[go: up one dir, main page]

WO2007121614A1 - Procédé de traduction automatisée pour la traduction d'une langue en plusieurs langues - Google Patents

Procédé de traduction automatisée pour la traduction d'une langue en plusieurs langues Download PDF

Info

Publication number
WO2007121614A1
WO2007121614A1 PCT/CN2006/000805 CN2006000805W WO2007121614A1 WO 2007121614 A1 WO2007121614 A1 WO 2007121614A1 CN 2006000805 W CN2006000805 W CN 2006000805W WO 2007121614 A1 WO2007121614 A1 WO 2007121614A1
Authority
WO
WIPO (PCT)
Prior art keywords
language
information code
thinking
meaning
voice
Prior art date
Application number
PCT/CN2006/000805
Other languages
English (en)
Chinese (zh)
Inventor
Wenhe Xu
Original Assignee
Wenhe Xu
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Wenhe Xu filed Critical Wenhe Xu
Priority to PCT/CN2006/000805 priority Critical patent/WO2007121614A1/fr
Publication of WO2007121614A1 publication Critical patent/WO2007121614A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/263—Language identification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/40—Processing or translation of natural language

Definitions

  • the present invention relates to a method of translating a language into another language or simultaneously translating into multiple languages, and is particularly suitable for translation in a field of information such as a network language, and can be applied in information industries such as computers, mobile phones, and fixed telephones.
  • the homonym and homograph have very different meanings, such as Chinese “big”, ⁇ can be interpreted as a confirmation Kind of tea. 2 can be interpreted as the specification of the wood (the language in the industry). 3 can be interpreted as unscrupulous, unconstrained, not tacky. In English, for example, "bank” can be interpreted as a bank that can also be interpreted as a river bank.
  • the object of the present invention is to provide an automatic translation method that does not rely on speech or text for translation.
  • speech or text is sorted by a language thinking recognition system, it is converted into an item, and the meaning is converted to Information code (it can be numbers, letters, symbols), information ⁇ becomes no language and text features, becomes a common standard of various languages, automatically exchanges with various languages on the information code exchange platform, through the mind recognition system information code Turn it into a meaning and switch to another voice or text.
  • the principle of the present invention It is well known that words constitute vocabulary, and the purpose of vocabulary is to express meanings.
  • the traditional input method of general vocabulary input (one party) is that when the thinking choice intention is determined, the expression method is a meaning item, and the expression method is a vocabulary representing the meaning item.
  • the expression method In order to express the vocabulary in words, it is necessary to consider inputting the vocabulary in the order of the alphabet or the strokes, and then input it.
  • the main working method of the invention is to convert the meaning item into the information code through the language thinking recognition system according to the human thinking habit
  • the exchange platform converts the information code into the meaning item, converts the meaning item into the vocabulary, and only needs to pass the thinking intention in the specific application. If you choose the meaning, if you have the intention to think, and you can't find the right words in your memory to express the meaning of the intention, what should you do? For example: In the case of reluctance to agree with something, in the choice of vocabulary, only the (identification, recognition, permission, permission, consent) is obviously inaccurate. The traditional input method can only be used.
  • the language thinking recognition system decomposes the polysemous items of the word into multiple meaning items, and multiple meaning items are displayed in the dialog box by voice or man-machine.
  • the selected single item is transmitted to the information code exchange platform, and the information code exchange platform is automatically exchanged with a plurality of languages or words, and finally the desired voice or text is output through the language broadcast system or the display system.
  • T N2006/000805 The above language thinking recognition system sets a multilingual vocabulary storage database and a mind recognition network information code base.
  • the present invention Compared with the prior art, the present invention has the following advantages: the present invention provides an automatic translation method that does not rely on voice or text for translation.
  • the present invention realizes the basis of language translation without voice and text. It is the conversion of speech and text through the language identification system into the exchange of thinking information codes, realizing the one-to-one exchange of thinking information codes in various languages, ensuring the accuracy of language intention exchange.
  • 1 is a schematic diagram showing the working principle of the language thinking recognition system of the present invention.
  • FIG. 2 is a schematic diagram of an identification information code network in the language thinking recognition system of the present invention. (It can be in any language, but the information code does not change.)
  • Figure 3 is a schematic diagram of the database of thinking recognition languages of the present invention.
  • Figure 4 is a diagram showing the overall translation system structure in which a language is simultaneously translated into multiple languages.
  • Figure 5 is a schematic diagram of an automatic translation of a language that can be translated into multiple languages. Detailed ways:
  • the vocabulary expression is selected through the intention of Figure 1.
  • the speech signal input method is used to input the language signal to the language thinking recognition system.
  • the language thinking recognition system will sort the vocabulary of multiple meanings (basic meaning, extended meaning, metaphorical meaning). Decomposed into multiple meanings, multiple meanings are displayed in the dialog box by voice prompts or texts. For the following cases, the solution is determined by splitting.
  • the change of intonation causes the change of meaning terms of homographs to be common features of various languages in human society, such as: four sounds, light sounds, accented sounds, vowels, consonants, vibrato, and unvoiced sounds.
