JP2004062726A - Translation device, translation method, program and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a translation device that more accurately specifies a similar translation expression referred to for correct translation of an input expression. <P>SOLUTION: The translation device comprises a translation expression storage part 12 storing translation expressions consisting of sets of source language expressions and target language expressions corresponding to the source language expressions, a source-language semantic distance calculation part 13 for calculating semantic distances between source language expressions, a target-language semantic distance calculation part 15 for calculating semantic distances between target language expressions, a similar translation expression specification part 14 for controlling the source-language semantic distance calculation part 13 to specify the most similar translation expression to an input expression, a translation expression application part 161 for translating the input expression with the specified translation expression, and an additional word selection part 162 for controlling the target-language semantic distance calculation part 15 to select a word of a portion whose word can not be decided by the translation expression application part 14. In example translation, when a source language expression in an example is obscure in meaning, the meaning is specified in advance with semantic distances in the target language, so that an example to be referred to is specified more correctly. <P>COPYRIGHT: (C)2004,JPO

Description

【０１００】

【０１０１】
この結果、Ｓ２１との距離の方が近いため、Ｓ２１に対応する意味である動物のネズミの意味（０４）が、この対訳表現における”ｍｏｕｓｅ”の意味となる。
【０１０２】
ここで、用例記憶部３２には、以下の対訳表現が存在したとする。
【０１０３】
Ｅ２：　Ｉ　ｋｅｅｐ　ａ　ｍｏｕｓｅ　→　私はネズミを飼っている
Ｅ３：　Ｉ　ｋｅｅｐ　ａ　ｋｎｉｆｅ　→　私はナイフを持っている
【０１０４】
また入力表現として、以下が与えられたとする。
【０１０５】
Ｉ２：　Ｉ　ｋｅｅｐ　ａ　ｋｅｙｂｏａｒｄ．
【０１０６】
入力部３１から英語表現Ｉ２が入力されると、類似対訳表現特定部３６が対訳表現記憶部３２から最も入力に類似した原言語に対する対訳表現を特定する。
【０１０７】
この際、上記の対訳表現Ｅ２中の原言語表現の”ｍｏｕｓｅ”は動物のネズミの意味（図１０における（０５））であることが分かっており、しかも”ｋｅｙｂｏａｒｄ”（０９）と動物の”ｍｏｕｓｅ”（０５）の距離は、”ｋｅｙｂｏａｒｄ”（０９）と”ｋｎｉｆｅ”（０８）の距離より遠いため、第一の従来例と異なり参照する対訳表現として対訳表現Ｅ３が選ばれる。よってＩ２の翻訳に対訳表現Ｅ２を参照して、誤って「私はキーボードを飼っている」と翻訳せずに、対訳表現Ｅ３を参照して「私はナイフを持っている」と翻訳できるようになる。
【０１０８】
対訳表現の意味の特定は、入力表現が与えられた後に行っても、予め行っておいても良い。
【０１０９】
以上説明したように、本実施例においては、目的言語側の意味距離計算を行って予め各対訳表現の原言語表現の意味の曖昧性を解消しておくことで、第一の従来例では正しく類似対訳表現が特定できなかった場合に関しても、より正確に類似対訳表現を特定できるようになる、という作用効果を有する。
【０１１０】
［実施の形態３］
本発明は、規則ベース翻訳方式において、翻訳規則から予め類似規則を生成して、入力表現に対する翻訳規則のカバー率を上げる場合にも応用可能である。
【０１１１】
図６は、本発明の第３の実施の形態の構成を示す図である。図６を参照して、本発明の第３の実施の形態について説明する。本発明の第３の実施の形態は、翻訳規則を入力する翻訳規則入力部６１と、原言語表現間の意味距離を計算する原言語意味距離計算部６２と、目的言語表現間の意味距離を計算する目的言語意味距離計算部６３と、原言語意味距離計算部６２、目的言語意味距離計算部６３を利用して、入力された翻訳規則の適用条件の原言語表現の意味の曖昧性を解消する規則意味特定部６４と、特定された原言語表現の意味に基づいて類似する原言語表現を求め、入力された翻訳規則を前記類似する原言語表現によって置き換えて類似規則を生成する類似規則生成部６５と、置き換えた原言語表現に対する訳語を選択する規則中訳語選択部６６と、生成された類似規則を出力する出力部６７とを備える。
【０１１２】
上記構成において、原言語意味距離計算部６２、目的言語意味距離計算部６３は発明の実施の形態２における原言語意味距離計算部３３、目的言語意味距離計算部３４と同様である。
【０１１３】
翻訳規則入力部６１は、入力された原言語表現が特定の構造を持つ場合、その構造の目的言語への訳し方を定めた翻訳規則を入力する。
【０１１４】
図７に、原言語が英語、目的言語が日本語の場合の翻訳規則の一例を示す。
【０１１５】
規則意味特定部６４は、発明の実施の形態２における対訳表現意味特定部３５と同様の働きを持ち、入力された翻訳規則の適用条件中の原言語表現に意味の曖昧性がある場合は、その翻訳規則で原言語表現に与えられた訳から、原言語意味距離計算手段６２、目的言語意味距離計算手段６３を用いて意味の特定を行う。
