Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A Cross-Domain Sentiment Classification Method Based on Multi-source Domain Integrated Transfer

An emotion classification, cross-field technology, applied in the direction of equipment, marketing, data processing applications, etc., can solve the problems such as poor effect of emotion classifiers, and achieve the effect of improving the effect of emotion classification

Active Publication Date: 2021-06-18
KUNMING UNIV OF SCI & TECH
View PDF11 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The present invention provides a cross-domain emotion classification method based on multi-source domain integration migration to solve the problem that the effect of the emotion classifier trained using the label corpus of a source domain is not good, and the invention improves the effect of emotion classification in the target domain

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
  • A Cross-Domain Sentiment Classification Method Based on Multi-source Domain Integrated Transfer
  • A Cross-Domain Sentiment Classification Method Based on Multi-source Domain Integrated Transfer
  • A Cross-Domain Sentiment Classification Method Based on Multi-source Domain Integrated Transfer

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0026] Embodiment 1: as Figure 1-4 As shown, the cross-domain sentiment classification method based on multi-source domain integration transfer,

[0027] The concrete steps of described emotion classification method are as follows:

[0028] Step1. Using the Feature-extending NeuralStructural Correspondence Learning (FNSCL) model to obtain the source domain D s Migrate to the features of the target domain Dt, and then train different logistics classifiers, D s1 to D t Training to get logistics classifier 1, D s2 to D t Training to get logistics classifier 2, D s3 to D t Train to get logistics classifier 3;

[0029] As a preferred solution of the present invention, the FNSCL model uses the extended pivot feature to perform feature migration, and not only considers the mutual information MI value between the feature and the label in the training set when screening the pivot feature, but also considers the word frequency of the feature, And the TF-IDF feature selection al...

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 a cross-domain emotion classification method based on multi-source domain integration migration, and belongs to the field of computer and information technology. The present invention comprises the steps: first, the neural structure corresponding learning FNSCL model of feature expansion is used to obtain the source domain D s Migrate to target domain D t features, and train different logistics classifiers; then according to the principle of ensemble consistency, calculate the weight of the logistics classifier; finally use the simulated annealing algorithm to optimize the weight of each classifier. The present invention can improve the emotion classification effect of cross-domain target domain, and the product review (electronic product review, book review, kitchen utensils review, DVD review) emotion classification experiment result of Amazon's four different fields shows that the multi-source domain-based A cross-domain sentiment classification approach that integrates transfer is effective.

Description

technical field [0001] The invention relates to a cross-domain emotion classification method based on multi-source domain integration migration, and belongs to the field of computer and information technology. Background technique [0002] At present, e-commerce has had a huge impact on our lives. The comment texts of historical consumers on e-commerce products are subtly affecting the purchase behavior of other consumers. Therefore, research on sentiment classification of product reviews on e-commerce platforms has gradually become a new research hotspot. However, the sentiment classification of product reviews is very domain-specific, and the performance of a sentiment classifier trained on labeled data in a certain field will drop more when applied to other fields. In response to this problem, the existing technologies are: Aue et al. proposed to train a classifier by mixing a small amount of labeled data and unlabeled data, the effect is good, and the trained model can...

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 Patents(China)
IPC IPC(8): G06K9/62G06Q30/06G06Q30/02
CPCG06Q30/0623G06Q30/0203G06F18/24G06F18/214
Inventor 相艳陆婷余正涛郭军军线岩团许莹
Owner KUNMING UNIV OF SCI & TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com