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

Methods and systems for establishing language model of address book names and searching voice

A technology of language model and establishment method, which is applied in speech analysis, speech recognition, special data processing applications, etc., can solve problems such as the difficulty in realizing speech recognition of names and the inability to have a good accuracy rate of speech search in address books, and achieve improvement The effect of voice search accuracy

Inactive Publication Date: 2012-10-10
BEIJING BAIDU NETCOM SCI & TECH CO LTD
View PDF8 Cites 27 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The popularity of smart phone terminals makes it possible to realize voice search anytime and anywhere through smart phone terminals, which brings convenience to user operations. Among them, the voice query of the address book is when the user wants to input the voice of a person's name, he can query the address book corresponding to the name information, but it is precisely because of the difficulty of voice recognition of names that the voice search of the address book has not been able to have a good accuracy rate.

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
  • Methods and systems for establishing language model of address book names and searching voice
  • Methods and systems for establishing language model of address book names and searching voice
  • Methods and systems for establishing language model of address book names and searching voice

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0046] figure 1 The flow chart of the method for establishing a language model of the name of the address book provided by Embodiment 1 of the present invention, such as figure 1 As shown, the method includes the following steps:

[0047] Step 101: Collect name information from the search log of the text search.

[0048] The text search involved in this step may include, but is not limited to: text search through a browser, and search after voice is converted into text during voice search.

[0049]When collecting personal name information, it can be realized manually, it can also be extracted from text based on preset extraction rules, and it can also be obtained by matching with the personal name database of the input method. Taking Chinese personal names as an example, since the first character of Chinese personal names is surname, the surname can be enumerated, and other characters are first names, and the length of the first name is usually within three characters, so th...

Embodiment 2

[0063] figure 2 The voice search flow chart of the name of the address book provided by Embodiment 2 of the present invention, such as figure 2 As shown, the process may include the following steps:

[0064] Step 201: the client receives the user's voice request, and sends the voice request to the server.

[0065] In this embodiment, the user's voice request mainly refers to a voice request including name information, that is, the user inputs the name by voice, and the client first sends the voice request to the server for recognition.

[0066] Step 202: The server uses the name syllable language model or the final language model to recognize the voice request, and returns the recognized name syllables to the client.

[0067] The server matches the voice request with the name syllable language model or the final language model, and determines the matched name syllable, which is the recognition result. As for the recognition process through the language model, it is the sam...

Embodiment 3

[0074] image 3 The voice search flow chart of the phone number in the address book provided by Embodiment 3 of the present invention, such as image 3 As shown, the process may include the following steps:

[0075] Step 301: the client receives the user's voice request, and sends the voice request to the server.

[0076] In this embodiment, the user's voice request refers to a voice request including phone number information, that is, the user inputs a phone number by voice.

[0077] Step 302: The server uses the final language model to recognize the voice request, and returns the recognized number string to the client.

[0078] The server matches the voice request with the final language model, and determines the matched digit string, which is the recognition result of the phone number.

[0079] Step 303: The client uses the number string returned by the server to match in the local address book, and displays the matching result to the user.

[0080] In this step, the nu...

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 methods and systems for establishing a language model of address book names and searching voice. The method for establishing the language model of the address book names comprises the following steps of: acquiring name information from a search log of text search; marking syllables for the acquired name information; training a name syllable language model by using syllable marking of the name information; and combining the name syllable language model and a digital probability model, setting the probability from syllables to digits and from digits to syllables to be 0, and thus obtaining the final language model. The method for searching the voice comprises the following steps of: receiving a voice request of a user and sending the voice request to a server by a client; identifying the voice request by using the name syllable language model or the final language model and returning the identified name syllables to the client by the server; and matching the name syllables returned by the server in a local address book and revealing the matching result to the user by the client. By the methods and the systems, the voice search accuracy of the address book can be improved.

Description

【Technical field】 [0001] The invention relates to the field of computer application technology, in particular to a method for establishing a language model of names in an address book, a voice search method and a system thereof. 【Background technique】 [0002] With the rapid development of speech recognition technology, the recognition of common speech has been able to achieve a high accuracy rate, because the name basically does not have linguistic information, it is difficult to be used as a "word pair", so the speech recognition of personal names has always been an important field in this field. difficulty. The popularity of smart phone terminals makes it possible to realize voice search anytime and anywhere through smart phone terminals, which brings convenience to user operations. Among them, the voice query of the address book is when the user wants to input the voice of a person's name, he can query the address book corresponding to the name information, but it is pr...

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): G06F17/30G10L15/08
Inventor 曹立新万广鲁
Owner BEIJING BAIDU NETCOM SCI & TECH CO LTD
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