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

Natural language understanding method in small sample scene

A natural language understanding and small-sample technology, applied in semantic analysis, digital data processing, special data processing applications, etc., can solve problems such as difficulty in obtaining manually labeled data

Active Publication Date: 2021-08-31
国科(吉林)知识产权运营有限公司
View PDF8 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] For dialogue systems, deep learning technology can use a large amount of data to learn intent recognition and slot mapping in natural language understanding. However, current deep learning methods require a large amount of labeled training data. In real landing scenarios, a large number of artificial Annotated data are often difficult to obtain

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
  • Natural language understanding method in small sample scene
  • Natural language understanding method in small sample scene
  • Natural language understanding method in small sample scene

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0099] According to the technical solution of the present invention, a multi-round dialogue system has been developed to demonstrate the natural language understanding and recognition effect of this patent in small sample scenarios. The system is distributed according to three levels of the WeChat applet front-end, middle control layer, and back-end system . The front-end of the applet is mainly responsible for receiving the user's input sentences, sending them to the dialogue understanding module, and at the same time generating corresponding responses from the system to display the user, realizing multiple rounds of interaction between the user and the machine. The middle layer is responsible for connecting the front-end and the back-end, controlling the back-end system according to the input and semaphore of the front-end, and receiving the running results of the back-end to feed back to the front-end interface. The background system is mainly natural language understanding...

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 provides a natural language understanding method in a small sample scene; the method proposes pre-training model language semantic representation, intention recognition and slot position recognition, and introduces tag semantics; then a linear space mapping method is used to draw a semantic representation distance, a gating network is established, slot information and intention information are fused, and abstract label transition probability is used to achieve the purpose of rapid learning and understanding in different fields. According to the method, the intention of the problem can be better judged in a small sample scene, and the slot position of the problem is identified, so that the problems of insufficient data, too high data labeling cost and too high model migration cost under a natural language understanding task of a task type dialogue system are well solved.

Description

technical field [0001] The invention belongs to the field of natural language understanding, and in particular relates to a natural language understanding method in a small sample scenario. Background technique [0002] The human-computer dialogue system is a human-computer two-way information interaction system that regards the machine as a cognitive subject. It is a way to realize human-computer interaction; this technology can make human-computer interaction as convenient as human-to-human communication. In recent years, more and more dialogue systems have emerged. According to their specific applications, these dialogue systems can be roughly divided into two categories: one is task-oriented dialogue systems, such as Ali Xiaomi, Xiaomi's Xiaoai Assistant, etc.; One type is non-task-oriented chatbots, such as Microsoft Xiaobing. In a task-oriented dialogue system, users have a clear purpose and hope to obtain information or services that meet characteristic constraints, ...

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): G06F16/332G06F16/35G06F40/30
CPCG06F16/3329G06F16/35G06F40/30
Inventor 赵铁军朱聪慧郑德权衣景龙曹海龙徐冰杨沐昀
Owner 国科(吉林)知识产权运营有限公司
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