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

A Method of Correcting Text After Speech Recognition Based on Domain Recognition

A speech recognition and text error correction technology, applied in speech recognition, speech analysis, text database query, etc., can solve problems such as a large number of manual interventions, low error correction efficiency, and inability to correct errors by proper names, and reduce a lot of time loss. , the data is accurate and true, the effect of enhancing practicability and robustness

Active Publication Date: 2021-01-26
SICHUAN CHANGHONG ELECTRIC CO LTD
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The technical problem to be solved by the present invention is: to propose a method for correcting text after speech recognition based on field recognition, which solves the problem that the processing method in the traditional technology requires a lot of manual intervention, the error correction efficiency is low, and the proper name cannot be corrected. error correction problem

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 Method of Correcting Text After Speech Recognition Based on Domain Recognition
  • A Method of Correcting Text After Speech Recognition Based on Domain Recognition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0037] The present invention aims to propose a method for text error correction after speech recognition based on field recognition, which solves the problems that the processing method in the traditional technology requires a lot of manual intervention, the error correction efficiency is low, and the proper name cannot be corrected.

[0038] The present invention adopts the Bigram model and the whoosh search engine to judge the domain of the input text. By introducing the Markov hypothesis, Bigram solves the problem of data sparseness and too large parameter space in n-grams, assuming that the appearance of a word only depends on the previous A word that appears, thus establishing the relationship between characters. The whoosh search engine helps to establish domain discrimination, and builds an index based on the input text, which can quickly realize the identification of candidate sets for fuzzy matching, and improve the speed of text error correction after semantic recogni...

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 belongs to the field of speech recognition and text processing, and discloses a method for correcting text after speech recognition based on field recognition, which solves the problem that the processing method in the traditional technology requires a lot of manual intervention, the error correction efficiency is low, and it cannot correct the proprietary The problem of name correction. The method includes the following steps: a. performing an error recognition analysis on the text after the speech recognition, and preliminarily determining the field to which the text sentence belongs; b. segmenting the sentence to be corrected according to the predefined grammatical rules, and dividing it into redundant parts and core parts ; c. Use the search engine to perform string fuzzy matching to determine the candidate proprietary thesaurus set for the core part of the sentence; d. Calculate the similarity score according to the edit distance, and correct the redundant part and the core part respectively. e. Fuse the error-corrected redundant part with the core part, and then output the error-corrected result.

Description

technical field [0001] The invention belongs to the field of speech recognition and text processing, and in particular relates to a method for correcting text after speech recognition based on field recognition. Background technique [0002] In recent years, the demand and development of artificial intelligence have been increasing day by day, and it has become a top priority for computers to correctly understand human language. Speech recognition can be mainly divided into pre-processing and post-processing processes. The pre-processing process mainly includes the process of voice signal processing, and extracts and analyzes the parameters of human / user speech, focusing on the processing of voice signals; The conversion of syllables to Chinese characters, in other words, is the process of converting voice signal information into internal codes recognizable by computers. In the actual post-processing of speech recognition, due to the possible psychological or emotional ups ...

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 Patents(China)
IPC IPC(8): G06F40/232G06F40/211G06F40/253G06F40/284G06F16/33G06F16/903G10L15/183G10L15/26
CPCG10L15/183G10L15/26G06F16/3343G06F16/90344G06F40/211G06F40/232G06F40/253G06F40/284
Inventor 杨鑫刘楚雄唐军
Owner SICHUAN CHANGHONG ELECTRIC 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