JPH11259474A - Machine translation system and machine translating method - Google Patents
Machine translation system and machine translating methodInfo
- Publication number
- JPH11259474A JPH11259474A JP10057799A JP5779998A JPH11259474A JP H11259474 A JPH11259474 A JP H11259474A JP 10057799 A JP10057799 A JP 10057799A JP 5779998 A JP5779998 A JP 5779998A JP H11259474 A JPH11259474 A JP H11259474A
- Authority
- JP
- Japan
- Prior art keywords
- phrase
- occurrence information
- preposition
- occurrence
- unit
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Landscapes
- Machine Translation (AREA)
- Document Processing Apparatus (AREA)
Abstract
Description
【0001】[0001]
【発明の属する技術分野】本発明は、原言語の文を目標
言語の文に翻訳する機械翻訳装置及び機械翻訳方法に関
するものである。The present invention relates to a machine translation device and a machine translation method for translating a source language sentence into a target language sentence.
【0002】[0002]
【従来の技術】従来の機械翻訳装置では、ある原言語の
原文を原言語の文法に従って構文解析を行い翻訳し目標
言語の翻訳文を得る装置であった。2. Description of the Related Art In a conventional machine translation apparatus, a source sentence of a certain source language is analyzed by parsing according to the grammar of the source language and translated to obtain a target language translation.
【0003】[0003]
【発明が解決しようとする課題】しかしながら、自然言
語は、構文解析において同じ句や文に複数の構文木が与
えられる場合、その句や文は文法的に曖昧な句や文にな
るという文法的曖昧性を内在していることが知られてい
る。即ち、ある修飾語句に対し、文法的に修飾すること
が可能な複数の被修飾句が存在する場合がある。例え
ば、英文「I saw agirl with a telescope.」の場合、
修飾語句である前置詞句「with a telescope」の修飾先
として、「a girl」と「saw」との二通りの可能性が考
えられる。前置詞句の用法として、該前置詞句が名詞句
(前記例では「a girl」)を修飾する用法を形容詞的用
法と呼び、該前置詞句が動詞句、形容詞句、副詞句(前
記例では「saw」)を修飾する用法を副詞的用法と呼
ぶ。前置詞句が形容詞的用法をとるか副詞的用法をとる
かは、それぞれの場合に依存し、決まった法則はない。However, in the natural language, when a plurality of parse trees are given to the same phrase or sentence in parsing, the phrase or sentence becomes a grammatically ambiguous phrase or sentence. It is known that ambiguity is inherent. That is, there may be a plurality of modified phrases that can be grammatically modified for a given modifier. For example, in the case of the English sentence "I saw agirl with a telescope."
There are two possibilities for modifying the preposition phrase “with a telescope”, which is a modifier, “a girl” and “saw”. As a usage of the preposition phrase, a usage that modifies a noun phrase (“a girl” in the above example) is called an adjective usage, and the preposition phrase is a verb phrase, an adjective phrase, or an adverbial phrase (in the above example, “saw”). )) Is called an adverbial usage. Whether a prepositional phrase takes an adjective or adverbial usage depends on each case, and there is no set rule.
【0004】従来の機械翻訳装置では、上記自然言語の
文法的曖昧性による前置詞句に対する被修飾句の曖昧性
を除去することができず、正確な翻訳結果が得られない
という問題を有していた。[0004] The conventional machine translator cannot remove the ambiguity of the qualified phrase with respect to the prepositional phrase due to the grammatical ambiguity of the natural language, and has a problem that an accurate translation result cannot be obtained. Was.
【0005】本発明は上記従来の課題を解決するもの
で、原言語の一般の文章における大量の解析済文を集め
たコーパスから前置詞句の修飾先に関する統計情報(前
置詞句の前置詞、前置詞句の主名詞、副詞的用法の前置
詞句が従属する被修飾句、形容詞的用法の前置詞句が従
属する被修飾句、の組合せに対して決まる前置詞句の用
法に関する情報。「前置詞句共起情報」と呼ぶ)を取り
出し、予め前置詞句共起情報のリスト(共通共起情報リ
スト)を作成し、更に利用者が独自の前置詞句共起情報
を登録可能な前置詞句共起情報のリスト(ユーザ共起情
報リスト)を備え、翻訳時に共通共起情報リスト及びユ
ーザ共起情報リストを参照することで、前置詞句の用法
の統計データに基づく特定及び利用者の指定による前置
詞句用法の特定を行うことによって、前置詞句の被修飾
句の曖昧性を排除し、翻訳精度の高い機械翻訳装置及び
機械翻訳方法を提供することを目的とする。SUMMARY OF THE INVENTION The present invention solves the above-mentioned conventional problems. The present invention solves the above problem by collecting statistical information (prepositions of prepositional phrases, prepositional phrases of prepositional phrases, etc.) from a corpus collecting a large amount of analyzed sentences in general sentences of the source language. Information on the usage of prepositional phrases, which is determined by the combination of the main noun, the qualified phrase to which the prepositional phrase of the adverbial usage depends, and the qualified phrase to which the prepositional phrase of the adjective usage depends. Call), a list of prepositional phrase co-occurrence information (common co-occurrence information list) is created in advance, and a list of prepositional phrase co-occurrence information (user co-occurrence) that allows the user to register unique prepositional phrase co-occurrence information Information list), and by referring to the common co-occurrence information list and the user co-occurrence information list during translation, identification based on statistical data on usage of preposition phrases and identification of preposition phrase usage by user designation are performed. It allows to eliminate the ambiguity of the qualifiers in the prepositional phrase, and an object thereof is to provide a highly translation accuracy machine translation apparatus and machine translation method.
【0006】[0006]
【課題を解決するための手段】上記課題を解決するため
に本発明の機械翻訳装置は、原言語の文を目標言語の文
に翻訳する機械翻訳装置であって、原言語の前置詞句の
係り先に関する前置詞句共起情報を記憶する共通共起情
報リストと、利用者により登録された原言語の前置詞句
の係り先に関する前置詞句共起情報を記憶する追加登録
可能なユーザ共起情報リストと、ユーザ共起情報リスト
に前置詞句共起情報を登録する共起情報登録部と、原言
語の原文の前置詞句の係り先に関する前置詞句共起語を
抽出する共起語抽出部と、前置詞句共起語に対応する前
置詞句共起情報を共起情報リストから検索する共起情報
検索部と、前置詞句共起情報により前置詞句の係る被修
飾句を決定し原文を翻訳する翻訳部とを備えた構成より
成る。According to the present invention, there is provided a machine translation apparatus for translating a sentence in a source language into a sentence in a target language. A common co-occurrence information list for storing prepositional phrase co-occurrence information related to the destination, and an additional registrable user co-occurrence information list for storing prepositional phrase co-occurrence information for the destination of the prepositional phrase in the source language registered by the user. A co-occurrence information registration unit that registers prepositional phrase co-occurrence information in a user co-occurrence information list, a co-occurrence word extraction unit that extracts a prepositional phrase co-occurrence word related to a destination of a prepositional phrase in the source language of the source language, A co-occurrence information search unit that searches for a preposition phrase co-occurrence information corresponding to a co-occurrence word from a co-occurrence information list, and a translation unit that determines a qualified phrase of the preposition phrase based on the preposition phrase co-occurrence information and translates the original sentence It has a configuration provided.
【0007】この構成により前置詞句の被修飾句の曖昧
性を排除し、翻訳精度の高い機械翻訳装置を提供するこ
とができる。With this configuration, it is possible to eliminate the ambiguity of the modified phrase of the prepositional phrase and provide a machine translation device with high translation accuracy.
【0008】なお、「前置詞句共起語」とは、ある前置
詞句に対する、該前置詞句の前置詞、該前置詞句の主名
詞、副詞的用法の該前置詞句が従属する被修飾語候補、
形容詞的用法の該前置詞句が従属する被修飾語候補、の
4つの語のことを言う。また、「前置詞句共起語に対応
する前置詞句共起情報」とは、該前置詞句共起語と該前
置詞句共起情報との前置詞句の前置詞が等しく、前置詞
句の主名詞、形容詞的用法の前置詞句が従属する被修飾
語候補、副詞的用法の前置詞句が従属する被修飾語候
補、のそれぞれの非活用形が等しいことを意味する。[0008] The "preposition phrase co-occurrence word" is a preposition of a preposition phrase, a main noun of the preposition phrase, a candidate of a qualified word to which the preposition phrase of the adverbial usage depends on a certain preposition phrase,
It refers to the four qualified words that are the candidates of the qualified word to which the preposition phrase of adjective usage depends. Further, `` preposition phrase co-occurrence information corresponding to a preposition phrase co-occurrence word '' means that the prepositions of the preposition phrases of the preposition phrase co-occurrence word and the preposition phrase co-occurrence information are equal, the main noun of the preposition phrase, and the adjective. This means that the unmodified forms of the qualified word candidate to which the preposition phrase of the usage depends and the qualified word candidate of the qualified word to which the preposition phrase of the adverbial usage depends are equal.
