Supercharge Your Innovation With Domain-Expert AI Agents!

Neural network training method, text similarity detection method and dialogue system

A neural network training and text similarity technology, applied in the field of information retrieval, can solve problems such as low accuracy rate, inconsistent training and prediction distribution, and achieve the effect of providing accuracy rate

Pending Publication Date: 2022-07-22
广州天宸健康科技有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The disadvantage of this method is that the cosine similarity is not added during the training, but the cosine similarity is calculated in the inference stage, resulting in inconsistent distributions of training and prediction, resulting in low accuracy

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
  • Neural network training method, text similarity detection method and dialogue system
  • Neural network training method, text similarity detection method and dialogue system
  • Neural network training method, text similarity detection method and dialogue system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0038] see figure 1 , this application discloses a neural network training method for the technical defects of the prior art, including the steps:

[0039] S1: Label the three sentences S1, S2, and S3;

[0040] S2: Input two of the marked three sentences into the neural network to obtain sentence features SF1, SF2, and SF3 respectively;

[0041] S3: Calculate the cosine similarity COS(1,2) of SF1 and SF2;

[0042] S4: Calculate the loss function J of positive and negative samples;

[0043] S5: Set the loss function L to enhance the distance and discrimination between positive and negative samples;

[0044] S6: Calculate the final loss function T, and train the neural network.

[0045] The technical solutions of the present application are further described below with reference to various preferred embodiments.

[0046] S1: Label the three sentences S1, S2, S3:

[0047] The purpose of this step is to turn the input sentence into a sentence format that can be processed by ...

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 a neural network training method, a text similarity detection method and a dialogue system. The neural network training method comprises the following steps: S1, labeling three sentences S1, S2 and S3; s2, inputting two sentences in the three marked sentences into a neural network to respectively obtain sentence features SF1, SF2 and SF3; s3, calculating the cosine similarity COS (1, 2) of the SF1 and the SF2; s4, calculating a loss function J of the positive and negative samples; s5, setting a loss function L, and enhancing the distance and the distinction degree of the positive and negative samples; and S6, calculating a final loss function T, and training the neural network. According to the technical scheme adopted by the invention, the neural network system obtained by training is closer to a real application result, so that the accuracy is greatly improved.

Description

technical field [0001] The invention relates to the technical field of information retrieval, in particular to a neural network training method, a text similarity detection method and a dialogue system. Background technique [0002] Natural language processing (NLP) is an important research direction in computer science and artificial intelligence. It mainly studies various theories and methods that can realize effective communication between humans and computers using natural language. It is a subject integrating linguistics, computer science, and mathematics. [0003] Question Answering is an important research field of Natural Language Processing, which can automatically give answers to questions posed by humans in the form of natural language. [0004] The current question answering system has been widely used, for example, it has been widely used in some daily searches, intelligent customer service, chat robots, etc., and has long been popular in our lives. For exampl...

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): G06N3/04G06N3/08G06F16/33G06F40/211G06K9/62
CPCG06N3/04G06N3/08G06F16/3344G06F40/211G06F18/22
Inventor 曾祥云朱姬渊
Owner 广州天宸健康科技有限公司
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