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

A model training method, text similarity determination method and device

A text similarity and model training technology, applied in character and pattern recognition, instrumentation, calculation, etc., can solve the problem of model accuracy reduction

Active Publication Date: 2021-03-16
HUNDSUN TECH
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] With the rapid development of the Internet and artificial intelligence technology, similar search and question and answer based on natural language have become necessary skills for major websites, APP (Application, application program), intelligent customer service system, etc. The text similarity model is the key model in the process. The text similarity model is used to judge whether two texts are similar and the degree of similarity between the two texts. However, the current text diversification reduces the model accuracy of the text similarity 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
  • A model training method, text similarity determination method and device
  • A model training method, text similarity determination method and device
  • A model training method, text similarity determination method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0037] Through research on the current text similarity model, the applicant found that the training difficulty of the text similarity model is mainly concentrated in two aspects. Robustness problem of metric definition.

[0038] Among them, regarding the model training and model parameter adjustment for various text similarity data annotation, the current text similarity annotation includes two text annotation methods, one is the triple annotation form (text A, text P, text N), which satisfies The similarity between text A and text P is greater than the similarity between text A and text N. Generally, the text similarity model is trained by constructing the triplet loss loss function. The other is the 0-1 labeling form (text A, text B, label 0 / 1), 0 means text A is not similar to text B, 1 means text A is similar to text B, generally based on classification loss function training text similarity Model. These two text annotation methods have their own advantages and disadvant...

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 application provides a model training method, a text similarity determination method and a device to obtain a text set used for training the text similarity model, the text set includes a first text group and a second text group, and the text in the first text group Each text is marked with the first text markup form, and each text in the second text group is marked with the second text markup form; each text is marked with the target text markup form, so as to unify each text in the first text group and the second text group The labeling form of the text; the loss value of the first text group is obtained based on the first loss function, and the loss value of the second text group is obtained based on the second loss function; based on the loss value of the first text group and the loss value of the second text group, Determine the loss value of the text set; use the loss value of the text set to adjust the model parameters of the text similarity model to train the text similarity model, so that the training method of the text similarity model has stronger compatibility and improves the accuracy of the model .

Description

technical field [0001] The present application belongs to the technical field of artificial intelligence, and in particular, relates to a model training method, a text similarity determination method, and an apparatus. Background technique [0002] With the rapid development of the Internet and artificial intelligence technology, natural language-based similarity search and question answering have become essential skills for major websites, APPs (Applications), intelligent customer service systems, etc. The text similarity model is the key model in the process. The text similarity model is used to determine whether two texts are similar and the degree of similarity when the two texts are similar. However, the current text diversity expression reduces the model accuracy of the text similarity model. . SUMMARY OF THE INVENTION [0003] In view of this, the purpose of the present application is to provide a model training method, text similarity determination method and appa...

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): G06F40/194G06K9/62
CPCG06F40/194G06F18/22G06F18/214
Inventor 王炯亮娄东方林金曙高峰陈哲许浩
Owner HUNDSUN TECH
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com