Model

Abstract

In the task of Chinese text sentiment classification, the method based on sentiment dictionary and machine learning is suitable for the situation of small amount of text corpus data and simple text semantics. With the explosive growth of online text information and the continuous enrichment of expression methods, many researchers have gradually applied deep learning methods to text sentiment classification and made breakthrough progress. Convolutional neural network (CNN) can effectively capture the local feature information of spatial structure, but it lacks the ability to learn the contextual relevance of words. Recurrent Neural Network (RNN) can solve the semantic context problem well, but the phenomenon of gradient explosion and gradient disappearance is easy to occur in the training process. To solve these problems, this paper designs a word-tag relation network model based on LSTM (Long Short-Term Memory) Chinese text sentiment classification model, which can acquire deeper tag dependency and enrich the feature information of the text.