Film review sentiment analysis method based on improved convolutional neural network model

Abstract

The invention claims to protect a film review emotion analysis method based on an improved convolutional neural network model. Introducing a weight distribution layer between the input layer and the convolutional layer can analyze important parts of the text, reduce noise, and improve the processing characteristics. Convolution is used to build a model. The convolution method is to generate local features around words, and then use local maxima to combine to create fixed-size features. The gradient descent method is used in the convolutional layer to calculate, gradient dispersion may occur, and the gating mechanism is introduced to reduce the dispersion; secondly, the softmax layer is canceled in the new model, and the support vector machine layer is added; finally, the conditional random field is used Not only the feature function of the traditional model on the i-th label is processed, but also the information feature function of its front and rear positions. The invention improves on the traditional convolutional neural network and adds a conditional random field layer, so that high-level abstract features can be extracted and better classification ability is achieved.

Description

technical field [0001] The invention belongs to Chinese text sentiment analysis, in particular to a film review sentiment analysis method based on an improved convolutional neural network model. Background technique [0002] In recent years, many people have begun to express their thoughts and opinions on the Internet. After watching the movie, leave your own movie reviews on Douban and other places to express your views on the movie. Data analysis of these massive emotional texts will help users have a better experience on the Internet and help the website operate better. The traditional movie recommendation method mainly uses the scoring record of the target customer to find users who are similar to him, or uses the user's historical preferences to find the first n movies similar to those he has watched in the past for recommendation. This kind of recommendation method is easy to operate. Higher precision. The disadvantages are also obvious. Some viewers may randomly sc...

AI Technical Summary

Problems solved by technology

[0005] The present invention aims to solve the above-mentioned problems that the existing film review text is short, there are many new words, and the word order is unreasonable, and the traditional sentiment analysis method can no longer reflect the correct semantics of the sentence. A method based on an improved convolutional neural network is proposed Modeling Method for Sentiment Analysis of Film Criticism

Images