【０１１６】
類似規則生成部６５は、規則意味特定部６４によって特定された原言語の意味に基づいて、適用条件の原言語表現を類似表現に置き換えた規則を生成する。
【０１１７】
図８に、置き換え規則の一例を示す。
【０１１８】
規則中訳語選択部６６は、前記第１の発明の実施の形態における追加訳語選択部１６２と同様の働きを持ち、置き換えられた類似規則の適用条件中の原言語表現の訳語を、目的言語意味距離計算部６３を利用して選択する。ここでは、追加訳語選択部１６２と同様の方法で、前記原言語表現の各訳語のうち、前記表現に対応する類似規則生成元翻訳規則の適用条件中の表現に振られた訳語と目的言語意味距離が近くなる訳語を選択する。
【０１１９】
次に、図９は、本発明の第３の実施の形態の動作を説明するためのフローチャートである。図９を参照して、本実施の形態の動作について説明する。
【０１２０】
まず、翻訳規則入力部６１から翻訳規則を入力し（ステップ９０１）、入力された翻訳規則の適用条件の原言語表現に意味の曖昧性のある場合は、規則意味特定部６４において、原言語意味距離計算部６２、目的言語意味距離計算部６３を用いて、翻訳規則で原言語の表現に与えられた訳から意味の曖昧性を解消する（ステップ９０２）。
【０１２１】
意味の曖昧性を解消するステップ９０２の詳細は、図５のフローチャートと同様であるため、説明の詳細は省略する。
【０１２２】
次に、類似規則生成部６５において、原言語意味距離計算部６２を利用して、入力された翻訳規則の適用条件中の原言語表現を意味的に類似した原言語表現に置き換えた類似規則を生成して出力する（ステップ９０３）。
【０１２３】
規則中訳語選択部６６において、類似規則を受け取り、目的言語意味距離計算部６３を利用して、類似規則中の置き換えられた原言語表現に対する訳語を決定する（ステップ９０４）。
【０１２４】
最後に、翻訳結果を出力部６７から出力する（ステップ９０５）。
【０１２５】
この実施の形態において、翻訳規則入力部６１と、原言語意味距離計算部６２と、目的言語意味距離計算部６３と、規則意味特定部６４と、類似規則生成部６５と、規則中訳語選択部６６の各処理と、出力部６７の処理制御は、好ましくは、コンピュータで実行されるプログラムによって実現される。
【０１２６】
［実施例３］
次に、具体的な動作例（実施例３）について説明する。この第３の実施例では、前記第３の実施の形態において、英語から日本語への翻訳規則を入力とし、また、図８の翻訳規則の構造を以下のように簡略に表現することにする。
【０１２７】
Ｒ１：　（ｆｅｅｄ　（ＯＢＪ　ｍｏｕｓｅ））→（餌を与える　（に　ネズミ））
【０１２８】
図６の翻訳規則入力部６１にＲ１が入力されると、まず、Ｒ１の適用条件中の原言語表現”ｍｏｕｓｅ”がマウス（パソコン周辺機器）とネズミ（動物）という意味の曖昧性を持つため、規則意味特定部６４において、図３の対訳表現意味特定部３５と同様の曖昧性解消を行い、この翻訳規則における”ｍｏｕｓｅ”の意味が動物のネズミの意味（図１０での（０５））であると特定する。次に、動物の意味の”ｍｏｕｓｅ”の類似表現を原言語意味距離計算部６２を用いて求め、入力された翻訳規則の適用条件中の原言語表現を置き換える。
【０１２９】
（ｆｅｅｄ　（ＯＢＪ　ｍｏｕｓｅ））→（餌を与える　（に　ネズミ））
⇒（ｆｅｅｄ　（ＯＢＪ　ｃａｔ））→（餌を与える　（に　？？？））
⇒（ｆｅｅｄ　（ＯＢＪ　ｐｉｇ））→（餌を与える　（に　？？？））
⇒（ｆｅｅｄ　（ＯＢＪ　ｓｅａｌ））→（餌を与える　（に　？？？））
・・・
【０１３０】
この時、図６の規則意味特定部６４による意味の特定により、”ｍｏｕｓｅ”のもう一つの意味であるパソコン周辺機器のマウス（図１０における（０６））を用いて以下のような誤った類似規則が生成されるのを防いでいる。
【０１３１】
×（ｆｅｅｄ　（ＯＢＪ　ｋｅｙｂｏａｒｄ））→（餌を与える　（に　？？？））
×（ｆｅｅｄ　（ＯＢＪ　ｍｏｄｅｍ））→（餌を与える　（に　？？？））
【０１３２】
次に、翻訳規則の適用条件中の置き換えた原言語表現に対応する訳語を選択する。選択方法は、発明の実施例２の追加訳語選択部１６２と同様に、元となる翻訳規則の置き換え部（”ｍｏｕｓｅ”）に与えられた訳語（「ネズミ」）と、目的言語の意味距離が近くなるような訳語を選択する。
【０１３３】
例えば、”ｓｅａｌ”に置き換えられた類似規則について考えると、”ｓｅａｌ”は「印章」と「アザラシ」という２つの訳語を持つが、例えば図１１のシソーラスを参照すると、ネズミ（０４）と印章（０６）の距離よりネズミ（０４）とアザラシ（０５）の方が距離が近いため、「アザラシ」が訳語として選択される。
【０１３４】
（ｆｅｅｄ　（ＯＢＪ　ｍｏｕｓｅ））→（餌を与える　（に　ネズミ））
⇒（ｆｅｅｄ　（ＯＢＪ　ｃａｔ））→（餌を与える　（に　ネコ））
⇒（ｆｅｅｄ　（ＯＢＪ　ｐｉｇ））→（餌を与える　（に　ブタ））
⇒（ｆｅｅｄ　（ＯＢＪ　ｓｅａｌ））→（餌を与える　（に　アザラシ））
・・・
【０１３５】
なお、本実施例において、規則意味特定部６４と規則中訳語選択部６６のどちらか一方のみを持つ構成としても良い。
【０１３６】
以上説明したように、本実施例においては、翻訳実行時または翻訳実行前に前処理として翻訳規則の類似規則を自動生成する場合に、目的言語側の意味距離計算を行って予め各規則の原言語表現の意味の曖昧性を解消しておくことで、誤った類似規則が生成されることを抑制し、また、類似規則生成時に元となる翻訳規則から置き換えられた原言語表現の訳語をより正確に選択できるようになる、という作用効果を有する。
【０１３７】
［実施の形態４］
本発明の第４の実施の形態について説明する。本発明は、規則主導型翻訳方式において、入力表現に適用条件が合致する翻訳規則が存在しない場合に、類似の翻訳規則を実行時に動的に生成する場合にも応用可能である。この実施の形態は、本発明の第１、第２の実施の形態における対訳表現として翻訳規則を用いるのに相当する。
【０１３８】
図１２は、本発明の第４の実施の形態の構成を示す図である。図１２を参照して、本発明の第４の実施の形態について説明する。本発明の第４の実施の形態は、原言語の入力表現を入力する入力部１３１と、翻訳規則を記憶する規則記憶部１３２と、原言語表現間の意味距離を計算する原言語意味距離計算部１３３と、目的言語表現間の意味距離を計算する目的言語意味距離計算部１３４と、原言語意味距離計算部１３３、目的言語意味距離計算部１３４を利用して、翻訳規則記憶部１３２中の各翻訳規則の適用条件の原言語表現の意味の曖昧性を解消する規則意味特定部１３５と、特定された意味に基づいて、入力表現に類似する規則の特定を行う類似規則特定部１３６と、翻訳を行う翻訳部１３７と、変換された目的言語表現を出力する出力部１３８とを備える。翻訳部１３７は更に、特定された類似規則により翻訳を行う規則適用部１３７１と、規則適用部１３７１で訳が決定しなかった部分の訳語を選択する追加訳語選択部１３７２を含む。
【０１３９】
