Emotion analysis method based on improved CNN-LDA

Abstract

The invention discloses an emotion analysis method based on improved CNN-LDA. The method comprises the following steps of preprocessing the crawler douban film review data; obtaining a corpus set wordvector by adopting Word2vec; adding an emotional tendency weight by using an emotional dictionary; obtaining an emotional tendency word vector by using a word vector obtained by Word2vec and an emotional tendency weight vector machine; adopting an LDA topic generation model to obtain a corpus set topic vector; superposing the emotional tendency word vector and the topic vector to obtain an inputrepresentation of the corpus set; inputting a training set word vector representation into a CNN model to establish an emotion analysis model; and finally, inputting a test set word vector representation into a trained network for classification and evaluation. According to the emotion analysis method based on the improved CNN-LDA provided by the invention, by combining the emotion tendency word vector with the theme vector, and considering the word granularity and text granularity level information containing the text emotion information at the same time, the text information can be better utilized in the training process, and the accuracy of the text emotion analysis is improved.

Description

technical field [0001] The invention relates to a sentiment analysis method based on improved CNN-LDA, which belongs to the field of sentiment analysis of wood text. Background technique [0002] Sentiment analysis is a common application of natural language processing methods, especially in classification methods aimed at extracting the emotional content of text. Text sentiment analysis is the process of analyzing, processing, inducing, and inferring subjective texts with emotional color, such as analyzing users' attributes such as "picture, sound effects, plot, and cast" of a certain movie from comment texts. emotional tendencies. At present, sentiment analysis technology is widely used in Internet applications such as online translation, user evaluation analysis, and opinion mining. For merchants, quickly obtaining the emotional tendencies of user reviews can facilitate advertising and hot topic pushes; for users, accurately obtaining the emotional tendencies of other u...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is to overcome the defects of a single convolutional neural network and provide a combination of the LDA topic generation model and the convolutional neural network, which can mine the deep semantic information of the text and improve the accuracy of text sentiment analysis. A Sentiment Analysis Method Based on Improved CNN-LDA

Images