Unlock instant, AI-driven research and patent intelligence for your innovation.

Emotion early warning system and method based on multi-head attention mechanism and convolution fusion

A technology of attention and attention model, which is applied in biological neural network models, instruments, electrical digital data processing, etc., and can solve problems such as no fine-grained emotional prediction

Pending Publication Date: 2022-02-08
NANJING YUNCHUANG LARGE DATA TECH CO LTD
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] At present, most of the emotional early warning schemes, that is, to predict the emotion of the text and the early warning feedback, are to make an overall emotional prediction of the entire text, and do not perform fine-grained emotional prediction

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Emotion early warning system and method based on multi-head attention mechanism and convolution fusion
  • Emotion early warning system and method based on multi-head attention mechanism and convolution fusion
  • Emotion early warning system and method based on multi-head attention mechanism and convolution fusion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0051] Embodiments of the present invention will be further described in detail below in conjunction with the accompanying drawings.

[0052] Such as figure 1 As shown, an emotional early warning system based on the fusion of multi-head attention mechanism and convolution, including:

[0053] A text monitoring and acquisition module is used to acquire and monitor the comment text information to be detected in real time;

[0054] A preprocessing module is used to preprocess the text obtained in the text monitoring module;

[0055] The sentiment analysis module is used to perform fine-grained sentiment analysis on the preprocessed text using the BERT-DPCNN model and the multi-head sub-attention model;

[0056] An alarm module is configured to generate alarm information for feedback when the detected text information reaches an alarm threshold.

[0057] An emotional early warning method based on the fusion of multi-head attention mechanism and convolution, including:

[0058]...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an emotion early warning system and method based on multi-head attention mechanism and convolution fusion. The method comprises the following steps: acquiring and monitoring comment text information to be detected in real time; preprocessing the obtained text; utilizing a BERT-DPCNN model and a multi-sub attention model to carry out fine-grained sentiment analysis on the preprocessed text; when it is detected that the text information reaches an alarm threshold value, generating alarm information for feedback. According to the method, real-time emotion early warning can be carried out on various comment information including take-out comments, hotel comments, microblog comments, film and television comments and the like in the whole network. Based on the system, the emotion trend of each aspect expressed in the comments can be effectively predicted, and early warning information is generated in time to facilitate rectification or public opinion control.

Description

technical field [0001] The invention belongs to the technical field of natural language processing, and in particular relates to an emotion early warning system and method based on the fusion of multi-head attention mechanism and convolution. Background technique [0002] Sentiment analysis is ubiquitous and it is a classification technique based on natural language processing. The main problem it solves is to judge the emotional tendency of a given sentence. For example, on Amazon.com or Twitter, people post comments about a product, event or person. Merchants can use sentiment analysis tools to know users' experience and evaluation of their products. [0003] When large-scale sentiment analysis is required, the processing power of the naked eye becomes very limited. The essence of sentiment analysis is to speculate whether the text is positive or negative based on known text and emotional symbols. After dealing with sentiment analysis, it can greatly improve the effici...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F16/35G06F16/33G06F40/216G06F40/242G06N3/04
CPCG06F16/35G06F16/3335G06F16/3344G06F40/216G06F40/242G06N3/047G06N3/045
Inventor 刘鹏张真张堃曹骝高中强严乐乐
Owner NANJING YUNCHUANG LARGE DATA TECH CO LTD