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

Spoken language understanding method based on knowledge graph and semantic graph technology

A technology of spoken language understanding and knowledge graph, which is applied in the field of detecting the gist and detailed points of sentences, and can solve the problems of lack of effective use of sentence semantics

Active Publication Date: 2019-08-30
杭州淘艺数据技术有限公司
View PDF4 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, these works often focus on the structural information of the sentence itself, and lack the effective use of sentence semantics

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 understanding method based on knowledge graph and semantic graph technology
  • Spoken language understanding method based on knowledge graph and semantic graph technology
  • Spoken language understanding method based on knowledge graph and semantic graph technology

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0059] The present invention will be further described below in conjunction with the accompanying drawings.

[0060] refer to figure 1 and figure 2 , step 1, training sequence to action sequence neural network model, the model architecture is as follows image 3 As shown, training a neural network for spoken language understanding based on sentences and sentence logical expressions, the model architecture is as follows Figure 4 shown;

[0061] Step 2, read the natural language sentence that needs to be parsed;

[0062] Step 3. Use the text mapping algorithm to scan and replace the part of the sentence that matches the semantic resource in the knowledge graph;

[0063] Step 4. Use the sequence-to-action sequence neural network model to read the sentence after replacing the semantic resource, then execute the corresponding action sequence to generate a semantic graph that matches the semantics of the sentence, and then use the depth-first algorithm to traverse the semanti...

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 spoken language understanding method based on a knowledge graph and a semantic graph technology. The method comprises the following steps: 1, training a sequence to an actionsequence neural network model, and training a spoken language understanding neural network based on sentences and sentence logic expressions; 2, reading a natural language sentence needing to be analyzed; 3, scanning and replacing the part, matched with the semantic resources in the knowledge graph, in the sentence by using a text mapping algorithm; 4, reading the sentences with the semantic resources replaced through a sequence-to-action sequence neural network model, then executing corresponding action sequences to generate semantic graphs conforming to sentence semantics, and then using adepth-first algorithm for traversing the semantic graphs to obtain logical expressions of the sentences; and 5, reading sentences and logic expressions by using a spoken language understanding neuralnetwork, and generating intention information and slot position information. The invention provides a method for parsing spoken language by combining a knowledge graph and understanding sentence semantics.

Description

technical field [0001] The invention relates to the field of text matching, in particular to a method for understanding oral language based on knowledge graph and semantic graph technology, and a method for detecting the gist and detailed points of a sentence in a specified text. Background technique [0002] Human-machine dialogue systems need to recognize information in human language in order to perform specific tasks, such as answering questions, booking air tickets, etc. This process is also called spoken language analysis. In the spoken language parsing task, the task of identifying the gist of the sentence is called "intent detection", and screening requirements from sentences according to different intentions is called "slot filling". [0003] With the development of artificial intelligence technology, researchers have shifted from the traditional maximum entropy Markov model, conditional random field and other schemes to various models based on neural networks, and ...

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): G06F17/27G06F16/36G06F16/35
CPCG06F16/367G06F16/35G06F40/30
Inventor 姜明滕海滨张旻汤景凡戚铖杰张雯
Owner 杭州淘艺数据技术有限公司
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