【0009】上記課題を解決するために本発明の機械翻
訳方法は、原言語の文を目標言語の文に翻訳する機械翻
訳方法であって、原言語の原文の前置詞句の係り先に関
する前置詞句共起語を抽出する共起語抽出過程と、原言
語の前置詞句の係り先に関する前置詞句共起情報を記憶
された予め用意された共通共起情報リストと利用者によ
り登録された原言語の前置詞句の係り先に関する前置詞
句共起情報が記憶された追加登録可能なユーザ共起情報
リストとから前記前置詞句共起語に対応する前置詞句共
起情報を検索する共起情報検索過程と、前置詞句共起情
報により前置詞句の係る被修飾句を決定し原文を翻訳す
る翻訳過程とを備えた構成より成る。In order to solve the above problem, a machine translation method according to the present invention is a machine translation method for translating a sentence in a source language into a sentence in a target language, wherein the preposition phrase relates to a destination of a preposition phrase in the source language. A co-occurrence word extraction process for extracting co-occurrence words, a pre-prepared common co-occurrence information list storing preposition phrase co-occurrence information relating to the destination of preposition phrases in the source language, and a source language registered by the user. A co-occurrence information search step of searching for a preposition phrase co-occurrence information corresponding to the preposition phrase co-occurrence information from an additionally registerable user co-occurrence information list in which preposition phrase co-occurrence information regarding a prepositional phrase is related, And a translation process of translating the original sentence by determining the qualified phrase of the preposition phrase based on the preposition phrase co-occurrence information.
【0010】この構成により前置詞句の被修飾句の曖昧
性を排除し、翻訳精度の高い機械翻訳方法を提供するこ
とができる。With this configuration, it is possible to eliminate the ambiguity of the modified phrase of the prepositional phrase and provide a machine translation method with high translation accuracy.
【0011】[0011]
【発明の実施の形態】本発明の請求項1に記載の機械翻
訳装置は、原言語の文を目標言語の文に翻訳する機械翻
訳装置であって、原言語の前置詞句の係り先に関する前
置詞句共起情報を記憶する共通共起情報リストと、利用
者により登録された原言語の前置詞句の係り先に関する
前置詞句共起情報を記憶する追加登録可能なユーザ共起
情報リストと、ユーザ共起情報リストに前置詞句共起情
報を登録する共起情報登録部と、原言語の原文の前置詞
句の係り先に関する前置詞句共起語を抽出する共起語抽
出部と、前置詞句共起語に対応する前置詞句共起情報を
共起情報リストから検索する共起情報検索部と、前置詞
句共起情報により前置詞句の係る被修飾句を決定し原文
を翻訳する翻訳部とを備えたこととしたものであり、こ
の構成により、前置詞句の修飾先の曖昧性が排除され翻
訳精度が向上するという作用を有する。DESCRIPTION OF THE PREFERRED EMBODIMENTS A machine translation apparatus according to claim 1 of the present invention is a machine translation apparatus for translating a sentence in a source language into a sentence in a target language. A co-occurrence information list that stores phrase co-occurrence information; a user co-occurrence information list that additionally stores prepositional phrase co-occurrence information regarding a destination of a prepositional phrase in a source language registered by a user; A co-occurrence information registration unit that registers prepositional phrase co-occurrence information in the occurrence information list, a co-occurrence word extraction unit that extracts prepositional phrase co-occurrence words related to the destination of the prepositional phrase in the source language source text, and a prepositional phrase co-occurrence word A co-occurrence information search unit that searches for a preposition phrase co-occurrence information corresponding to a preposition phrase from a co-occurrence information list, and a translation unit that determines the qualified phrase of the preposition phrase based on the preposition phrase co-occurrence information and translates the original sentence With this configuration, Lyrics phrase modified destination ambiguity is eliminated translation accuracy has the effect of improving.
【0012】本発明の請求項2に記載の機械翻訳装置
は、請求項1に記載の機械翻訳装置において、原言語の
用法情報付加文から前置詞句共起情報を作成する共起情
報作成部を備えたこととしたものであり、この構成によ
り、前置詞句共起情報のユーザ共起情報リストへの登録
時に、前置詞句共起語の原文からの抽出と、抽出された
それぞれの前置詞句共起語の活用形又は複数形からの単
数原形への変換と、前置詞句共起情報の作成とを自動化
することが可能となり、前置詞句共起情報登録時におけ
る利用者の操作性が向上するという作用を有する。A machine translation apparatus according to a second aspect of the present invention is the machine translation apparatus according to the first aspect, further comprising a co-occurrence information creating unit for creating preposition phrase co-occurrence information from the usage information additional sentence of the source language. With this configuration, when the prepositional phrase co-occurrence information is registered in the user co-occurrence information list, the extraction of the prepositional phrase co-occurrence word from the original text and the extraction of each extracted prepositional phrase co-occurrence It is possible to automate the conversion of the inflected form or plural form to the singular form and to create the preposition phrase co-occurrence information, thereby improving the operability of the user when registering the preposition phrase co-occurrence information. Having.
【0013】本発明の請求項3に記載の機械翻訳装置
は、請求項1に記載の機械翻訳装置において、情報を表
示する表示面を有する表示部と、表示面上の特定の位置
を指示する指示部と、前置詞句共起情報を一時的に記憶
する共起情報一時記憶部と、利用者の前置詞句用法の変
更指示により前記前置詞句の用法を変更する用法変更部
とを備えたこととしたものであり、この構成により、前
置詞句共起情報のユーザ共起情報リストへの登録時に、
ユーザ共起情報リストへの前置詞句共起情報の登録操作
が簡便化され、前置詞句共起情報登録時における利用者
の操作性が向上するという作用を有する。According to a third aspect of the present invention, there is provided a machine translation apparatus according to the first aspect, wherein a display unit having a display surface for displaying information and a specific position on the display surface are indicated. An instruction unit, a co-occurrence information temporary storage unit that temporarily stores preposition phrase co-occurrence information, and a usage change unit that changes the usage of the preposition phrase according to a user's instruction to change the preposition phrase usage. With this configuration, when the prepositional phrase co-occurrence information is registered in the user co-occurrence information list,
The operation of registering the prepositional phrase co-occurrence information in the user co-occurrence information list is simplified, and the operability of the user at the time of registering the prepositional phrase co-occurrence information is improved.
【0014】本発明の請求項4に記載の機械翻訳方法
は、原言語の文を目標言語の文に翻訳する機械翻訳方法
であって、原言語の原文の前置詞句の係り先に関する前
置詞句共起語を抽出する共起語抽出過程と、原言語の前
置詞句の係り先に関する前置詞句共起情報を記憶された
予め用意された共通共起情報リストと利用者により登録
された原言語の前置詞句の係り先に関する前置詞句共起
情報が記憶された追加登録可能なユーザ共起情報リスト
とから前置詞句共起語に対応する前置詞句共起情報を検
索する共起情報検索過程と、前置詞句共起情報により前
置詞句の係る被修飾句を決定し原文を翻訳する翻訳過程
と、を備えたこととしたものであり、この構成により、
前置詞句の修飾先の曖昧性が排除され翻訳精度が向上す
るという作用を有する。A machine translation method according to a fourth aspect of the present invention is a machine translation method for translating a source language sentence into a target language sentence, wherein a preposition phrase relating to a destination of a preposition phrase in the source language original sentence is shared. A co-occurrence word extraction process for extracting a word, a pre-prepared common co-occurrence information list in which preposition phrase co-occurrence information relating to the destination of the preposition phrase in the source language is stored, and a source language preposition registered by the user. A co-occurrence information search process for searching for a preposition / phrase co-occurrence information corresponding to a preposition / phrase co-occurrence word from an additionally registerable user co-occurrence information list in which preposition / phrase co-occurrence information relating to a phrase destination is stored; And a translation process of translating the original text by determining the qualified phrase of the prepositional phrase based on the co-occurrence information.
This has the effect of eliminating the ambiguity of the modification destination of the prepositional phrase and improving the translation accuracy.
【0015】(実施の形態1)図1は、本発明の実施の
形態1における機械翻訳装置の機能ブロック図である。(Embodiment 1) FIG. 1 is a functional block diagram of a machine translation apparatus according to Embodiment 1 of the present invention.
