Dictionary correction and implicit emotion recognition method based on multi-task learning

A multi-task learning and emotion recognition technology, applied in the field of computer text emotion analysis, can solve the problems of poor generalization, time-consuming feature construction of recognition methods, etc., to achieve good generalization and improve the effect of recognition.

Active Publication Date: 2021-07-23
SHANXI UNIV
View PDF12 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] In the task of rhetoric recognition, the feature construction of existing recognition methods is time-consuming, and the specific model only solves one kind of rhetoric recognition, which has poor gener

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
  • Dictionary correction and implicit emotion recognition method based on multi-task learning
  • Dictionary correction and implicit emotion recognition method based on multi-task learning
  • Dictionary correction and implicit emotion recognition method based on multi-task learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0084] Such as figure 1 As shown, a rhetoric and implicit emotion recognition method based on multi-task learning of the present invention divides rhetoric and emotion recognition into three sub-modules, and each module is connected layer by layer, and finally it is fused by a multi-task mechanism. Training, specifically includes the following steps:

[0085] Step 1, semantic information encoding: for a sentence containing N words S={w 1 ,w 2 ,...,w N}, using the BERT model to capture the semantic representation sr of the sentence S sem , the specific steps are:

[0086] Step 1.1, normalize the sentence into the format required by the BERT model, that is, add [CLS] representation at the beginning of the sentence;

[0087] Step 1.2, using the output of [CLS] as the semantic representation of the entire sentence, as shown in formula (1):

[0088] sr sem =BERT(S)(1)

[0089] Among them, S stands for sentence, sr sem is the semantic representation of S, i.e. srsem for d...

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 relates to the field of computer text sentiment analysis, in particular to a dictionary correction and implicit emotion recognition method based on multi-task learning. The method is provided for recognizing and correcting dictionaries and emotions. Firstly, semantic and syntactic expressions of sentences are captured by utilizing BERT and Tree-LSTMs; on the basis, a dictionary revision classifier of a gating mechanism and an emotion classifier based on semantics are designed, and the dictionary revision of the sentences and the association distribution representation of the emotions are obtained respectively; and then, multi-label prediction integrated with the associated representation is built to obtain a label set of a sentence dictionary and an emotion.

Description

technical field [0001] The invention relates to the field of computer text emotion analysis, in particular to a rhetoric and implicit emotion recognition method based on multi-task learning. Background technique [0002] The implicit emotional expression of rhetoric exists widely in literary works, product reviews and other texts. Carrying out research on rhetoric and sentiment analysis can provide technical support for smart education and product public opinion analysis. In intelligent education, answering reading comprehension questions about language appreciation in literary works, such as "reading materials express reverence for life from multiple perspectives, please select a detail to analyze language characteristics", requires the support of rhetorical and emotional knowledge . The automatic recognition technology of rhetoric and emotion can help students quickly analyze and answer exercises and consolidate relevant knowledge points, thereby helping students improve ...

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
IPC IPC(8): G06F40/211G06F40/30G06K9/62G06N3/04G06N3/08
CPCG06F40/211G06F40/30G06N3/08G06N3/048G06N3/044G06F18/241
Inventor 陈鑫王素格李德玉
Owner SHANXI UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap