Supercharge Your Innovation With Domain-Expert AI Agents!

Statistics machine translation-based language generation method

A statistical machine translation and language generation technology, applied in natural language translation, instruments, computing, etc., can solve the problem that the statistical machine translation system is not an ideal language generation method, and achieve the effect of improving language generation performance

Active Publication Date: 2017-12-29
SUZHOU UNIV
View PDF14 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0015] The purpose of the present invention is to overcome the problem that the existing traditional standard statistical machine translation system is not an ideal language generation method

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
  • Statistics machine translation-based language generation method
  • Statistics machine translation-based language generation method
  • Statistics machine translation-based language generation method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0043] The present invention will be further described below in conjunction with the accompanying drawings.

[0044] Such asfigure 1 As shown, the language generation method based on statistical machine translation of the present invention comprises the following steps,

[0045] Step (A), corpus preprocessing, transforms the semantic expression of the source-end tree structure into a semantic expression of natural language, and the semantic expression of the source-end tree structure is a function-parameter form (also called tree structure expression formula), for example, the source-side semantic expression is "answer(len(river(riverid('colorado')))", the semantic expression of each natural language at the target side, Chinese: how long is the Colorado River; English: how long is the colorado river; tree structure expression such as figure 2 As shown; convert these semantic expressions into a series of strings in natural language at the source, the example of the preprocess...

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 statistics machine translation-based language generation method. The method comprises the steps of converting a semantic expression of a tree structure of a source end into a natural language; taking preprocessed corpora as a source end language of hierarchical phrase-based statistics machine translation; generating an n-best translation result by utilizing a hierarchical phrase-based statistics machine translation decoder; filtering the n-best translation result generated in the previous step to obtain a translation result with the same answer type as an input sentence of the source end; and in an evaluation stage, increasing the number of reference statements of Chinese and English to 3 from 1, thereby ensuring that an evaluation value can better reflect a real translation effect. The method has good application prospects.

Description

technical field [0001] The technical field of language generation of the present invention, be specifically related to a kind of language generation method based on statistical machine translation. Background technique [0002] The task of language generation is to analyze the semantic expression and convert it into natural language. Many early methods of language generation were rule-based, which generally only focused on surface realization, that is, adjusting the order and changing words, and then began to appear many methods based on corpus probability, many of which required semantic expressions to be of special form, For example, a tree structure expression. [0003] Much early language generation did not use probabilistic methods. In the early days, there was a method of using artificial labeling rules to generate natural language from extended predicate logic expressions, and then there was a semantic-driven method, which was based on rules written in a logic progr...

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): G06F17/28
CPCG06F40/44G06F40/58
Inventor 李军辉柴强孔芳周国栋
Owner SUZHOU UNIV
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com