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

Spoken language semantic comprehension method, device and system

A semantic understanding and spoken language technology, applied in semantic analysis, natural language data processing, electronic digital data processing, etc., can solve problems such as lack of pan-domain ability, unclear hierarchy, and sentence confusion

Pending Publication Date: 2021-03-12
GREE ELECTRIC APPLIANCES INC +1
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The template rule matching method is better for a small amount of semantic processing, but the gradual increase in the corpus will cause intention conflicts, chaotic sentence patterns, and unclear levels, which are difficult to maintain
The model trained by the machine learning method has domain adaptability to the labeled domain, but lacks the pan-domain capability. For new domain semantics, it needs to be re-labeled and trained to recognize

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
  • Spoken language semantic comprehension method, device and system
  • Spoken language semantic comprehension method, device and system
  • Spoken language semantic comprehension method, device and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0040] In order to solve the above-mentioned technical problems existing in the prior art, an embodiment of the present invention provides a method for understanding spoken language semantics.

[0041] The inventors found that colloquial short texts usually do not follow grammatical rules, are short in length, do not have enough information for statistical inference, and are difficult to make accurate inferences in a limited context. For example: the syntax structure can be arranged in any order, "the air conditioner is turned on" means the intention to turn on the air conditioner, and "is the air conditioner turned on" means to query the state of the air conditioner.

[0042] Embodiments of the present invention combine rule-based matching and machine learning for colloquial short texts:

[0043] Train text classification models and entity extraction models on the basis of existing label data to achieve semantic understanding in existing fields;

[0044] For semantic underst...

Embodiment 2

[0096] The embodiment of the present invention also provides a spoken language semantic understanding device, which includes:

[0097] A text classification unit, configured to perform text classification processing on the input spoken text based on the text classification model, to obtain the corresponding domain and intent of the spoken text;

[0098] An entity extraction unit, configured to perform entity extraction processing on the input spoken text based on the entity extraction model, to obtain entity parameters corresponding to the spoken text;

[0099] a scoring unit, configured to score the accuracy of the text classification process and the entity extraction process, and obtain a score value;

[0100] An output judging unit, configured to output structured information according to the domain, the intention and the entity parameter when the scoring value is greater than or equal to a preset scoring threshold.

[0101] The embodiment of the present invention also pro...

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 spoken language semantic comprehension method, device and system, and the method comprises the steps: carrying out the text classification of an inputted spoken language textbased on a text classification model, and obtaining a field and an intention corresponding to the spoken language text; performing entity extraction processing on the input spoken language text basedon an entity extraction model to obtain entity parameters corresponding to the spoken language text; scoring the accuracy of the text classification processing and the entity extraction processing toobtain a score value; and when the score value is greater than or equal to a preset score threshold, outputting structured information according to the domain, the intention and the entity parameters.According to the method, rule matching and a machine learning model are combined, on the basis of rapid and accurate recognition of spoken language semantics, matching and label setting of new fielddata are completed, and understanding of the new field spoken language data is achieved.

Description

technical field [0001] The invention belongs to the field of intelligent control, and in particular relates to a spoken language semantic understanding method, device and system. Background technique [0002] In the field of voice interaction, converting natural language into semantics that machines can understand is the core technical content. Semantics not only expresses the essence of things, but also expresses various logical relationships between things, such as causality, subordination, and facts. Semantic analysis and understanding is to identify the semantics contained in information and establish a model to make it understand related texts, among which Semantic understanding is inseparable from technologies such as rule matching, text classification, entity recognition and extraction. [0003] The current semantic understanding technology is mainly divided into two types: [0004] (1) Template rule matching method: Accurate matching is performed by setting spoken ...

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): G06F16/35G06F40/30G06F40/289
CPCG06F16/35
Inventor 李明杰宋德超贾巨涛吴伟黄姿荣
Owner GREE ELECTRIC APPLIANCES INC
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