上記構成において、入力部１３１、原言語意味距離計算部１３３、目的言語意味距離計算部１３４、出力部１３８は、前記第１の実施の形態１における入力部１１、原言語意味距離計算部１３、目的言語意味距離計算部１５、出力部１７と同様である。
【０１４０】
規則意味特定部１３５は、前記第３の実施の形態における規則意味特定部６４と同様である。
【０１４１】
翻訳規則記憶部１３２は、原言語表現が特定の統語構造を持つ場合のその構造の目的言語での訳し方を定めた翻訳規則を記憶する。
【０１４２】
図７には、原言語が英語、目的言語が日本語の場合の翻訳規則の例が示されている。
【０１４３】
類似規則特定部１３６は、第１の実施の形態における類似対訳表現特定部１４に対応し、規則意味特定手段１３５において翻訳規則の適用条件中の原言語表現の特定された意味に基づいて、入力表現に類似する原言語表現に対して適用可能な翻訳規則を特定する。
【０１４４】
規則適用部１３７１は、第１の発明の実施の形態における対訳表現適用部１６１に対応し、特定された翻訳規則により翻訳を行う。
【０１４５】
追加訳語選択部１３７２は、第１の発明の実施の形態における追加訳語選択部１６２に対応し、規則適用部１３７１で訳が決定されなかった部分の訳語を選択する。ここでは、追加訳語選択部１６２と同様の方法で、前記部分の各訳語のうち、前記部分に対応する類似規則生成元翻訳規則の適用条件中の表現に振られた訳語と目的言語意味距離が近くなる訳語を選択する。
【０１４６】
図１３は、本発明の第４の実施の形態の動作を説明するためのフローチャートである。図１２及び図１３を参照して、本発明の第４の実施の形態の動作について説明する。
【０１４７】
予め規則記憶部１３２中の各翻訳規則について、適用条件中の原言語表現に意味の曖昧性のある場合は、規則意味特定部１３５において、原言語意味距離計算部１３３、目的言語意味距離計算部１３４を用いて、原言語表現に与えられた訳から意味の曖昧性を解消しておく（ステップ１４０１）。
【０１４８】
ステップ１４０１の詳細は、前記第３の発明の実施の形態におけるステップ９０２（図９参照）と同様であるため、詳細は省略する。
【０１４９】
その後、入力部に入力表現が入力されると（ステップ１４０２）、類似規則特定部１３６が、原言語意味距離計算部１３３を制御して、規則記憶部１３２から入力表現に類似する原言語表現に適用可能な対訳表現を特定する（ステップ１４０３）。ここでは、予め規則意味特定部１３６によって規則の適用条件中の原言語表現の意味の曖昧性が解消されているため、意味的に入力表現と食い違った原言語表現に適用される規則が特定されることはない。
【０１５０】
次にステップ１４０３で特定された対訳表現から入力表現を翻訳し（ステップ１４０４）、最後に、翻訳結果を出力部から出力する（ステップ１４０６）。
【０１５１】
この実施の形態において、原言語意味距離計算部１３３と、目的言語意味距離計算部１３４と、規則意味特定部１３５と、類似規則特定部１３６と、規則適用部１３７１と、追加訳語選択部１３７２の各処理と、入力部１３１と出力部１３８の処理制御は、好ましくは、コンピュータで実行されるプログラムによって実現される。
【０１５２】
［実施例４］
次に、具体的な動作例（実施例４）について説明する。この第４の実施例では、前記第４の実施の形態において、英語から日本語への翻訳動作について説明する。
【０１５３】
本発明の第４の実施例では、予め、規則記憶部１３２中の各規則の適用条件中の原言語表現の意味の曖昧性解消を行う。
【０１５４】
例として、以下の規則を挙げる。
【０１５５】
Ｒ２：　（ｋｅｅｐ　（ＯＢＪ　ｍｏｕｓｅ））　→　（飼う　（を　ネズミ））
【０１５６】
”ｍｏｕｓｅ”には、動物のネズミの意味（ｍ１）と、パソコン周辺機器のマウスの意味（ｍ２）があるが、Ｒ２においては”ｍｏｕｓｅ”の訳語としてネズミが与えられているため、発明の実施例２のステップ５０２からステップ５０５と同様の方法で、Ｒ２の”ｍｏｕｓｅ”の意味が動物の「ネズミ」の意味（図１０での（０５））であることが分かる。
【０１５７】
ここで、規則記憶部１３２には以下の翻訳規則が存在したとする。
【０１５８】
Ｒ２：　（ｋｅｅｐ　（ＯＢＪ　ｍｏｕｓｅ））　→　（飼う　（を　ネズミ））
Ｒ３：　（ｋｅｅｐ　（ＯＢＪ　ｋｎｉｆｅ））　→　（持つ　（を　ナイフ））
【０１５９】
また入力表現として、以下が与えられたとする。
【０１６０】
Ｉ２：　Ｉ　ｋｅｅｐ　ａ　ｋｅｙｂｏａｒｄ
【０１６１】
入力部１３１から英語表現Ｉ２が入力されると、類似規則特定部１３６が規則記憶部１３２から入力に類似した原言語表現に対して適用可能な翻訳規則のうち、最も類似度の高い翻訳規則を特定する。この際、上記の規則Ｒ１中の原言語表現の”ｍｏｕｓｅ”は動物のネズミの意味（図１０の（０５））であることが分かっており、”ｍｏｕｓｅ”（０５）と”ｋｅｙｂｏａｒｄ”（０９）の距離は、”ｋｎｉｆｅ”（０８）と”ｋｅｙｂｏａｒｄ”（０９）の距離より遠くなる。
【０１６２】
よって、
誤ってＲ２が利用されて、
「私はキーボードを飼っている」とは翻訳はされず、
Ｒ３が利用されて、
「私はキーボードを持っている」
と翻訳される。
【０１６３】
また、入力表現として、以下が与えられたとする。
【０１６４】
Ｉ１：　Ｉ　ｓｅｒｖｅ　ａ　ｄｉｓｈ
【０１６５】
入力部１３１から上記英語表現が入力されたとき、類似規則特定部１３６は上記の方法と同様に以下の規則を特定する。
【０１６６】
Ｒ４：　（ｓｅｒｖｅ　（ＯＢＪ　ｐｉｅ））　→　（出す　（を　パイ））
【０１６７】
規則適用部１３７１において、規則Ｒ３から「ｓｅｒｖｅ　〜」の部分に関しては、「〜を出す」と訳が定まる。
【０１６８】
”ｄｉｓｈ”の訳語として、「皿」「料理」の２つの訳語があった場合、追加訳語選択部１３７２において、第１の実施の形態における追加訳語選択部１６２（図１参照）と同様にして、訳語の選択を行う。訳語として、「料理」を選択する。
【０１６９】
更に、別の翻訳規則や単語辞書から、「Ｉ」の訳を「私は」と定める。この結果、出力部１３８から、
「私は料理を出す」
という翻訳結果が出力される。
【０１７０】
本実施例において、翻訳規則の適用条件中の原言語表現の意味の特定は、入力表現が与えられた後に行っても、あるいは、予め行っておいても良い。また、本実施例において、規則意味特定部１３５と追加訳語選択部１３７２のいずれか一方のみを持つ構成としても良い。
【０１７１】
以上説明したように、本実施例においては、入力に対し適用可能な翻訳規則がなく、類似の翻訳規則で代用することを考えた場合に、目的言語側の意味距離計算を行って予め各規則の適用条件中の原言語表現の意味の曖昧性を解消しておくことで、誤った翻訳規則が利用されることを抑制し、また、利用された元の翻訳規則から置き換えられた部分の原言語表現の訳語をより正確に選択できるようになる、という作用効果を持つ。
【０１７２】
以上本発明を上記各実施の形態に即して説明したが、本発明は、上記実施の形態の構成にのみ限定されるものではなく、特許請求の範囲の各請求項の発明の範囲内で当業者であればなし得るであろう各種変形、修正を含むことは勿論である。
【０１７３】
【発明の効果】