【0016】図1において、100は原言語の原文を入
力する機能を有する入力部、102は原文の前置詞句共
起語を抽出する機能を有する共起語抽出部、103は共
通共起情報リスト103a及びユーザ共起リスト103
bを記憶する機能を有する共起情報記憶部、104は前
置詞句共起語に対応する前置詞句共起情報を共通共起情
報リスト103a及びユーザ共起情報リスト103bか
ら検索する機能を有する共起情報検索部、105はユー
ザ共起情報リストに共起情報を登録する機能を有する共
起情報登録部、106は前置詞句共起情報により前置詞
句の被修飾句を決定し原言語の原文から目標言語の翻訳
文を生成する機能を有する翻訳部、107は翻訳文を出
力する機能を有する出力部、108は、入力部100、
共起語抽出部102、共起情報検索部104、共起情報
登録部105、翻訳部106、出力部108、の間のデ
ータの流れを後述の動作の流れに従って制御する機能を
有する制御部である。In FIG. 1, reference numeral 100 denotes an input unit having a function of inputting an original sentence of a source language; 102, a co-occurrence word extraction unit having a function of extracting a prepositional phrase co-occurrence word of the original sentence; 103a and user co-occurrence list 103
The co-occurrence information storage unit 104 having a function of storing the b. The co-occurrence information storage unit 104 has a function of searching for the preposition phrase co-occurrence information corresponding to the preposition phrase co-occurrence word from the common co-occurrence information list 103a and the user co-occurrence information list 103b. An information retrieval unit, 105 is a co-occurrence information registration unit having a function of registering co-occurrence information in a user co-occurrence information list, and 106 determines a qualifying phrase of a preposition phrase based on preposition phrase co-occurrence information, and sets a target from a source language original. A translation unit having a function of generating a translation of a language; 107, an output unit having a function of outputting a translation; 108, an input unit 100;
A control unit having a function of controlling the flow of data among the co-occurrence word extraction unit 102, the co-occurrence information search unit 104, the co-occurrence information registration unit 105, the translation unit 106, and the output unit 108 in accordance with the operation flow described below. is there.
【0017】図2は、本発明の実施の形態1における共
通共起情報リストの内部状態を示す図である。共通共起
情報リストは、副詞的用法の前置詞句が従属する被修飾
語候補、形容詞的用法の前置詞句が従属する被修飾語候
補、前置詞句の前置詞、前置詞句の主名詞、前置詞句の
用法、の4つの語及び用法で構成される要素を集成した
リストである。FIG. 2 is a diagram showing an internal state of the common co-occurrence information list according to the first embodiment of the present invention. The common co-occurrence information list is a list of qualified word candidates to which the prepositional phrase of adverbial usage depends, qualified word candidates to which the prepositional phrase of adjective usage depends, prepositions of prepositional phrases, main nouns of prepositional phrases, and usage of prepositional phrases , Is a list of elements composed of four words and usages.
【0018】例えば、英文「I saw a girl with a tele
scope.」の中の前置詞句「with a telescope」と該副詞
的用法の前置詞句に係る被修飾句「saw」及び形容詞的
用法に係る被修飾句「a girl」の場合、共通共起情報リ
ストには該前置詞句及び該被修飾句に対応する前置詞句
共起情報として「see,girl,with,telescope,副詞的
用法」のような4個の語及び用法から成る要素が登録さ
れている。For example, the English sentence "I saw a girl with a tele
In the case of the preposition phrase "with a telescope" in "scope." and the qualified phrase "saw" related to the preposition phrase of the adverbial usage and the qualified phrase "a girl" related to the adjective usage, the common co-occurrence information list In the field, elements consisting of four words and usage such as "see, girl, with, telescope, adverb usage" are registered as preposition phrase co-occurrence information corresponding to the preposition phrase and the subject phrase.
【0019】図3は、本発明の実施の形態1における機
械翻訳装置の装置ブロック図である。FIG. 3 is a device block diagram of the machine translation device according to the first embodiment of the present invention.
【0020】図3において、301は入力装置で、例え
ば、キーボード、マウス・タッチパネル・トラックボー
ル等のポインティングデバイス、ディスク読みとり装
置、光学式読みとり装置、音声入力装置等の1乃至複数
の組み合わせにより構成される。302は中央演算処理
装置(CPU)、303は記憶装置で、ROM(リード
オンリーメモリー)若しくはRAM(ランダムアクセス
メモリー)、又はその組み合わせ、或いはディスク装置
により構成される。304は出力装置で、ディスプレ
イ、プリンタ、ディスク書き込み装置、光学式書き込み
装置、音声出力装置等の1乃至複数の組み合わせで構成
される。305は読取装置でフロッピーディスクドライ
ブ、CD−ROMドライブ、磁気テープリーダー等の1
乃至複数の組み合わせで構成される。306は読取装置
305で読み取り可能な記録媒体で、フロッピーディス
ク、CD−ROM、磁気テープ等で構成される。In FIG. 3, reference numeral 301 denotes an input device, which is constituted by one or more combinations of a keyboard, a pointing device such as a mouse, a touch panel and a trackball, a disk reading device, an optical reading device, and a voice input device. You. Reference numeral 302 denotes a central processing unit (CPU), and reference numeral 303 denotes a storage device, which is constituted by a ROM (read only memory) or a RAM (random access memory), a combination thereof, or a disk device. An output device 304 includes one or more combinations of a display, a printer, a disk writing device, an optical writing device, an audio output device, and the like. Reference numeral 305 denotes a reading device, such as a floppy disk drive, CD-ROM drive, or magnetic tape reader.
Or a plurality of combinations. Reference numeral 306 denotes a recording medium that can be read by the reading device 305, and includes a floppy disk, a CD-ROM, a magnetic tape, and the like.
【0021】この装置において、入力部100は入力装
置301により実現され、共起情報記憶部103は記憶
装置303により実現され、共起語抽出部102・共起
情報検索部104・共起情報登録部105・翻訳部10
6・制御部108はCPU302により実現され、出力
部107は出力装置304により実現される。In this apparatus, the input unit 100 is realized by the input device 301, the co-occurrence information storage unit 103 is realized by the storage device 303, and the co-occurrence word extraction unit 102 / co-occurrence information search unit 104 / co-occurrence information registration Unit 105 / Translation unit 10
6. The control unit 108 is realized by the CPU 302, and the output unit 107 is realized by the output device 304.
【0022】なお、本実施の形態では、CPU302
は、予めROMなどの記憶装置303に記憶された制御
プログラムを実行することにより各種制御を行うことと
しているが、制御プログラムは、コンピュータ読み取り
可能な記録媒体(例えばCD−ROMなど)に記憶され
ていても良く、このような形態では、CPU302は、
読取装置305(例えばCD−ROMドライブなど)を
介して記録媒体から制御プログラムを読み込んで、読み
込んだ制御プログラムをRAMやハードディスクなどの
記憶装置303に展開した後、制御プログラムを実行す
る。このような形態とすることにより、本発明を汎用コ
ンピュータで容易に実現することが可能となる。In this embodiment, the CPU 302
Performs various controls by executing a control program stored in a storage device 303 such as a ROM in advance. The control program is stored in a computer-readable recording medium (for example, a CD-ROM). In such a form, the CPU 302
The control program is read from a recording medium via a reading device 305 (for example, a CD-ROM drive), and the read control program is loaded into a storage device 303 such as a RAM or a hard disk, and then the control program is executed. With such an embodiment, the present invention can be easily realized by a general-purpose computer.
【0023】次に、本実施の形態に係る機械翻訳装置の
動作の流れについて説明する。翻訳を行う場合、利用者
は、入力部100に翻訳機能選択指示2aを入力し、続
いて翻訳したい原文1を入力する。制御部108は、入
力部100から翻訳機能選択指示2aを受けると翻訳動
作状態となり、次いで原文1を受けると、原文1を共起
語抽出部102に送り、原文1から前置詞句共起語3を
抽出する。次に、制御部108は、前置詞句共起語3を
共起情報検索部104に送り、共通共起情報リスト10
3a及びユーザ共起情報リスト103bから前置詞句共
起語3に対応する前置詞句共起情報3aを検索抽出す
る。次に、制御部108は、前置詞句共起情報3aと原
文1とを翻訳部106に送り、目的言語の翻訳文4を得
る。Next, an operation flow of the machine translation apparatus according to the present embodiment will be described. When performing translation, the user inputs the translation function selection instruction 2a into the input unit 100, and then inputs the original text 1 to be translated. When receiving the translation function selection instruction 2a from the input unit 100, the control unit 108 enters the translation operation state. When receiving the original sentence 1, the control unit 108 sends the original sentence 1 to the co-occurrence word extraction unit 102, and the preposition phrase co-occurrence word 3 Is extracted. Next, the control unit 108 sends the preposition phrase co-occurrence word 3 to the co-occurrence information search unit 104, and
The preposition phrase co-occurrence information 3a corresponding to the preposition phrase co-occurrence word 3 is retrieved and extracted from 3a and the user co-occurrence information list 103b. Next, the control unit 108 sends the preposition phrase co-occurrence information 3a and the original sentence 1 to the translating unit 106, and obtains the translated sentence 4 of the target language.
【0024】最後に、制御部108は翻訳文4を出力装
置107に送り、翻訳文4を出力する。Finally, the control unit 108 sends the translated sentence 4 to the output device 107, and outputs the translated sentence 4.
【0025】なお、入力部100に原文1が入力された
時点で、原文1中に前置詞句が含まれていない場合に
は、制御部108は原文1を翻訳部106に送り翻訳文
4を得た後、翻訳文4を出力部107に送り出力する。If the preamble phrase is not included in the original sentence 1 when the original sentence 1 is input to the input unit 100, the control unit 108 sends the original sentence 1 to the translating unit 106 to obtain the translated sentence 4. After that, the translation sentence 4 is sent to the output unit 107 and output.
