Training method, device, equipment and medium for cross-domain text sentiment classification model

A sentiment classification and cross-domain technology, applied in the field of natural language processing, can solve the problems of inaccurate sentiment classification of cross-domain text sentiment classification models, and the inability to provide cross-domain text sentiment classification models, etc., to achieve the effect of improving the accuracy of sentiment classification

Active Publication Date: 2021-06-04
SHENZHEN UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a training method, device, equipment and medium for a cross-domain text sentiment classification model, aiming to solve the problem of cross-domain The problem of inaccurate sentiment classification of text sentiment classification model

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
  • Training method, device, equipment and medium for cross-domain text sentiment classification model
  • Training method, device, equipment and medium for cross-domain text sentiment classification model
  • Training method, device, equipment and medium for cross-domain text sentiment classification model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0039] figure 1 The implementation process of the training method of the cross-domain text sentiment classification model provided by the first embodiment of the present invention is shown. For the convenience of explanation, only the parts related to the embodiment of the present invention are shown, and the details are as follows:

[0040] In step S101, the pre-built cross-domain text sentiment classification model is initially trained according to the preset source domain sample set and the preset target domain sample set.

[0041] The embodiments of the present invention are applicable to text or natural language processing platforms, systems or devices, such as personal computers, servers, and the like. The pre-built cross-domain text sentiment classification model is first trained according to the preset source domain sample set and the preset target domain sample set, wherein the source domain sample set contains labeled samples and unlabeled samples, and the target dom...

Embodiment 2

[0068] figure 2 The structure of the training device for the cross-domain text sentiment classification model provided by Embodiment 2 of the present invention is shown. For the convenience of description, only the parts related to the embodiment of the present invention are shown, including:

[0069] The model initial training unit 21 is used to perform initial training on the pre-built cross-domain text sentiment classification model according to the preset source domain sample set and the preset target domain sample set;

[0070] The emotion classification unit 22 is used to perform emotion classification on the target field sample set according to the trained cross-domain text emotion classification model, and obtain the emotion prediction label corresponding to each target sample in the target field sample set and each target sample belonging to each preset The membership degree of an emotional category;

[0071] The fuzzy value calculation unit 23 is used to calculate ...

Embodiment 3

[0080] Figure 4 The structure of the computing device provided by the third embodiment of the present invention is shown, and for the convenience of description, only the parts related to the embodiment of the present invention are shown.

[0081] The computing device 4 of the embodiment of the present invention includes a processor 40 , a memory 41 and a computer program 42 stored in the memory 41 and operable on the processor 40 . When the processor 40 executes the computer program 42, it realizes the steps in the embodiment of the training method of the above-mentioned cross-domain text sentiment classification model, for example figure 1 Steps S101 to S105 are shown. Alternatively, when the processor 40 executes the computer program 42, the functions of the units in the above-mentioned device embodiments are realized, for example figure 2 Function of units 21 to 25 shown.

[0082] In the embodiment of the present invention, after the initial training of the cross-doma...

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 present invention is applicable to the technical field of natural language processing, and provides a training method, device, equipment and medium for a cross-domain text sentiment classification model. The method includes: classifying the cross-domain text sentiment according to the source domain sample set and the target domain sample set After the initial training of the model, according to the trained cross-domain text sentiment classification model, the target field sample set is sentimentally classified, and the sentiment prediction label corresponding to each target sample in the target field sample set and the sentiment prediction label of each target sample belonging to each emotion category are obtained. According to the degree of membership, the emotional fuzzy value of each target sample is calculated by the fuzzy value formula, and the target samples whose emotional fuzzy value is lower than the fuzzy threshold and the corresponding emotional prediction labels of the target samples are added to the source domain sample set, according to the The source domain sample set and the target domain sample set retrain the cross-domain text sentiment classification model, thereby improving the sentiment classification accuracy of the cross-domain text sentiment classification model.

Description

technical field [0001] The invention belongs to the technical field of natural language processing, and in particular relates to a training method, device, equipment and medium for a cross-domain text emotion classification model. Background technique [0002] Text sentiment analysis is domain-sensitive, since reviews in different domains have different word distributions, for example, the word "clean" is common in hotel reviews, but rarely used in book reviews, so only labeled If the sentiment classifier trained on book review data predicts the sentiment tendency of unlabeled hotel review data, it cannot obtain satisfactory results in hotel reviews. [0003] Domain adaptive algorithms aim to improve the prediction performance of target domain samples by using a large number of labeled samples from related domains (source domain). Domain Adversarial Neural Network (DANN for short) uses domain classifiers and gradient inversion layers for unsupervised cross-domain sentiment ...

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): G06F16/35
Inventor 傅向华刘旺旺
Owner SHENZHEN 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