以上説明したように、本発明によれば、［従来の技術］で説明した上記第一の従来例では正しく決定されなかった可能性のある、対訳表現からは直接に訳語が決定されない入力表現中の一部分に関して、目的言語の意味距離を用いた制約を設けることで、より正確に訳語の決定を行うことができる、という効果を奏する。
【０１７４】
本発明によれば、目的言語側の意味距離計算を行って、予め各対訳表現の原言語表現の意味の曖昧性を解消しておくことにより、上記第一の従来例では正しく類似対訳表現が特定できなかった場合に関しても、より正確に、類似対訳表現を特定することができる、という効果を奏する。
【０１７５】
本発明によれば、翻訳実行時または翻訳実行前に前処理として、翻訳規則の類似規則を自動生成する場合に、目的言語側の意味距離計算を行って、予め各規則の原言語表現の意味の曖昧性を解消しておくことにより、誤った類似規則が生成されることを抑制し、類似規則生成時に元となる翻訳規則から置き換えられた原言語表現の訳語を、より正確に選択することができる、という効果を奏する。
【０１７６】
本発明によれば、入力に対し適用可能な翻訳規則がなく、類似の翻訳規則で代用する場合、目的言語側の意味距離計算を行って予め各規則の適用条件中の原言語表現の意味の曖昧性を解消しておくことで、誤った翻訳規則が利用されることを抑制し、利用された元の翻訳規則から置き換えられた部分の原言語表現の訳語をより正確に選択することができる、という効果を奏する。
【図面の簡単な説明】
【図１】本発明の第１の実施の形態の構成を示す図である。
【図２】本発明の第１の実施の形態における機械翻訳の処理手順を示すフローチャートである。
【図３】本発明の第２の実施の形態の構成を示す図である。
【図４】本発明の第２の実施の形態における機械翻訳の処理手順を示すフローチャートである。
【図５】図４のステップ４０１（サブルーチン）の処理手順を示すフローチャートである。
【図６】本発明の第３の実施の形態をなす類似規則生成装置の構成を示す図である。
【図７】翻訳規則の一例を示す図である。
【図８】翻訳規則からの類似規則生成の一例を示す図である。
【図９】本発明の第３の実施の形態における類似規則生成の処理手順を示すフローチャートである。
【図１０】英語から日本語への翻訳時の原言語の英語シソーラスの一例を示す図である。
【図１１】英語から日本語への翻訳時の目的言語の日本語シソーラスの一例を示す図である。
【図１２】本発明の第４の実施形態の構成を示す図である。
【図１３】本発明の第４の実施の形態における機械翻訳の処理手順を示すフローチャートである。
【図１４】従来の機械翻訳装置の構成を示す図である。
【符号の説明】
１１　入力部
１２　対訳表現記憶部
１３　原言語意味距離計算部
１４　類似対訳表現特定部
１５　目的言語意味距離計算部
１６　翻訳部
１６１　対訳表現適用部
１６２　追加訳語選択部
１７　出力部[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention translates a sentence in an input source language, a source language, into a sentence in a target language, a target language, and outputs it. The present invention relates to a translation apparatus that mechanically executes translation of a phrase from one language to another, such as an inter-language search system for searching.
[0002]
[Prior art]
Conventionally, as a machine translation method that takes an expression in the source language as input, mechanically translates it into an expression in the target language, and outputs it, the result of semantic analysis such as morphological analysis and syntax analysis of the source language expression 2. Description of the Related Art There is known a rule-based translation method (hereinafter, referred to as a “first conventional example”) that accumulates translation rules defining a method and translates an input sentence based on the translation rules.
[0003]
To briefly explain the operation of the first conventional example, when an input sentence of the source language is input, first, morphological analysis, language analysis such as syntax analysis is performed, and then, the result of the semantic analysis is The translation rules are applied and replaced with the target language.
[0004]
Finally, a replacement result is generated and output as a sentence in the target language.
[0005]
However, since the applicable range of a translation rule is limited by the applicable conditions of the translation rule, a large number of translation rules covering the input expression are required to comprehensively correspond to the input expression. Was.
[0006]