【0026】例えば、原文1として英文「I watched a
chick with a telescope.」が入力された場合、制御部
108は原文1から前置詞句共起語3「watch,chick,
with,telescope」を抽出する。この時点では、前置詞
句は副詞的用法(「with a telescope」が「watched」
に係る)か形容詞的用法(「with a telescope」が「ac
hick」に係る)かの判断はされていない。次に制御部1
08は、共起情報記憶部に記憶されている共通共起情報
リスト103a内及びユーザ共起情報リスト103b内
から、前置詞句共起語に対応する前置詞句共起情報を検
索し、前置詞句共起情報3a「watch,chick,with,te
lescope,副詞的用法」を得る。共通共起情報リスト
は、図2に示すように、副詞的用法の前置詞句が従属す
る被修飾語候補、形容詞的用法の前置詞句が従属する被
修飾語候補、前置詞句の前置詞、前置詞句の主名詞、前
置詞句の用法、の4つの語及び用法で構成される要素の
集合から成る。従って、前置詞句共起情報が検索抽出さ
れた時点で該前置詞句の用法が形容詞的用法若しくは副
詞的用法に確定するため、前置詞句の係る被修飾句の曖
昧性は排除され被修飾句が正しく決定される。次に、制
御部108は前置詞句共起情報3aと原文1を翻訳部1
06に送り、翻訳部106は前置詞句共起情報3aの前
置詞句用法の情報を基に原文を翻訳し、翻訳文4「私は
望遠鏡で雛鳥を見た。」を得る。共起情報3aの情報が
ない場合、翻訳文4は「私は望遠鏡を持っている雛鳥を
見た。」と「私は望遠鏡で雛鳥を見た。」との2通りの
可能性を持つが、前置詞句共起情報3aを参照すること
により翻訳文4は一意的に決定され、正しい翻訳文を得
ることが可能となる。For example, as an original sentence 1, the English sentence "I watched a
When “chick with a telescope.” is input, the control unit 108 starts from the original sentence 1 and the preposition phrase co-occurrence word 3 “watch, chick,
with, telescope ”is extracted. At this point, the prepositional phrase is adverbial ("with a telescope" becomes "watched"
) Or adjective usage ("with a telescope" becomes "ac
hick ") has not been determined. Next, the control unit 1
08 searches for preposition phrase co-occurrence information corresponding to the preposition phrase co-occurrence word from the common co-occurrence information list 103a and the user co-occurrence information list 103b stored in the co-occurrence information storage unit. Ki 3a "watch, chick, with, te
lescope, adverbial usage. " As shown in FIG. 2, the common co-occurrence information list includes, as shown in FIG. 2, a qualified word candidate to which the preposition phrase of the adverbial usage depends, a qualified word candidate to which the preposition phrase of the adjective usage depends, a preposition of the prepositional phrase, It consists of a set of elements consisting of four words and usages: main nouns, prepositional phrase usage. Therefore, when the preposition phrase co-occurrence information is retrieved and extracted, the usage of the preposition phrase is determined to be an adjective usage or an adverb usage, so that the ambiguity of the qualified phrase related to the preposition phrase is eliminated and the qualified phrase is correct. It is determined. Next, the control unit 108 converts the preposition phrase co-occurrence information 3a and the original sentence 1 into the translation unit 1
Then, the translation unit 106 translates the original sentence based on the information of the prepositional phrase usage of the prepositional phrase co-occurrence information 3a, and obtains a translated sentence 4 "I saw the chick with a telescope." If there is no information of the co-occurrence information 3a, the translation 4 has two possibilities, "I saw a chick with a telescope." And "I saw a chick with a telescope." The translated sentence 4 is uniquely determined by referring to the preposition phrase co-occurrence information 3a, and a correct translated sentence can be obtained.
【0027】ユーザ共起リストに共起情報の登録を行う
場合、利用者は、まず入力部100に登録機能選択指示
2bを入力し、続いて登録したい前置詞句共起情報5を
入力する。制御部108は、入力部100から登録機能
選択指示2bを受けると登録動作状態となり、次いで前
置詞句共起情報5を受けると、前置詞句共起情報5を共
起情報登録部105に送り、前置詞句共起情報5をユー
ザ共起情報リスト103bに登録する。When registering co-occurrence information in the user co-occurrence list, the user first inputs a registration function selection instruction 2b to the input unit 100, and then inputs preposition phrase co-occurrence information 5 to be registered. When receiving the registration function selection instruction 2b from the input unit 100, the control unit 108 enters a registration operation state. When receiving the preposition phrase co-occurrence information 5, the control unit 108 sends the preposition phrase co-occurrence information 5 to the co-occurrence information registration unit 105, and The phrase co-occurrence information 5 is registered in the user co-occurrence information list 103b.
【0028】例えば、英文「I saw chicks in the nes
t.」に関する前置詞句共起情報が登録されておらず正し
く翻訳されなかった場合、利用者はまず入力部100に
登録機能選択指示2bを入力し、続いて前置詞句共起情
報5「see,chick,in,nest,形容詞的用法」を入力す
ることによって、ユーザ共起情報リスト103bに前置
詞句共起情報を登録することができる。このとき、利用
者に対する利便性を向上させるため、例えば、入力部に
キーボード及びマウス、出力部にディスプレイを備え、
ディスプレイに図4に示したような前置詞共起情報入力
ウィンドウを表示させ、キーボード及びマウスにより入
力するような構成としてもよい。このように共起情報リ
ストに不足している前置詞句共起情報をユーザ共起情報
リスト103bに登録し補充することにより、例えば、
次に利用者が英文「He saw a black chick in the big
nest.」を翻訳しようとすると、共起情報検索部104
はユーザ共起情報リスト103b中に前回補充した前置
詞句共起情報5「see,chick,in,nest,形容詞的用
法」を発見し正しい翻訳文「彼はその大きな巣の中にい
る黒い雛鳥を見た。」を得ることが可能となる。For example, the English sentence "I saw chicks in the nes"
If the preposition phrase co-occurrence information relating to “t.” is not registered and is not correctly translated, the user first inputs the registration function selection instruction 2b to the input unit 100, and then the preposition phrase co-occurrence information 5 “see, By inputting "chick, in, nest, adjective usage", preposition phrase co-occurrence information can be registered in the user co-occurrence information list 103b. At this time, in order to improve the convenience for the user, for example, a keyboard and a mouse in the input unit, a display in the output unit,
A preposition co-occurrence information input window as shown in FIG. 4 may be displayed on the display, and the input may be made by a keyboard and a mouse. By registering and supplementing the preposition phrase co-occurrence information missing in the co-occurrence information list in the user co-occurrence information list 103b, for example,
Next, the user reads the English sentence "He saw a black chick in the big
When trying to translate “nest.”, the co-occurrence information search unit 104
Found the preposition phrase co-occurrence information 5 "see, chick, in, nest, adjective usage" supplemented last time in the user co-occurrence information list 103b, and translated the correct translation " I saw. "
【0029】以上のように、本実施の形態によれば、前
置詞句共起情報を各共起情報リストから検索することに
より前置詞句の被修飾句の曖昧性を排除され翻訳精度が
向上し、また、追記可能なユーザ共起情報リスト103
bを備えることにより、利用者が共起情報リストに不足
している前置詞句共起情報を補充することが可能とな
り、補充を行うことによって翻訳精度が更に向上し、利
用者毎の特有な表現に対応して翻訳精度が向上する。な
お、本実施の形態では原言語を英語、目標言語を日本語
として例示したが、原言語及び目標言語はこれらに限定
されるものではない。As described above, according to the present embodiment, by searching for the preposition phrase co-occurrence information from each co-occurrence information list, the ambiguity of the qualified phrase of the preposition phrase is eliminated, and the translation accuracy is improved. Also, a user co-occurrence information list 103 that can be additionally written
By providing b, the user can supplement the preposition phrase co-occurrence information that is lacking in the co-occurrence information list, and by performing the supplement, the translation accuracy is further improved, and a unique expression for each user is provided. The translation accuracy is improved in accordance with. In the present embodiment, the source language is exemplified as English and the target language is exemplified as Japanese, but the source language and the target language are not limited to these.
【0030】(実施の形態2)図5は本発明の実施の形
態2における機械翻訳装置の機能ブロック図である。図
5において、実施の形態1で説明した各部については同
符号を付して説明は省略する。本実施の形態は、実施の
形態1の機械翻訳装置において、前置詞句の前置詞に前
置詞句用法に関する付加情報を付けた原言語の文(用法
情報付加文)から前置詞句共起情報を作成する機能を有
する共起情報作成部401を備えることにより、前置詞
句共起情報登録操作における利用者の操作性の向上を図
ったものである。(Embodiment 2) FIG. 5 is a functional block diagram of a machine translation apparatus according to Embodiment 2 of the present invention. In FIG. 5, the components described in the first embodiment are denoted by the same reference numerals, and description thereof is omitted. According to the present embodiment, in the machine translation apparatus according to the first embodiment, a function of generating preposition phrase co-occurrence information from a source language sentence (usage information additional sentence) in which a preposition of a preposition phrase is attached with additional information regarding preposition phrase usage. By providing the co-occurrence information creating unit 401 having the above, the operability of the user in the preposition phrase co-occurrence information registration operation is improved.