  • Chinese commonly used Chinese characters are less than two thousand.
  • About two-fifths of the polyphonic words have three sounds and four sounds, and each has five sounds.
  • Each change of a tone has a corresponding change. Not only that, but the difference between light and accent is even greater.
  • the automatic language translation system of the present invention can be implemented in two ways: one is voice, the other is text input, and when voice or text enters the language recognition system, it automatically expresses the voice or text.
  • the meaning of the word--the above-mentioned meanings are identified separately.
  • the whole solution can be split in the system (as can be seen in the working principle of the mind recognition system), After the independent meaning, enter the logical thinking sorting detection program. This program is divided into several files according to the concept of the meaning of the word, different file affiliation, parallel relationship with the same position, anyway relationship...
  • Fig. 1 the automatic language translation system of the present invention
  • B, B03, B0303, and B030302 are subordinates, and B01, B02, B03, etc. are parallel relationships, anyway relationship, etc.
  • the number of information codes is small, and the concept meaning is large. On the contrary, the larger the number, the smaller the meaning. Straight line can be divided.
  • Multi-language information automatic exchange system platform (it is translated by language database, information code, logic processing and corresponding code.)
  • a 1 - ⁇ multiple languages are connected to the automatic exchange platform
  • the platform When the information code exchange platform E receives the information code and the language selection instruction F, the platform switches the information code from the database through the logical processor according to the language instruction (see FIG. 3), and switches the meaning term corresponding to the information code.
  • N2006/000805 See Figure 1, Figure 4, after voice output or text display reaches 0
  • Figure 5 is a process of transferring the transfer between various languages from Table 4 Apple Information Code B030302.
  • the specific implementation is as follows: see Figure 1. After opening, click the automatic language translation system to operate the keyboard control keys. The mind recognition network system and the logical processor enter the working state. Click the language selection button. For example, after selecting Chinese confirmation, the Chinese data is turned on.
  • Pop-up dialog box see Table 1, select the input method, confirm the input method by pressing the button, such as confirming the thinking input method, its operation is as follows: Take the word [Apple] as an example, first judge its attribute, select by keyboard operation" B" column, that is, “economic”, see Table 2, pop-up dialog box 1-2 or voice prompt, then select "B03” column, that is, "forestry”, see Table 3, pop-up dialog box 1-3 or voice prompt, Then select the "B0303” column, that is, “fruit wood”, see Table 4, pop-up dialog box 1-4 or voice prompt, then select the "B030302” column, that is, [Apple], confirm and input.
  • the voice input method when the voice input is recognized by the voice recognition system, the monophonic word will automatically enter the selection procedure, and the homophone will appear in the parallel word window, see Table 5, dialog box 1-5 or voice prompt, for example, input " When majiang", it will appear in the dialog 1-5, after a second selection confirmation.
  • the language thinking recognition system sets a multilingual vocabulary storage database and a mind recognition network information code base.
  • the single meaning determined by the selection is converted into an information code in the thinking recognition system, such as "Apple” in the dialog box 1-4, and the information code is "B30302", and the information code is transmitted to the information code exchange platform in the system.
  • the related information code is automatically exchanged with the multi-language or text on the information exchange platform, and finally the desired voice or text is output through the language broadcast system or the display system, as shown in Figures 4 and 5.
  • the invention adopts a voice input method in the application, and is suitable for the purpose of dialogue and communication, especially the simultaneous interpretation of the news channel.
  • the voice of the person is received, and is sorted into a number of meanings by the mind recognition language identification system, the output of the righteous item is selected and transmitted to the information code exchange platform to form an automatic exchange with multiple languages, and vice versa, as long as any language Synchronously mounted on the exchange platform, the purpose of automatic language exchange can be achieved.
  • the language thinking recognition system of the present invention is a two-way section, which can convert voice and text into meaning items and convert meaning items into voice or text.
  • the invention adopts the thinking input method in the application, and the method is applicable to all languages.
  • This method is suitable for editing documents, literary creation, non-manuscript and people with language barriers, especially blind and deaf people.
  • Table 1 dialog box 1-1

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Machine Translation (AREA)

Abstract

L'invention concerne un procédé de traduction automatisée qui permet de traduire une langue en plusieurs langues. Selon ce procédé, un signal de langue est entré dans un système d'identification de réflexion de langue au moyen d'un procédé d'entrée de caractères ou de données vocales; le système d'identification de réflexion de langue trie le signal et le divise en plusieurs sémantèmes; et les sémantèmes sont affichés dans une boîte de dialogue. Un sémantème qui est déterminé sur la base d'une sélection est transmis à une plate-forme d'échange de code d'informations. La plate-forme d'échange de code d'informations assure un échange automatisé avec plusieurs langues ou caractères, et émet la langue ou le caractère désiré au moyen d'un système de diffusion de langue ou d'un système d'affichage.
PCT/CN2006/000805 2006-04-26 2006-04-26 Procédé de traduction automatisée pour la traduction d'une langue en plusieurs langues WO2007121614A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2006/000805 WO2007121614A1 (fr) 2006-04-26 2006-04-26 Procédé de traduction automatisée pour la traduction d'une langue en plusieurs langues

