A Chinese text sentiment analysis method based on deep learning

Abstract

The invention discloses a Chinese text sentiment analysis method based on deep learning, and belongs to the technical field of natural language processing. The defects of an unsupervised sentiment analysis method based on English are overcome. The method comprises the following steps: after converting an obtained corpus text into pinyin, pre-training a constructed language model to obtain a pre-trained language model; obtaining a small amount of text data which is in the same field as the corpus text and has emotion categories, converting the text in the text data into pinyin, and training a constructed emotion classification model based on a pre-trained language model to obtain a trained emotion analysis model; and carrying out sentiment classification on the unlabeled text by utilizing the trained sentiment analysis model to obtain a corresponding sentiment category label. The method is used for Chinese text sentiment analysis.

Description

technical field [0001] A Chinese text sentiment analysis method based on deep learning is used for sentiment analysis of Chinese text, and belongs to the technical field of natural language processing. Background technique [0002] Text sentiment analysis refers to judging the emotional tendency of text. [0003] The language model is used to calculate the probability of a sentence and judge whether a sentence is reasonable. [0004] RNN is a recurrent neural network, a type of neural network for processing sequence data. [0005] LSTM is a long short-term memory network, a special type of RNN, which can learn long-term dependent information. [0006] GRU is a gated recurrent unit, a variant of LSTM, which simplifies the structure of the LSTM model. [0007] In the existing Chinese text sentiment analysis, most of the texts are classified using Chinese characters, and the accuracy of sentiment classification prediction is low through Chinese characters; while most of the ...

AI Technical Summary

Problems solved by technology

[0011] For the problems of above-mentioned research, the purpose of the present invention is to provide a kind of Chinese text emotion analysis method based on deep learning, solve the weak point of the unsupervised emotion analysis method based on English in the prior art: (1) this method cannot be used in Chinese (2) The language model training in this method is time-consuming, the model network parameters are more, and the pre-training language model takes a lot of time; (3) This method only uses the last character of each sentence text as the sentence text The feature representation of the last character, and the word vector of the last character cannot represent the entire sentence text, which will reduce the accuracy of emotion classification for each sentence text