【0031】本実施の形態に係る機械翻訳装置の装置ブ
ロック図は、図3に示した形態となる。この装置におい
て、入力部100は入力装置301により実現され、共
起情報記憶部103は記憶装置303により実現され、
共起語抽出部102・共起情報検索部104・共起情報
登録部105・翻訳部106・共起情報作成部401・
制御部108はCPU302により実現され、出力部1
07は出力装置304により実現される。An apparatus block diagram of the machine translation apparatus according to the present embodiment has the form shown in FIG. In this device, the input unit 100 is realized by the input device 301, the co-occurrence information storage unit 103 is realized by the storage device 303,
Co-occurrence word extraction unit 102 / co-occurrence information search unit 104 / co-occurrence information registration unit 105 / translation unit 106 / co-occurrence information creation unit 401
The control unit 108 is realized by the CPU 302 and the output unit 1
07 is realized by the output device 304.
【0032】なお、本実施の形態では、CPU302
は、予めROMなどの記憶装置303に記憶された制御
プログラムを実行することにより各種制御を行うことと
しているが、制御プログラムは、コンピュータ読み取り
可能な記録媒体(例えばCD−ROMなど)に記憶され
ていても良く、このような形態では、CPU302は、
読取装置305(例えばCD−ROMドライブなど)を
介して記録媒体から制御プログラムを読み込んで、読み
込んだ制御プログラムをRAMやハードディスクなどの
記憶装置303に展開した後、制御プログラムを実行す
る。このような形態とすることにより、本発明を汎用コ
ンピュータで容易に実現することが可能となる。In this embodiment, the CPU 302
Performs various controls by executing a control program stored in a storage device 303 such as a ROM in advance. The control program is stored in a computer-readable recording medium (for example, a CD-ROM). In such a form, the CPU 302
The control program is read from a recording medium via a reading device 305 (for example, a CD-ROM drive), and the read control program is loaded into a storage device 303 such as a RAM or a hard disk, and then the control program is executed. With such an embodiment, the present invention can be easily realized by a general-purpose computer.
【0033】例えば、利用者が英文「I saw chicks in
the nest.」の前置詞句「in the nest」に関する前置詞
句共起情報「see,chick,in nest,形容詞的用法」を
ユーザ共起情報リスト103bに登録しようとする場
合、まず入力部100に登録機能選択指示2bを入力
し、続いて用法情報付加文6「I saw chicks ppn^in t
henest.」を入力する。ここで、「ppn ^」は形容詞的
用法の前置詞句の前置詞に付加する付加情報で、「ppv
^」は副詞的用法の前置詞句の前置詞に付加する付加情
報を示す。制御部108は、入力部100から登録機能
選択指示2bを受けると登録動作状態となり、次いで用
法情報付加文6を受けるとそれを共起情報作成部201
に送り前置詞句共起情報5を得る。共起情報作成部40
1は、用法情報付加文6を受け取ると、副詞的用法の前
置詞句の係る被修飾語、前置詞句用法の前置詞句の係る
被修飾語、前置詞句の前置詞、前置詞句の主名詞、を抽
出し、抽出されたそれぞれの語が活用形であれば原形に
戻し、複数形であれば単数形に戻し、付加情報から前置
詞句用法を決定し、前置詞句共起情報5を作成し制御部
108に返す。最後に制御部108は前置詞句共起情報
5を共起情報登録部105に送りユーザ共起情報リスト
103bに登録する。[0033] For example, when a user writes an English sentence "I saw chicks in
When the preposition phrase co-occurrence information “see, chick, in nest, adjective usage” relating to the preposition phrase “in the nest.” of the nest. is to be registered in the user co-occurrence information list 103b, first, it is registered in the input unit 100. The function selection instruction 2b is input, and then the usage information addition sentence 6 “I saw chicks ppn ^ int
henest. " Here, “ppn ^” is additional information to be added to the preposition of the preposition phrase in the adjective usage, and “ppv
“^” indicates additional information to be added to the preposition of the preposition phrase in the adverb usage. The control unit 108 enters a registration operation state when the registration function selection instruction 2b is received from the input unit 100, and then receives the usage information additional sentence 6 and, upon receipt of the usage information addition sentence 6, transmits it to the co-occurrence information creation unit 201
To obtain prepositional phrase co-occurrence information 5. Co-occurrence information creation unit 40
1 receives the usage information addition sentence 6 and extracts a qualified word of the preposition phrase of the adverbial usage, a qualified word of the preposition phrase of the prepositional phrase usage, a preposition of the preposition phrase, and a main noun of the preposition phrase. If each extracted word is inflected, it is returned to the original form, if it is plural, it is returned to the singular, the preposition phrase usage is determined from the additional information, the preposition phrase co-occurrence information 5 is created, and the return. Finally, the control unit 108 sends the preposition phrase co-occurrence information 5 to the co-occurrence information registration unit 105 and registers it in the user co-occurrence information list 103b.
【0034】以上のように、本実施の形態によれば、用
法情報付加文から共起情報を作成する共起情報作成部4
01を備えることにより、前置詞句共起語の原文からの
抽出と、抽出されたそれぞれの前置詞句共起語の活用形
又は複数形からの単数原形への変換と、前置詞句共起情
報5の作成とを自動化することが可能となり、前置詞句
共起情報登録時における利用者の操作性が向上する。な
お、本実施の形態では原言語を英語、目標言語を日本語
として例示したが、原言語及び目標言語はこれらに限定
されるものではない。As described above, according to the present embodiment, the co-occurrence information creating unit 4 for creating co-occurrence information from the usage information additional sentence
01, the preposition phrase co-occurrence is extracted from the original sentence, the extracted preposition phrase co-occurrence is converted from the inflected form or plural form to the singular original form, and the preposition phrase co-occurrence information 5 is extracted. It is possible to automate the creation and the operability of the user when registering the prepositional phrase co-occurrence information is improved. In the present embodiment, the source language is exemplified as English and the target language is exemplified as Japanese, but the source language and the target language are not limited to these.
【0035】(実施の形態3)図6は本発明の実施の形
態3における機械翻訳装置の機能ブロック図である。図
6において、実施の形態1で説明した各部については同
符号を付して説明は省略する。本実施の形態は、実施の
形態1の機械翻訳装置において、原文の前置詞句共起情
報を一時的に記憶する共起情報一時記憶部501と、利
用者の変更指示により前置詞句の用法を変更する用法変
更部502と、出力部の一つとして情報を表示する表示
面107bを有する表示部107aと、入力部の一つと
して表示面107b上の特定の位置を指示する指示部1
00aとを備えることにより、前置詞句共起情報登録操
作における利用者の操作性の向上を図ったものである。(Embodiment 3) FIG. 6 is a functional block diagram of a machine translation apparatus according to Embodiment 3 of the present invention. In FIG. 6, the components described in the first embodiment are denoted by the same reference numerals, and description thereof will be omitted. In this embodiment, a co-occurrence information temporary storage unit 501 that temporarily stores prepositional phrase co-occurrence information of an original sentence and a usage of a preposition phrase are changed by a user's change instruction in the machine translation apparatus of the first embodiment. The usage changing unit 502, a display unit 107a having a display surface 107b for displaying information as one of the output units, and an instruction unit 1 for specifying a specific position on the display surface 107b as one of the input units
By providing 00a, the operability of the user in the prepositional phrase co-occurrence information registration operation is improved.
【0036】図7は、本発明の実施の形態3における機
械翻訳装置の装置ブロック図である。FIG. 7 is a device block diagram of a machine translation device according to the third embodiment of the present invention.
【0037】図7において、実施の形態1で説明した各
装置については同符号を付して説明は省略する。301
aは入力装置の一である指示装置で、マウス・タッチパ
ネル・トラックボール等のポインティングデバイス等に
より構成される。304aは出力装置の一である表示装
置で、ディスプレイ等で構成される。305は読取装置
でフロッピーディスクドライブ、CD−ROMドライ
ブ、磁気テープリーダー等の1乃至複数の組み合わせで
構成される。306は読取装置305で読み取り可能な
記録媒体で、フロッピーディスク、CD−ROM、磁気
テープ等で構成される。In FIG. 7, the same reference numerals are given to the devices described in the first embodiment, and the description is omitted. 301
A pointing device a is one of the input devices, and is configured by a pointing device such as a mouse, a touch panel, and a trackball. A display device 304a is one of the output devices, and includes a display and the like. A reading device 305 includes one or more combinations of a floppy disk drive, a CD-ROM drive, a magnetic tape reader, and the like. Reference numeral 306 denotes a recording medium that can be read by the reading device 305, and includes a floppy disk, a CD-ROM, a magnetic tape, and the like.
