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

A sentence similarity recognition method for voice answer system

A sentence similarity and question answering system technology, applied in the computer field, can solve problems such as not being 100% accurate, and the recognition rate is declining

Inactive Publication Date: 2007-12-12
INST OF AUTOMATION CHINESE ACAD OF SCI
View PDF0 Cites 30 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Compared with the general question answering system, it has two outstanding features: (1) the sentence input by voice has the characteristics of spoken language; (2) the result of the user's input after voice recognition is not 100% accurate, and the voice recognition rate is related to the surrounding environment, When the noise is large, the recognition rate will drop significantly
Since the traditional sentence similarity calculation method is aimed at accurate text input and does not take into account the above two characteristics, it is necessary to design a sentence similarity calculation method specifically for voice question answering systems

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 sentence similarity recognition method for voice answer system
  • A sentence similarity recognition method for voice answer system
  • A sentence similarity recognition method for voice answer system

Examples

Experimental program
Comparison scheme
Effect test

example 1

[0046] User input: Hello, what's your name, please?

[0047] Speech recognition result: How are you? What is your name, please?

[0048] Step S1: It involves three questions in the FAQ knowledge base: "Hello", "What's your name", "What's your name?" They are respectively combined with keywords "Hello", "What's your name", " What's the name" to indicate;

[0049] Step S2: Extract keywords from the speech recognition results to obtain the keyword combination: "What's your name?" At this time, n=3, W 1 = Hello, W 2 = What is it called, W 3 = name,

[0050] When k=1, sequence one: hello m 1 = 1

[0051] Sequence 2: What is it called m 1 = 2

[0052] Sequence three: name m 1 =3

[0053] When k=2,

[0054] Sequence Four: What's your name m 1 = 1, m 2 = 2

[0055] Sequence Five: Hello name m 1 = 1, m 2 =3

[0056] Sequence Six: What's the name m 1 = 2, m 2 =3

[0057] When k=3,

[0058] Sequence Seven: Hello, what's your name m 1 = 1, m 2 = 2, m 3 =3

[0059]...

example 2

[0069] User input: when were you born

[0070] Speech recognition results: when was the ability to be born

[0071] Step S1: It involves two questions in the FAQ knowledge base: "What are your abilities" and "When were you born", which are represented by the keyword combination "Abilities" and "When were you born" respectively;

[0072] Step S2: Extract keywords from the speech recognition results to obtain the keyword combination: "Time to be born", at this time, n=3, W 1 = time, W 2 = ability, W 2 = born

[0073] When k=1,

[0074] Sequence one: time m 1 = 1

[0075] Sequence Two: Skills m 1 = 2

[0076] Sequence Three: Birth m 1 =3

[0077] When k=2,

[0078] Sequence 4: Time and skill m 1 = 1, m 2 = 2

[0079] Sequence Five: Time Birth m 1 = 1, m 2 =3

[0080] Sequence Six: Ability to be born m 1 = 2, m 2 =3

[0081] When k=3,

[0082] Sequence Seven: Time Matters Birth m 1 = 1, m 2 = 2, m 3 =3

[0083] According to the above method, two question sen...

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 relates to a sentence similarity identification method in audio answering system that comprises the key word combination representing the usual questions in the knowledge base, abstracting the key word in the audio identification result, matching with the normal knowledge base to get the candidate questions, deciding the similarity equation and identification result matching procedure.It solves the accuracy issue of dialectic audio identification, using sentence similarity value to get the intent of users in a relatively accurate way.

Description

technical field [0001] The present invention belongs to the field of computer technology and relates to a voice question answering system, in particular to a sentence similarity recognition method applied in a voice question answering system, especially a sentence similarity recognition method that considers the characteristics of spoken language and the accuracy of voice recognition results . Background technique [0002] The calculation of sentence similarity is an important theoretical basis in automatic question answering system. In the question answering system based on Frequently-Asked Question (FAQ for short), the FAQ knowledge base is an integral part of the automatic question answering system, which stores frequently asked questions and related answers by users. For the question input by the user, the answer is first searched in the FAQ knowledge base, and if the corresponding question can be found, the answer corresponding to the question is directly returned to t...

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): G10L15/08G10L15/02G10L15/00G10L17/14
Inventor 李成荣高倩倩
Owner INST OF AUTOMATION CHINESE ACAD OF SCI
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