Therefore, as a machine translation method, there is an example in which a source language expression and a set of translations of the target language expression corresponding to the source language expression, which are created by hand or the like in advance, are stored as an example, and the input completely matches the input. In the case where the translation is performed, the corresponding translation is output, and even if there is no example that exactly matches the input, an example similar to the input is specified, and the input is translated with reference to the specified example. Has been proposed.
[0007]
As a publication related to this type of conventional translation system, for example, Japanese Patent No. 2609173 (hereinafter, referred to as “second conventional example”) is referred to. FIG. 14 is a diagram showing the configuration of the second conventional example. Referring to FIG. 14, in a second conventional example, an input unit 121 for inputting an input expression of a source language, an example storage unit 122 storing a set of a source language expression and a target language expression as a translation thereof, A source language semantic distance calculation unit 123 that calculates a semantic distance between linguistic expressions, a similar example specifying unit 124 that uses the source language semantic distance calculation unit 123 to specify an example whose semantic distance is closest to the input as a similar example, A translation unit 125 that performs translation using the specified similar example and an output unit 126 that outputs the translated target language expression are provided.
[0008]
Next, the operation of the second conventional example will be described with reference to FIG. Hereinafter, a case where a Japanese sentence is translated into English will be described.
[0009]
It is assumed that “meeting in Tokyo” is input to the input unit 121.
[0010]
Then, the similar example specifying unit 124 refers to the source language semantic distance calculation unit (source language thesaurus) 123 and reads from the example storage unit 122 a source language expression similar to the input sentence (“meeting in Tokyo”). Select an example.
[0011]
For example, in an example,
“Stay in Kyoto → stay in Kyoto” (similarity to input 0.4),
"Meeting in Kyoto → conference in Kyoto" (similarity to input 1.0)
If there is
More similar
"Conference in Kyoto → Conference in Kyoto"
Is selected.
[0012]
Next, using the translation “conference in Kyoto” for “meeting in Kyoto”, “conference in Tokyo” is output as a translation for the input.
[0013]
[Problems to be solved by the invention]
However, in the second conventional example, since the similar example is referred to at the time of translation, the following translation error may occur.
[0014]
In the second conventional example, an erroneous translation may be selected for a part of the input expression for which a translation is not determined from the similar example. First, it is assumed that the following example exists.
[0015]
E0: I serve the master → I serve my master
E1: I serve a pie → I serve pie
[0016]
Further, it is assumed that the translations of “dish” include “dish” (first translation) and “cooking” (second translation).
[0017]
Here, a case where the following English expression is translated into Japanese will be described.
[0018]
I1: I serve a dish.