【0038】入力部100は入力装置301により実現
され、指示部100aは指示装置301aにより実現さ
れ、共起情報記憶部103・共起情報一時記憶部501
は記憶装置303により実現され、共起語抽出部102
・共起情報検索部104・共起情報登録部105・翻訳
部106・用法変更部502・制御部108はCPU3
02により実現され、表示部107aは表示装置304
aにより実現され、出力部107は出力装置304によ
り実現される。The input unit 100 is realized by the input device 301, and the instruction unit 100a is realized by the instruction device 301a, and the co-occurrence information storage unit 103 and the co-occurrence information temporary storage unit 501 are used.
Is realized by the storage device 303, and the co-occurrence word extraction unit 102
A co-occurrence information search unit 104; a co-occurrence information registration unit 105; a translation unit 106; a usage change unit 502;
02, and the display unit 107a includes a display device 304
a, and the output unit 107 is realized by the output device 304.
【0039】なお、本実施の形態では、CPU302
は、予めROMなどの記憶装置303に記憶された制御
プログラムを実行することにより各種制御を行うことと
しているが、制御プログラムは、コンピュータ読み取り
可能な記録媒体(例えばCD−ROMなど)に記憶され
ていても良く、このような形態では、CPU302は、
読取装置305(例えばCD−ROMドライブなど)を
介して記録媒体から制御プログラムを読み込んで、読み
込んだ制御プログラムをRAMやハードディスクなどの
記憶装置303に展開した後、制御プログラムを実行す
る。このような形態とすることにより、本発明を汎用コ
ンピュータで容易に実現することが可能となる。In the present embodiment, the CPU 302
Performs various controls by executing a control program stored in a storage device 303 such as a ROM in advance. The control program is stored in a computer-readable recording medium (for example, a CD-ROM). In such a form, the CPU 302
The control program is read from a recording medium via a reading device 305 (for example, a CD-ROM drive), and the read control program is loaded into a storage device 303 such as a RAM or a hard disk, and then the control program is executed. With such an embodiment, the present invention can be easily realized by a general-purpose computer.
【0040】以下、本実施の形態における機械翻訳装置
の動作について説明する。まず、利用者は、入力部10
0に翻訳したい原文1を入力する。制御部108は、入
力部100から原文1を受けると、原文1を共起語抽出
部102に送り、原文1から前置詞句共起語3を抽出す
る。次に、制御部108は、前置詞句共起語3を共起情
報検索部104に送り、共通共起情報リスト103a及
びユーザ共起情報リスト103bから前置詞句共起語3
に対応する前置詞句共起情報3aを検索抽出する。次
に、制御部108は、前置詞句共起情報3aと原文1と
を翻訳部106に送り、目的言語の翻訳文4を得る。次
に、制御部108は原文1と翻訳文4を表示部107a
に送り、翻訳文4を表示する。利用者は、表示された翻
訳結果を読み、前置詞句の用法が誤っている可能性があ
ると判断した場合、表示面107b上に表示された該当
する前置詞句の前置詞を指示部100aにより指示す
る。制御部108は表示面107b上の指示部100a
で指示された位置に表示されている前置詞句の前置詞句
共起情報7を共起情報一時記憶部501から読み出し、
その前置詞句の現在の用法と他の用法とを表示部107
aに表示する。利用者は、正しいと思われる前置詞句用
法の表示箇所を指示部100aにより指示する。制御部
108は指示部100aで指示された位置に表示されて
いる前置詞句用法8を読み出し、前置詞句共起情報7と
前置詞句用法8とを用法変更部502に送り前置詞句共
起情報7の前置詞句用法を前置詞句用法8に変更し、変
更された前置詞句共起情報7aを共起情報登録部105
に送りユーザ共起情報リスト103bに登録し、原文1
と前置詞句共起情報7aとを翻訳部106に送り利用者
の指示する前置詞句用法8により翻訳された翻訳文4a
を取得し、翻訳文4aを表示部107aに送り表示す
る。Hereinafter, the operation of the machine translation apparatus according to the present embodiment will be described. First, the user enters the input unit 10
Enter the original sentence 1 you want to translate into 0. When receiving the original sentence 1 from the input unit 100, the control unit 108 sends the original sentence 1 to the co-occurrence word extraction unit 102, and extracts the preposition phrase co-occurrence word 3 from the original sentence 1. Next, the control unit 108 sends the preposition phrase co-occurrence word 3 to the co-occurrence information search unit 104, and extracts the preposition phrase co-occurrence word 3 from the common co-occurrence information list 103a and the user co-occurrence information list 103b.
Is retrieved and extracted. Next, the control unit 108 sends the preposition phrase co-occurrence information 3a and the original sentence 1 to the translating unit 106, and obtains the translated sentence 4 of the target language. Next, the control unit 108 displays the original sentence 1 and the translated sentence 4 on the display unit 107a.
To display the translated sentence 4. When the user reads the displayed translation result and determines that there is a possibility that the usage of the preposition phrase is incorrect, the user designates the preposition of the corresponding preposition phrase displayed on the display surface 107b by the instruction unit 100a. . The control unit 108 controls the indication unit 100a on the display surface 107b.
The preposition phrase co-occurrence information 7 of the preposition phrase displayed at the position indicated by the above is read out from the co-occurrence information temporary storage unit 501,
The display unit 107 displays the current usage of the preposition phrase and other usages.
Display in a. The user instructs the display part of the preposition phrase usage that seems to be correct by the instruction unit 100a. The control unit 108 reads the preposition phrase usage 8 displayed at the position indicated by the instruction unit 100a, sends the preposition phrase co-occurrence information 7 and the preposition phrase usage 8 to the usage change unit 502, and reads the preposition phrase co-occurrence information 7 The preposition phrase usage is changed to the preposition phrase usage 8, and the changed preposition phrase co-occurrence information 7a is stored in the co-occurrence information registration unit 105.
To the user co-occurrence information list 103b,
And the preposition phrase co-occurrence information 7a to the translation unit 106, and the translated sentence 4a translated by the preposition phrase usage 8 specified by the user.
Is acquired and the translated sentence 4a is sent to the display unit 107a for display.
【0041】例えば、指示部100aとしてマウス、表
示部107aとしてディスプレイを用いて構成した場
合、前置詞句共起情報の登録操作は次のようになる。ま
ず、原文1を入力すると翻訳が行われ、図8のように原
文1と翻訳文4との対応する文同士が分かり易いように
横に並べて表示される。利用者は、表示された翻訳結果
を読み、前置詞句の用法が誤っている可能性があると判
断した場合、ディスプレイ上に表示された該当する前置
詞句の前置詞にマウスカーソルを移動させ、マウスのマ
ウスボタンをクリックする。制御部108はマウス10
0aからマウスイベントを受けると、マウスカーソル位
置に表示されている前置詞句の前置詞句共起情報7を共
起情報一時記憶部501から読み出し、その前置詞句の
用法を、ディスプレイに表示する(図9参照)。利用者
は、正しいと思われる前置詞句用法の表示箇所にマウス
カーソルを移動しマウスボタンをクリックする(図10
参照)。制御部はマウスカーソル位置に表示されている
前置詞句用法8を読み出し、前置詞句共起情報7と前置
詞句用法8とを用法変更部に送り前置詞句共起情報7の
前置詞句用法を前置詞句用法8に変更し、変更された前
置詞句共起情報7aを共起情報登録部105に送りユー
ザ共起情報リスト103bに登録し、原文1と前置詞句
共起情報7aとを翻訳部106に送り利用者の指示する
前置詞句用法8により翻訳された翻訳文4aを取得し、
翻訳文4aをディスプレイに表示する(図11参照)。For example, when a mouse is used as the instruction unit 100a and a display is used as the display unit 107a, the operation of registering the preposition phrase co-occurrence information is as follows. First, when the original sentence 1 is input, translation is performed. As shown in FIG. 8, the corresponding sentences of the original sentence 1 and the translated sentence 4 are displayed side by side for easy understanding. When the user reads the translation result displayed and determines that the usage of the preposition phrase may be incorrect, the user moves the mouse cursor to the preposition of the corresponding preposition phrase displayed on the display, and Click the mouse button. The control unit 108 controls the mouse 10
When a mouse event is received from 0a, the preposition phrase co-occurrence information 7 of the preposition phrase displayed at the position of the mouse cursor is read from the co-occurrence information temporary storage unit 501, and the usage of the preposition phrase is displayed on the display (FIG. 9). reference). The user moves the mouse cursor to the position of the preposition phrase usage that seems to be correct and clicks the mouse button (FIG. 10).
reference). The control unit reads the preposition phrase usage 8 displayed at the position of the mouse cursor, sends the preposition phrase co-occurrence information 7 and the preposition phrase usage 8 to the usage change unit, and converts the preposition phrase usage of the preposition phrase co-occurrence information 7 to the preposition phrase usage. 8 and sends the changed preposition phrase co-occurrence information 7a to the co-occurrence information registration unit 105 to register it in the user co-occurrence information list 103b, and sends the original text 1 and the preposition phrase co-occurrence information 7a to the translation unit 106 for use. The translated sentence 4a translated by the preposition phrase usage 8 specified by the person,
The translated sentence 4a is displayed on the display (see FIG. 11).
