Method for analyzing affective perspectives of English composition

Abstract

The invention provides a method for analyzing affective perspectives of an English composition. The method comprises an analysis model consisting of an English composition and model composition preprocessing module, an English composition affective analysis module, an English composition perspective analysis module and an English composition perspective analysis result generation module which are successively connected with one another. After an English composition is processed by the analysis model, an affective perspective analysis result of the English composition can be finally obtained. By adopting the method, the affective perspective and perspective of the English composition can be analyzed by virtue of fewer English model compositions, so that the problems of the traditional English composition affective perspective analysis method that correlation between a vocabulary and a vocabulary cannot be analyzed and the accuracy for analyzing the affective perspective of the English composition is low can be solved.

Description

technical field [0001] The invention relates to natural language processing technology and English short text analysis technology, in particular to a method for excavating emotional viewpoints of English compositions by using a computer system. Background technique [0002] Analyzing the expression of emotion and viewpoint in English composition is one of the important indicators to evaluate the written content of English composition. The existing sentiment analysis techniques for English composition are mainly divided into latent Dirichlet distribution methods and machine learning-related methods, but most of these methods use "bag of words" as the basic unit, and there is a problem of ignoring the relationship between words and words. The accuracy of mining the emotional viewpoints of English compositions is low. Before mining the emotional viewpoints of English compositions, it is necessary to use a large number of English sample essays to train and build an analysis mode...

AI Technical Summary

Problems solved by technology

Existing sentiment analysis techniques for English composition are mainly divided into latent Dirichlet distribution methods and machine learning-related methods, but most of these methods use "bag of words" as the basic unit, and there is a problem of ignoring the relationship between words and words. The accuracy rate of mining the emotional viewpoints of English compositions is low. Before mining the emotional viewpoints of English compositions, it is necessary to use a large number of English sample essays to train and build an analysis model, and collecting a large number of English sample essays is time-consuming and labor-intensive.

Images