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

Attribute-level sentiment analysis method of complete attention mechanism

A technology of sentiment analysis and attention, applied in the field of deep learning, can solve the problems of time overhead and low accuracy, and achieve the effect of improving accuracy and reducing costs

Active Publication Date: 2020-06-09
GUILIN UNIV OF ELECTRONIC TECH
View PDF10 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The present invention provides an attribute-level sentiment analysis method with a full attention mechanism, aiming at the problem of low time overhead and low accuracy when using deep learning to conduct sentiment analysis in user comments

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
  • Attribute-level sentiment analysis method of complete attention mechanism
  • Attribute-level sentiment analysis method of complete attention mechanism
  • Attribute-level sentiment analysis method of complete attention mechanism

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0038] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with specific examples and with reference to the accompanying drawings.

[0039] See figure 1 , An attribute-level sentiment analysis method with complete attention mechanism, including the following steps:

[0040] 1) Use sample data to train a network model based entirely on the attention mechanism:

[0041] Step 1. Sample data preprocessing

[0042] The given user comment sentences with real sentiment label classification are used as sample data, and the sample data is preprocessed.

[0043] The emotional polarity mark content is the emotional polarity of the user's comment sentence under the corresponding feature, which specifically contains three kinds of emotions: positive emotion, negative emotion, and neutral emotion.

[0044] The purpose of data preprocessing is to standardize data and construc...

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 attribute-level sentiment analysis method of a complete attention mechanism. According to the method, vocabulary-level semantic features and sentence-level semantic featuresare generated based on a self-attention mechanism network SAM-NN and a specific aspect attention mechanism network AAM-NN, and finally, the emotional polarity of comment sentence content is calculated through a full-connection neural network FC-NN output layer. The method provided by the invention is of a parallel structure in implementation, and specific aspect information characteristics are fused in each network computing module, so that the emotion polarity of the user comment information about the specific attribute aspect of the target object can be further analyzed according to the specific aspect information as much as possible. Compared with the prior art, the method not only effectively improves the accuracy of emotion analysis tasks in specific aspects, but also effectively reduces the expenditure on model training time.

Description

Technical field [0001] The present invention relates to the technical field of deep learning, in particular to an attribute-level sentiment analysis method with a complete attention mechanism. Background technique [0002] The rapid development and popularization of Internet technology has prompted the generation of a large amount of online comment information, such as various e-commerce trading platforms, social networking platforms, etc. How to effectively dig useful information from a large number of comments has become a field of natural language processing in recent years Research hotspots, including opinion mining tasks in user comment information. The emotions expressed in online comment information are usually diverse, and a user comment may contain different emotions in multiple attributes of the comment target. For example, a user comment from a restaurant: "The noodles are very delicious, but the soup is awful". This comment contains the characteristic information of...

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): G06F16/35G06F40/211G06F40/289G06F40/30G06N3/04G06N3/08
CPCG06F16/35G06N3/08G06N3/045
Inventor 林煜明傅裕李优周娅张敬伟
Owner GUILIN UNIV OF ELECTRONIC TECH