【0042】以上のように、本実施の形態によれば、入
力部100の一として指示部100aと、出力部107
の一として表示部107aと、共起情報一時記憶部50
1と、用法変更部502とを備えることにより、ユーザ
共起情報リストへの前置詞句共起情報の登録操作が簡便
化され、前置詞句共起情報登録時における利用者の操作
性が向上する。なお、本実施の形態では原言語を英語、
目標言語を日本語として例示したが、原言語及び目標言
語はこれらに限定されるものではない。As described above, according to the present embodiment, the instructing unit 100 a and the output unit 107 are one of the input units 100.
The display unit 107a and the co-occurrence information temporary storage unit 50
1 and the usage change unit 502, the operation of registering the preposition phrase co-occurrence information in the user co-occurrence information list is simplified, and the operability of the user at the time of registering the preposition phrase co-occurrence information is improved. In this embodiment, the source language is English,
Although the target language is exemplified as Japanese, the source language and the target language are not limited to these.
【0043】[0043]
【発明の効果】以上のように請求項1に記載の機械翻訳
装置によれば、原言語の文を目標言語の文に翻訳する機
械翻訳装置であって、原言語の前置詞句の係り先に関す
る前置詞句共起情報を記憶する共通共起情報リストと、
利用者により登録された原言語の前置詞句の係り先に関
する前置詞句共起情報を記憶する追加登録可能なユーザ
共起情報リストと、ユーザ共起情報リストに前置詞句共
起情報を登録する共起情報登録部と、原言語の原文の前
置詞句の係り先に関する前置詞句共起語を抽出する共起
語抽出部と、前置詞句共起語に対応する前置詞句共起情
報を共起情報リストから検索する共起情報検索部と、前
置詞句共起情報により前置詞句の係る被修飾句を決定し
原文を翻訳する翻訳部とを備えたことにより、前置詞句
の修飾先の曖昧性が排除され翻訳精度が向上し、利用者
が共起情報リストに不足している前置詞句共起情報を補
充することが可能となり、補充を行うことによって翻訳
精度が更に向上し、利用者毎の特有な表現に対応して翻
訳精度が向上するという有利な効果が得られる。As described above, according to the machine translation device of the first aspect, the machine translation device translates a source language sentence into a target language sentence, and relates to a destination of a preposition phrase in the source language. A common co-occurrence information list for storing prepositional phrase co-occurrence information,
An additional registrable user co-occurrence information list that stores prepositional phrase co-occurrence information related to the destination of prepositional phrases in the source language registered by the user, and a co-occurrence that registers prepositional phrase co-occurrence information in the user co-occurrence information list An information registration unit, a co-occurrence word extraction unit that extracts a preposition phrase co-occurrence word relating to a destination of a preposition phrase in the source language of the source language, and a preposition phrase co-occurrence information corresponding to the preposition phrase co-occurrence information from the co-occurrence information list By providing a co-occurrence information search unit for searching and a translating unit for translating the original text by determining the qualifying phrase of the preposition phrase based on the preposition phrase co-occurrence information, the ambiguity of the modification destination of the preposition phrase is eliminated and translation is performed. The accuracy is improved, and the user can supplement the preposition phrase co-occurrence information that is lacking in the co-occurrence information list, and by performing the supplement, the translation accuracy is further improved, and the expression is unique to each user. Correspondence improves translation accuracy Advantageous effect that can be obtained.
【0044】請求項2の機械翻訳装置によれば、請求項
1に記載の機械翻訳装置において、原言語の用法情報付
加文から前置詞句共起情報を作成する共起情報作成部を
備えたことにより、前置詞句共起情報のユーザ共起情報
リストへの登録時に、前置詞句共起語の原文からの抽出
と、抽出されたそれぞれの前置詞句共起語の活用形又は
複数形からの単数原形への変換と、前置詞句共起情報の
作成とを自動化することが可能となり、前置詞句共起情
報登録時における利用者の操作性が向上するという有利
な効果が得られる。According to a second aspect of the present invention, there is provided a machine translation apparatus according to the first aspect, further comprising a co-occurrence information creating unit for creating preposition phrase co-occurrence information from the usage information additional sentence of the source language. Therefore, when registering the prepositional phrase co-occurrence information in the user co-occurrence information list, extraction of the prepositional phrase co-occurrence word from the original sentence, and singular original form from the inflected form or plural form of each extracted prepositional phrase co-occurrence word And the creation of prepositional phrase co-occurrence information can be automated, and an advantageous effect that user operability at the time of registration of prepositional phrase co-occurrence information is improved is obtained.
【0045】請求項3の機械翻訳装置によれば、請求項
1に記載の機械翻訳装置において、情報を表示する表示
面を有する表示部と、表示面上の特定の位置を指示する
指示部と、前置詞句共起情報を一時的に記憶する共起情
報一時記憶部と、利用者の前置詞句用法の変更指示によ
り前記前置詞句の用法を変更する用法変更部とを備えた
ことにより、前置詞句共起情報のユーザ共起情報リスト
への登録時に、ユーザ共起情報リストへの前置詞句共起
情報の登録操作が簡便化され、前置詞句共起情報登録時
における利用者の操作性が向上するという有利な効果が
得られる。According to a third aspect of the present invention, there is provided the machine translation apparatus according to the first aspect, further comprising: a display unit having a display surface for displaying information; and an instruction unit for designating a specific position on the display surface. A co-occurrence information temporary storage unit that temporarily stores preposition phrase co-occurrence information, and a usage changing unit that changes the usage of the preposition phrase according to a user's instruction to change the usage of the preposition phrase. When the co-occurrence information is registered in the user co-occurrence information list, the operation of registering the prepositional phrase co-occurrence information in the user co-occurrence information list is simplified, and the operability of the user when registering the prepositional phrase co-occurrence information is improved. The advantageous effect described above can be obtained.
【0046】本発明の請求項4に記載の機械翻訳方法に
よれば、原言語の文を目標言語の文に翻訳する機械翻訳
方法であって、原言語の原文の前置詞句の係り先に関す
る前置詞句共起語を抽出する共起語抽出過程と、原言語
の前置詞句の係り先に関する前置詞句共起情報を記憶さ
れた予め用意された共通共起情報リストと利用者により
登録された原言語の前置詞句の係り先に関する前置詞句
共起情報が記憶された追加登録可能なユーザ共起情報リ
ストとから前置詞句共起語に対応する前置詞句共起情報
を検索する共起情報検索過程と、前置詞句共起情報によ
り前置詞句の係る被修飾句を決定し原文を翻訳する翻訳
過程と、を備えたことにより、前置詞句の修飾先の曖昧
性が排除され翻訳精度が向上するという有利な効果が得
られる。According to a fourth aspect of the present invention, there is provided a machine translation method for translating a source language sentence into a target language sentence, comprising: A co-occurrence word extraction process for extracting phrase co-occurrence words, a pre-prepared common co-occurrence information list storing preposition phrase co-occurrence information related to the destination of the preposition phrase in the source language, and the source language registered by the user A co-occurrence information search process for searching for a preposition / phrase co-occurrence information corresponding to a preposition / phrase co-occurrence word from an additionally-registerable user co-occurrence information list in which preposition / phrase co-occurrence information relating to a destination of a preposition / phrase is stored; The translation process of determining the qualifying phrase of the preposition phrase based on the preposition phrase co-occurrence information and translating the original text, thereby eliminating the ambiguity of the modification destination of the preposition phrase and improving the translation accuracy. Is obtained.
【図1】本発明の実施の形態1における機械翻訳装置の
機能ブロック図FIG. 1 is a functional block diagram of a machine translation device according to a first embodiment of the present invention.
【図2】本発明の実施の形態1における共起情報リスト
の内部状態を示す図FIG. 2 is a diagram showing an internal state of a co-occurrence information list according to the first embodiment of the present invention.
【図3】本発明の実施の形態1における機械翻訳装置の
装置ブロック図FIG. 3 is a device block diagram of a machine translation device according to the first embodiment of the present invention.
【図4】本発明の実施の形態1における前置詞句共起情
報入力画面を示す図FIG. 4 is a diagram showing a prepositional phrase co-occurrence information input screen according to the first embodiment of the present invention;
【図5】本発明の実施の形態2における機械翻訳装置の
機能ブロック図FIG. 5 is a functional block diagram of a machine translation device according to a second embodiment of the present invention.
【図6】本発明の実施の形態3における機械翻訳装置の
機能ブロック図FIG. 6 is a functional block diagram of a machine translation device according to a third embodiment of the present invention.
【図7】本発明の実施の形態3における機械翻訳装置の
装置ブロック図FIG. 7 is a device block diagram of a machine translation device according to a third embodiment of the present invention.
【図8】本発明の実施の形態3における翻訳結果表示画
面を示す図FIG. 8 is a diagram showing a translation result display screen according to the third embodiment of the present invention.
【図9】本発明の実施の形態3における前置詞句用法表
示画面を示す図FIG. 9 is a diagram showing a preposition phrase usage display screen according to Embodiment 3 of the present invention.
【図10】本発明の実施の形態3における前置詞句用法
変更指示表示画面を示す図FIG. 10 is a diagram showing a preposition phrase usage change instruction display screen according to the third embodiment of the present invention.