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2006/000805 WO2007121614A1 (fr) 2006-04-26 2006-04-26 Procédé de traduction automatisée pour la traduction d'une langue en plusieurs langues

Publications (1)

Publication Number Publication Date
WO2007121614A1 true WO2007121614A1 (fr) 2007-11-01

Family

ID=38624522

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2006/000805 WO2007121614A1 (fr) 2006-04-26 2006-04-26 Procédé de traduction automatisée pour la traduction d'une langue en plusieurs langues

Country Status (1)

Country Link
WO (1) WO2007121614A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110807311A (zh) * 2018-07-18 2020-02-18 百度在线网络技术（北京）有限公司 用于生成信息的方法和装置

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1417707A (zh) * 2002-12-02 2003-05-14 刘莎 一种自然语言语义信息统一编码方法
CN1441402A (zh) * 2003-04-03 2003-09-10 上海交通大学 跨语种语音信息交流方法
WO2004021148A2 (fr) * 2002-08-30 2004-03-11 Speechgear, Inc. Systeme de traduction
CN1741009A (zh) * 2005-09-13 2006-03-01 徐国庆 一种使用翻译装置的翻译方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2004021148A2 (fr) * 2002-08-30 2004-03-11 Speechgear, Inc. Systeme de traduction
CN1417707A (zh) * 2002-12-02 2003-05-14 刘莎 一种自然语言语义信息统一编码方法
CN1441402A (zh) * 2003-04-03 2003-09-10 上海交通大学 跨语种语音信息交流方法
CN1741009A (zh) * 2005-09-13 2006-03-01 徐国庆 一种使用翻译装置的翻译方法

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110807311A (zh) * 2018-07-18 2020-02-18 百度在线网络技术（北京）有限公司 用于生成信息的方法和装置

Similar Documents

Publication Publication Date Title
JP3277123B2 (ja) 中国語テキストを処理するためのシステムおよび方法
Littell et al. Indigenous language technologies in Canada: Assessment, challenges, and successes
US7395203B2 (en) System and method for disambiguating phonetic input
US6442524B1 (en) Analyzing inflectional morphology in a spoken language translation system
CN101118541B (zh) 汉语语音码汉语语音识别方法
WO2000045377A9 (fr) Methode et appareil permettant d'effectuer une traduction de langage parle
WO2000045376A1 (fr) Methode et appareil permettant une reconnaissance de langage source interactive ainsi qu'une presentation et une selection d'hypotheses de remplacement
WO2000045374A9 (fr) Methode permettant d'effectuer une traduction de langage parle et appareil portable a cet effet
WO2000045290A1 (fr) Procede et appareil pour la construction d'hypotheses de reconnaissance vocale adaptative et selection dans un systeme de traduction de langage parle
CN102902660A (zh) 汉语语音码全拼和简拼汉语全息信息处理方法
Ablimit et al. A multilingual language processing tool for Uyghur, Kazak and Kirghiz
CN101430681A (zh) 内嵌翻译功能的输入方法
Jiang et al. Braille to print translations for Chinese
CN103164396B (zh) 使用汉维哈柯电子辞典自动转译汉维哈柯语的方法
KR102086601B1 (ko) 담화 성분과 화행을 고려한 한국어 대화체 코퍼스 분류 방법 및 시스템
KR100400220B1 (ko) 대화 모델을 이용한 자동 통역 장치 및 방법
CN111581991A (zh) 一种基于端到端神经机器翻译的汉盲翻译方法及系统
CN103164395A (zh) 汉柯电子辞典及其自动转译汉柯语的方法
WO2007121614A1 (fr) Procédé de traduction automatisée pour la traduction d'une langue en plusieurs langues
CN103853705A (zh) 计算机汉语语音与外语语音实时语音字幕翻译方法
WO2005089215A2 (fr) Procedes de frappe et de saisie phonetique de caracteres chinois et expressions chinoises
Rahate et al. An experimental technique on text normalization and its role in speech synthesis
Garvin et al. The conversion of phonetic into orthographic English: a machine-translation approach to the problem
Dawa et al. Multilingual Text–Speech Corpus of Mongolian
KR100463376B1 (ko) 원시언어를 대상언어로 번역하기 위한 번역엔진 장치 및 그 번역방법

Legal Events

Date Code Title Description
DPE2 Request for preliminary examination filed before expiration of 19th month from priority date (pct application filed from 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 06741731

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112 (1) EPC (EPOFORM 1205A DATED 18.02.2009)

122 Ep: pct application non-entry in european phase

Ref document number: 06741731

Country of ref document: EP

Kind code of ref document: A1