Method and device for classifying sentiment words in bullet screen text and storage medium

Abstract

The invention discloses a method and device for classifying sentiment words in a bullet screen text, and the method comprises the following steps: carrying out the preprocessing of bullet screen datacrawled by Python, judging whether to-be-classified bullet screen data contains the sentiment words or not, and directly carrying out the classification through a GRU classifier if the to-be-classified bullet screen data does not contain the sentiment words, and obtaining a classification result; if the emotion words are included, constructing a multi-dimensional bullet screen emotion dictionary in the bullet screen text field, and constructing a text emotion calculation classifier on the basis of the multi-dimensional bullet screen emotion dictionary; constructing a bullet screen emotion classification model by adopting an ensemble learning strategy of model fusion; and inputting the test set data into the bullet screen emotion classification model to obtain an emotion classification result of the bullet screen. According to the method, an emotion dictionary is expanded, GRU, naive Bayes and seven-dimensional emotion calculation classification methods are used as base classifiers, voting fusion is carried out according to results obtained by the base classifiers to output a final emotion classification result, and the bullet screen short text emotion word classification problem issolved.

Description

technical field [0001] The invention belongs to the field of network public opinion, and in particular relates to a method, device and storage medium for classifying emotional words in bullet chat text. Background technique [0002] In recent years, with the continuous development of the Internet, barrage video websites have become popular. More and more teenagers are used to expressing their opinions by sending barrage while watching this type of video. However, in an environment with social factors such as barrage videos, once public opinion or videos of hot events appear, they will spread rapidly and cause a huge impact. Most of the barrage senders are young people who are not deeply involved in the world, and they are easily led astray by the barrage that deliberately provokes them. These negative and provocative barrages can easily have a negative impact on them and hinder the development of young people's physical and mental health. Therefore, if it is not controlled...

AI Technical Summary

Problems solved by technology

But the disadvantage is that it needs to manually label the data category, and there are certain limitations in semantic expression

[0004]Nowadays, a large number of popular words have emerged on the Internet. With the continuous appearance of these words, the traditional emotional dictionary can no longer meet the needs of this new vocabulary. The follow-up analysis of the sentence, especially the sentence with the exclusive hot words in the barrage

Commonly used facial expressions, colloquialism, and symbolization are one of the most distinctive features of bullet-screen texts, and facial expressions are a relatively strong expression of emotion, while traditional emotional dictionaries choose to ignore these symbolic expressions. A large amount of information will be lost in the process of

Images