【図11】本発明の実施の形態3における前置詞句用法
変更後表示画面を示す図FIG. 11 is a diagram showing a display screen after a change of prepositional phrase usage according to Embodiment 3 of the present invention.
100 入力部 100a 指示部 102 共起語抽出部 103 共起情報記憶部 103a 共通共起情報リスト 103b ユーザ共起情報リスト 104 共起情報検索部 105 共起情報登録部 106 翻訳部 107 出力部 107a 表示部 108 制御部 301 入力装置 301a 指示装置 302 CPU 303 記憶装置 304 出力装置 304a 表示装置 305 読取装置 306 記録媒体 501 共起情報一時記憶部 502 用法変更部 Reference Signs List 100 input unit 100a instruction unit 102 co-occurrence word extraction unit 103 co-occurrence information storage unit 103a common co-occurrence information list 103b user co-occurrence information list 104 co-occurrence information search unit 105 co-occurrence information registration unit 106 translation unit 107 output unit 107a display Unit 108 control unit 301 input device 301a pointing device 302 CPU 303 storage device 304 output device 304a display device 305 reading device 306 recording medium 501 co-occurrence information temporary storage unit 502 usage change unit
Claims (4)
翻訳装置であって、原言語の前置詞句の係り先に関する
前置詞句共起情報を記憶する共通共起情報リストと、利
用者により登録された原言語の前置詞句の係り先に関す
る前置詞句共起情報を記憶する追加登録可能なユーザ共
起情報リストと、前記ユーザ共起情報リストに前置詞句
共起情報を登録する共起情報登録部と、原言語の原文の
前置詞句の係り先に関する前置詞句共起語を抽出する共
起語抽出部と、前記前置詞句共起語に対応する前置詞句
共起情報を前記共起情報リストから検索する共起情報検
索部と、前記前置詞句共起情報により前記前置詞句の係
る被修飾句を決定し原文を翻訳する翻訳部と、を備えた
ことを特徴とする機械翻訳装置。1. A machine translator for translating a source language sentence into a target language sentence, comprising: a common co-occurrence information list storing preposition phrase co-occurrence information relating to a destination of a preposition phrase in the source language; An additional registrable user co-occurrence information list that stores preposition phrase co-occurrence information related to the destination of the prepositional phrase in the source language registered by the co-occurrence information that registers prepositional phrase co-occurrence information in the user co-occurrence information list A registration unit, a co-occurrence word extraction unit for extracting a preposition phrase co-occurrence word relating to a destination of a preposition phrase in the source sentence of the source language, and a co-occurrence information list of preposition phrase co-occurrence information corresponding to the preposition phrase co-occurrence word A machine translation apparatus comprising: a co-occurrence information search unit that searches for a pre-phrase phrase from a search term; and a translation unit that determines a qualified phrase related to the preposition phrase based on the preposition phrase co-occurrence information and translates an original sentence.
情報を作成する共起情報作成部を備えることを特徴とす
る請求項1に記載の機械翻訳装置。2. The machine translation apparatus according to claim 1, further comprising a co-occurrence information creating unit for creating preposition phrase co-occurrence information from the usage information additional sentence of the source language.
前記表示面上の特定の位置を指示する指示部と、前記前
置詞句共起情報を一時的に記憶する共起情報一時記憶部
と、利用者の前置詞句用法の変更指示により前記前置詞
句の用法を変更する用法変更部とを備えることを特徴と
する請求項1に記載の機械翻訳装置。3. A display unit having a display surface for displaying information,
An instruction unit for indicating a specific position on the display surface; a co-occurrence information temporary storage unit for temporarily storing the preposition phrase co-occurrence information; and a usage of the preposition phrase according to a user's instruction to change the preposition phrase usage. 2. The machine translation device according to claim 1, further comprising a usage changing unit that changes the translation.
翻訳方法であって、原言語の原文の前置詞句の係り先に
関する前置詞句共起語を抽出する共起語抽出過程と、原
言語の前置詞句の係り先に関する前置詞句共起情報を記
憶された予め用意された共通共起情報リストと利用者に
より登録された原言語の前置詞句の係り先に関する前置
詞句共起情報が記憶された追加登録可能なユーザ共起情
報リストとから前記前置詞句共起語に対応する前置詞句
共起情報を検索する共起情報検索過程と、前記前置詞句
共起情報により前記前置詞句の係る被修飾句を決定し原
文を翻訳する翻訳過程と、を備えたことを特徴とする機
械翻訳方法。4. A machine translation method for translating a source language sentence into a target language sentence, comprising: a co-occurrence word extraction step of extracting a preposition phrase co-occurrence word relating to a destination of a preposition phrase in the source language original sentence; A pre-prepared common co-occurrence information list that stores prepositional phrase co-occurrence information relating to the source language preposition phrase and the preposition phrase co-occurrence information relating to the source language preposition phrase that is registered by the user are stored. A co-occurrence information search step of searching for a preposition phrase co-occurrence information corresponding to the preposition phrase co-occurrence word from the additionally registered user co-occurrence information list, and a subject of the preposition phrase according to the preposition phrase co-occurrence information. A translation process for determining a modifier and translating the original text.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP10057799A JPH11259474A (en) | 1998-03-10 | 1998-03-10 | Machine translation system and machine translating method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP10057799A JPH11259474A (en) | 1998-03-10 | 1998-03-10 | Machine translation system and machine translating method |
Publications (1)
Publication Number | Publication Date |
---|---|
JPH11259474A true JPH11259474A (en) | 1999-09-24 |
Family
ID=13065965
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
JP10057799A Pending JPH11259474A (en) | 1998-03-10 | 1998-03-10 | Machine translation system and machine translating method |
Country Status (1)
Country | Link |
---|---|
JP (1) | JPH11259474A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2011007627A1 (en) * | 2009-07-17 | 2011-01-20 | 日本電気株式会社 | Speech processing device, method, and storage medium |
CN109614624A (en) * | 2018-12-12 | 2019-04-12 | 广东小天才科技有限公司 | English sentence recognition method and electronic equipment |
-
1998
- 1998-03-10 JP JP10057799A patent/JPH11259474A/en active Pending
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2011007627A1 (en) * | 2009-07-17 | 2011-01-20 | 日本電気株式会社 | Speech processing device, method, and storage medium |
US9583095B2 (en) | 2009-07-17 | 2017-02-28 | Nec Corporation | Speech processing device, method, and storage medium |
CN109614624A (en) * | 2018-12-12 | 2019-04-12 | 广东小天才科技有限公司 | English sentence recognition method and electronic equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US4774666A (en) | Translating apparatus | |
JP2654001B2 (en) | Machine translation method | |
JP4319860B2 (en) | Method and apparatus for developing a transfer dictionary for use in a transfer-based machine translation system | |
JPS62203273A (en) | Mechanical translation system | |
JP2815714B2 (en) | Translation equipment | |
US4860206A (en) | Translation system allowing user designation of postpositional words | |
JP4875040B2 (en) | Machine translation system and machine translation program | |
JPH11259474A (en) | Machine translation system and machine translating method | |
JPS61260366A (en) | Mechanical translating system having learning function | |
JP6045948B2 (en) | Machine translation apparatus and machine translation program | |
JPS62191967A (en) | Context processor | |
JPH11259473A (en) | Machine translation system and machine translating method | |
JPH11282844A (en) | Preparing method of document, information processor and recording medium | |
JP3313810B2 (en) | Aspect processing device | |
JPH0410665B2 (en) | ||
JP4021813B2 (en) | Compound word registration program and registration device | |
JPS61260367A (en) | Mechanical translating system | |
JP2608384B2 (en) | Machine translation apparatus and method | |
JPH10293761A (en) | Machine translation device, machine translation method and recording medium recording machine translation program | |
JPH09185629A (en) | Machine translation method | |
JPH0444981B2 (en) | ||
JP2007095099A (en) | Translation support device and translation support program | |
JPH01207873A (en) | Translation supporting device | |
JP2004326302A (en) | Translation support device and translation support program | |
JPH0612449A (en) | System and method for machine translation |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
A621 | Written request for application examination |
Free format text: JAPANESE INTERMEDIATE CODE: A621 Effective date: 20050302 |
|
RD01 | Notification of change of attorney |
Free format text: JAPANESE INTERMEDIATE CODE: A7421 Effective date: 20050413 |
|
RD01 | Notification of change of attorney |
Free format text: JAPANESE INTERMEDIATE CODE: A7421 Effective date: 20050627 |
|
A131 | Notification of reasons for refusal |
Free format text: JAPANESE INTERMEDIATE CODE: A131 Effective date: 20070320 |
|
A521 | Written amendment |
Free format text: JAPANESE INTERMEDIATE CODE: A523 Effective date: 20070518 |
|
A131 | Notification of reasons for refusal |
Free format text: JAPANESE INTERMEDIATE CODE: A131 Effective date: 20070605 |
|
A02 | Decision of refusal |
Free format text: JAPANESE INTERMEDIATE CODE: A02 Effective date: 20070821 |