Text emotion information identification method and related device

Abstract

The invention discloses a text emotion information identification method, which comprises the following steps: preprocessing a to-be-predicted text by adopting a trained Word2vec model to obtain an input vector of the to-be-predicted text; carrying out feature recognition on the input vector by adopting a trained BiLSTM model to obtain context time sequence text information features; performing feature extraction on the context time sequence text information features through a trained CNN model to obtain target features; and identifying the target features by adopting the trained neural network to obtain an emotion identification result. Through the trained Word2vec model, BiLSTM model and CNN model, the time sequence characteristics of the context of the to-be-predicted text are extracted, the local characteristics and sequence information of the text are obtained, and the precision and accuracy of sentiment analysis are improved. The invention further discloses emotion information recognition equipment, a natural language analysis device and a computer readable storage medium which have the above beneficial effects.

Description

technical field [0001] The present application relates to the technical field of natural language analysis, and in particular to a text emotion information recognition method, emotion information recognition equipment, a natural language analysis device, and a computer-readable storage medium. Background technique [0002] Sentiment analysis is another analysis angle in the field of natural language processing, also known as tendency analysis, opinion extraction, opinion mining, emotion mining, subjective analysis, etc. It mainly analyzes, processes, summarizes and The process of reasoning, such as analyzing the user’s evaluation of the movie from the movie review, and analyzing the user’s emotional tendency to the product’s “price, size, weight, ease of use” and other attributes from the product review text. [0003] The main task of sentiment analysis for commodity reviews is to analyze, process, summarize and judge emotional texts. Traditional machine learning algorithms...

AI Technical Summary

Problems solved by technology

However, since each word or sentence in the text has a different determining effect on the emotional polarity of the entire text, in the above two neural networks, the former will ignore the contextual semantics of the word and a large amount of feature information will be lost during the maximum pooling operation. The latter has the problem of gradient disappearance and gradient dispersion.

Specifically, CNN is not fully suitable for learning time series, so various auxiliary processing is required, and the effect is not necessarily good

However, RNN only considers the one-way timing problem, does not fully integrate the context, and there will also be gradient disappearance and gradient explosion problems.

Both of the above situations will result in a decrease in the accuracy of the sentiment analysis of the text and reduce the accuracy of the sentiment analysis.

Images