[0019]
In the second conventional example, an example for a source language expression having the closest meaning distance to an input is obtained using a source language thesaurus. The thesaurus is a word that is semantically classified into a tree structure, and a specific example is shown in FIG. The semantic distance between words using the thesaurus is approximately the distance between nodes to which each word for which the semantic distance is to be determined belongs on the thesaurus.
[0020]
Here, referring to the thesaurus of FIG. 10 for the calculation of the semantic distance, the “dish” of (01) is larger than the “dish” of (01) or (02) and the “master” of (04). Since the meaning distance between “pie” and (03) is shorter, E1 is selected as a reference example.
[0021]
Next, with respect to the portion of "I reserved ~" of the input, a translation can be obtained from the example E1 as "I will issue ~". When the translation of "dish" having a plurality of translations is selected, the example E1 is used. Is not used. For example, if the first translation is used for this translation selection, "dish" is selected as the translation for "dish", and the translation is incorrect, such as "I will put out the plate". There is.
[0022]
Further, in the second conventional example, there is also a problem that an erroneous example may be referred to when the source language expression of the example has a plurality of possible meanings.
[0023]
First, assume that the following example exists,
[0024]
E2: I keep a mouse → I have a mouse
E3: I keep a knife → I have a knife
[0025]
It is also assumed that “mouse” has the meaning of a mouse (animal) and the meaning of a mouse (personal computer peripheral device).
[0026]
Here, a case where the following English expression is translated into Japanese will be described.
[0027]
I2: I keep a keyboard.
[0028]
Referring to the source language thesaurus shown in FIG. 10, "mouse" appears at a plurality of positions of an animal meaning position (05) and a personal computer peripheral device meaning position (06).
[0029]
In the second conventional example, the "mouse" (06) and the "keyboard" (09) as personal computer peripherals are longer than the distance between any of "mouse" and "knife" (08) in (05) and (06). ), The example E2 is erroneously determined to be semantically close to the input, and the input is translated as “I have a keyboard”.
[0030]
On the other hand, as another approach to the problem of the first conventional example, a method of automatically generating a semantically similar translation rule from the translation rule can be considered. However, in this method, similar to the second conventional example, if a similar rule is simply generated using the semantic distance of the source language, an erroneous similar rule may be generated. For example, consider generating a similar rule from the following translation rule from English to Japanese.
[0031]
R1: (feed (OBJ mouse)) → (feeding (in rats))
[0032]
When the word “mouse” exists in the slot of the OBJECT (object) of the word “feed”, R1 converts “feed” to “feed” and “mouse” to “mouse”, and converts the word “mouse” to “mouse”. The translation rule is to attach the particle “ni” to “rat”.
[0033]
Referring to the thesaurus of FIG. 10, "mouse" has the meaning of an animal (05) and the meaning of a personal computer peripheral device (06), and there is a synonym for each meaning. If a similar rule is generated using a synonym of the meaning (06) of the personal computer peripheral device, the following erroneous rule will be generated (“×” indicates impossible (NG)).
[0034]
× (feed (OBJ keyboard)) → (feeding (keyboard))
× (feed (OBJ modem)) → (feeding (modem))
[0035]
Furthermore, even if a similar rule is correctly generated with the meaning of an animal (05), it is necessary to select a translation for the word to which "mouse" is replaced in the original translation rule R1, but the translation may fail to be selected. There is.
[0036]
A case in which "mouse" in rule R1 is replaced with "seal" (seal) of an animal in the following manner will be described. "???" indicates that the translation is undecided.
R1 ': (feed (OBJ seal)) → (Feed (Ni ???))
[0037]
Here, "seal" has two translations of "seal" (first translation) and "seal" (second translation). If, for example, the first translation is selected for selecting the translation of "seal" of the similar rule R1 ', the following erroneous similar rule is generated.
[0038]
× (feed (OBJ seal)) → (feeding (sealed seal))
[0039]
Accordingly, the present invention has been made in view of the above problems, and a main object of the present invention is to provide a translation device, a method, a program, and a recording medium that can perform translation with higher accuracy. is there.
[0040]
[Means for Solving the Problems]
The present invention, which provides means for solving the above problems, is characterized by using a semantic distance of a target language in a translation process of inputting a source language expression, converting the expression into a target language expression, and outputting the converted expression.
[0041]
An apparatus according to one aspect of the present invention inputs a source language expression (abbreviated as “source language expression”), converts it into a target language expression (abbreviated as “target language expression”), and outputs it. A translation device, wherein a bilingual expression storage means storing a bilingual expression comprising a set of a source language expression and a target language expression corresponding to the source language expression, and a source language semantic distance calculating means for calculating a semantic distance between the source language expressions. A target language semantic distance calculating means for calculating a semantic distance between target language expressions, a similar bilingual expression specifying means for controlling the source language semantic distance calculating means to specify a bilingual expression most similar to the input expression, Using the selected bilingual expression, the bilingual expression applying means for translating the input expression and the target language semantic distance calculating means, and selecting a translated word of a part where the translated word is not determined by the bilingual expression applying means Translation It includes a selecting means.
[0042]
According to another aspect of the present invention, there is provided a translation apparatus that inputs a source language expression, converts the source language expression into a target language expression, and outputs the target language expression. A target language semantic distance calculating means for calculating a semantic distance between the expressions, and, when a target language expression corresponding to the source language expression is given, using the source language semantic distance calculating means and the target language semantic distance calculating means. And meaning specifying means for resolving the ambiguity of the meaning of the source language expression.
[0043]
According to another aspect of the present invention, there is provided a translation apparatus for converting a source language expression into a target language expression, comprising: a bilingual expression storage unit storing a bilingual expression including a set of the target language expression corresponding to the source language expression; From a source language semantic distance calculating means for calculating a semantic distance between linguistic expressions, a target language semantic distance calculating means for calculating a semantic distance between the target linguistic expressions, and a target language expression corresponding to the source language expression in the bilingual expression, Using the source language semantic distance calculating means and the target language semantic distance calculating means, controlling the rule semantic specifying means for resolving the ambiguity of the meaning of the source language expression of the bilingual expression, and controlling the source language semantic distance calculating means. A bilingual expression specifying unit that specifies a bilingual expression most similar to the input expression based on the meaning of the source language expression of the bilingual expression specified by the rule semantic specifying unit; Using a constant has been translated expression, and a, a translation means for performing translation of the input expression.
[0044]
According to another aspect of the present invention, there is provided a translation apparatus that inputs a source language expression, converts the source language expression into a target language expression, and outputs the target language expression. The translation device includes a source language expression and a target language expression corresponding to the source language expression. A bilingual expression storage unit that stores a bilingual expression, a source language semantic distance calculating unit that calculates a semantic distance between source language expressions, a target language semantic distance calculating unit that calculates a semantic distance between target language expressions, A bilingual expression semantic identifying unit for controlling the linguistic semantic distance calculating unit and the target language semantic distance calculating unit to eliminate ambiguity of the meaning of the source language expression of each bilingual expression in the bilingual expression storage unit; A similar bilingual expression specifying means for controlling the semantic distance calculating means to specify the bilingual expression most similar to the input expression, a translating means for translating the input expression based on the specified bilingual expression, and a converted target language table And a, and output means for outputting.
[0045]
In one aspect, the method according to the present invention solves the above-described problem. In one aspect, a translation process of inputting an expression in a source language, converting the expression into an expression in a target language, and outputting the translated expression is performed using a computer including an input unit and an output unit. In the translation method to be performed, a step of inputting a source language expression from the input unit and a process of calculating a semantic distance between the source language expressions are controlled so that a source language expression and a target language corresponding to the source language expression are controlled. Specifying a bilingual expression for a source language expression that is most similar to the input expression from a bilingual expression storage unit storing a bilingual expression composed of a set of expressions, and inputting the bilingual expression using the specified bilingual expression. Translating the expression in the source language, and controlling the process of calculating the semantic distance between the target language expressions, and controlling the translation of the words in the input expression for which the translation was not determined in the translating step. And determining a, and outputting the translation result from the output unit.
[0046]
A program according to the present invention that solves the above-described problem causes a computer including an input unit and an output unit to execute a translation process of inputting a source language expression, converting the source language expression into a target language expression, and outputting the target language expression. A program for inputting a source language expression from the input unit, a process for calculating a semantic distance between the source language expressions, a process for calculating a semantic distance between the target language expressions, By controlling the process of calculating the distance, the most similar to the input expression is obtained from a bilingual expression storage unit storing a bilingual expression comprising a set of a source language expression and a target language expression corresponding to the source language expression. Specifying the bilingual expression for the source language expression to be translated, translating the input expression using the specified bilingual expression, and calculating the semantic distance between the target language expressions And causing the computer to execute a process of determining a translation of a word in the input expression whose translation has not been determined in the process of translating the input expression, and a process of outputting a translation result from the output unit. Consists of a program. The program is stored on a recording medium.
[0047]
The present invention provides a mechanism for resolving the ambiguity of the meaning of the source language expression of the bilingual expression by calculating the semantic distance between the target language expressions and specifying the similar bilingual expression more correctly. It is characterized by having a mechanism for selecting a translation of the input sentence so as to be semantically close to the given translation. As a result, an operation and effect that translation processing can be performed with higher accuracy can be obtained.
[0048]
BEST MODE FOR CARRYING OUT THE INVENTION
An embodiment of the present invention will be described with reference to the drawings.
[0049]
[Embodiment 1]
A first embodiment of the present invention will be described. FIG. 1 is a diagram showing a configuration of the first exemplary embodiment of the present invention. Referring to FIG. 1, an apparatus according to a first embodiment of the present invention includes an input unit 11 for inputting an input sentence in a source language, a source language expression (hereinafter simply referred to as “source language expression”) and a source language expression. A bilingual expression storage unit 12 that stores a pair of corresponding target language expressions (hereinafter simply referred to as “target language expression”), a source language semantic distance calculation unit 13 that calculates a semantic distance between source language expressions, A similar bilingual expression specifying unit 14 for specifying a bilingual expression for the source language expression most similar to the input using the distance calculating unit 13; a target language semantic distance calculating unit 15 for calculating a semantic distance between the target language expressions; A translation unit 16 for translating the input sentence based on the translated expression, and an output unit 17 for outputting the converted target language expression.
[0050]
The translation unit 16 uses the bilingual expression application unit 161 that determines a translation that can be determined from the specified bilingual expression, and the target language semantic distance calculation unit 15 that translates the input word that is not determined by the bilingual expression application unit. And an additional translation word selection unit 162 to be determined.
[0051]
In the above configuration, regarding the input unit 11, the bilingual expression storage unit 12, the source language semantic distance calculation unit 13, and the similar bilingual expression specification unit 14, the input unit 121, the example storage unit 122 used in the first conventional example, A known unit having the same configuration as the source language semantic distance calculation unit 123 and the similar example specifying unit 124 is used. The bilingual expression application section 161 has a known configuration similar to that of a conventional translation section.
[0052]
The target language semantic distance calculator 15 calculates the semantic distance between words in the target language.
[0053]
The additional translation word selection unit 162 uses the target language semantic distance calculation unit 15 to select a translation of a word in the input expression whose translation has not been determined by the bilingual expression application unit 161. The additional translation word selection unit 162 selects, for example, a translation assigned to the word in the bilingual expression corresponding to the word and a translation word having a short meaning distance of the target language, from among the translation word candidates of the word.
[0054]
Next, FIG. 2 is a flowchart for explaining the operation of the first embodiment of the present invention. The operation of the first embodiment of the present invention will be described with reference to FIGS.
[0055]
First, when an input expression is input to the input unit 11 (step 201), the similar bilingual expression specifying unit 14 controls the source language semantic distance calculating unit 13 and the source language most similar to the input expression from the bilingual expression storage unit 12. A bilingual expression for the expression is specified (step 202).
[0056]
Next, the bilingual expression application unit 161 translates the input expression in the source language using the specified bilingual expression (step 203).
[0057]
The additional translation word selection unit 162 controls the target language semantic distance calculation unit 15, and the bilingual expression application unit 161 determines the translation of the word in the input expression of the source language, whose translation has not been determined (step 204).
[0058]
Finally, the translation result is output from the output unit (Step 205).
[0059]
In this embodiment, the processing of the source language semantic distance calculating unit 13, the similar bilingual expression specifying unit 14, the target language semantic distance calculating unit 15, the translating unit 16, the input / output of the input unit 11 and the output unit 17 The control is preferably realized by a program executed by a computer. In this case, the computer reads the program from a machine-readable storage medium that stores the program and loads the program into the main storage of the computer and executes the program, or downloads the program to the computer via a network. By executing, the apparatus according to the first embodiment of the present invention is realized. The same applies to the following embodiments.
[0060]
[Example 1]
Next, in order to describe the above-described first embodiment of the present invention in more detail, the operation of the first embodiment will be described with reference to a specific example (example). The configuration of the present embodiment is the same as that of the first embodiment. In the following, the translation from English to Japanese will be described.
[0061]
It is assumed that the following bilingual expressions exist in the bilingual expression storage unit 12 of FIG.
[0062]
E0: I serve the master → I serve my master
E1: I serve a pie → I serve pie
[0063]
Also, as a translation of "dish",
"Dish" (first translation),
"Cooking" (second translation)
Suppose there is.
[0064]
Here, it is assumed that the following is given as an input expression.
[0065]
I1: I serve a dish.
[0066]
When the English expression I1 is input from the input unit 11, the similar bilingual expression specifying unit 14 controls the source language semantic distance calculating unit 13 and, from the bilingual expression storage unit 12, the source expression that is semantically close to the input expression I1. Specify the bilingual expression for the language.
[0067]
Referring to the thesaurus illustrated in FIG. 10, the distance between “dish” (01) and “pie” (03) is greater than the distance between any of “dish” and “master” (04). Are close, E1 is selected as the bilingual expression to be referred to.
[0068]
Next, in the bilingual expression application unit 161, the part “I serve” in the input expression is referred to as the bilingual expression E1 in the same manner as in the conventional translating apparatus, and is translated as “I will output ~”. Is specified.
[0069]
Next, from the bilingual expression E1, a translation of "dish" of the word in the input expression for which no translation is determined is selected. “Dish” has “dish” and “dish” as translated words, but the additional translated word selection unit 162 controls the target language semantic distance calculation unit 15 to translate the translated expression E1 corresponding to “dish” of the input expression I1. The translation distance between the translated words “pie” and “dish” for the word “pie” in the middle is calculated.
[0070]
Referring to the thesaurus of FIG. 11, since the distance between “pie” (01) and the dish (02) is shorter than the distance between “pie” (01) and the plate (03), the translated word “dish” is selected. Finally, the translation unit generates the following as a translation for the input using the specified bilingual expression and the selected translation of “dish”, and outputs the translation from the output unit 17.
I serve.
[0071]
In this embodiment, a thesaurus is used for the source language semantic distance calculation unit 13 and the target language semantic distance calculation unit 15, but linguistic information obtained from a large-scale corpus, such as simple word co-occurrence frequency information or mutual information amount, etc. Statistical combination information, combinations thereof, and other language information may be used.
[0072]
The bilingual expression may be a set of sentences, phrases, and words in the source language and corresponding sentences, phrases, and words in the target language, as in the example of the first conventional example described in [Prior Art]. Then, the translation rule in the second conventional example described in [Prior Art] may be used, or a result of performing some semantic analysis on the source language expression and some semantic analysis on the corresponding target language expression May be a set of the results of the above.
[0073]
As described above, in the present embodiment, the meaning of the target language is used for a part of the input expression in which the translation is not directly determined from the bilingual expression, which may not have been correctly determined in the first conventional example. Due to the restriction using the distance, the translation word can be determined more accurately.
[0074]
[Embodiment 2]
A second embodiment of the present invention will be described with reference to FIG. In the second embodiment of the present invention, an input unit 31 for inputting an input expression of a source language, a bilingual expression storage unit 32 storing a pair of the source language expression and a target language expression as a translation thereof, Source language semantic distance calculation unit 33 that calculates the semantic distance between them, target language semantic distance calculation unit 34 that calculates the semantic distance between the target language expressions, source language semantic distance calculation unit 33, and target language semantic distance calculation unit 34 And a bilingual expression meaning specifying unit 35 for resolving the ambiguity of the meaning of the source language expression of each bilingual expression in the bilingual expression storage unit 32, based on the meaning of the specified source language expression in each bilingual expression. A similar bilingual expression specifying unit 36 for specifying a bilingual expression corresponding to the source language expression most similar to the input, a translating unit 37 for translating the input expression based on the specified bilingual expression, and outputting the converted target language expression Output unit 38 that performs
[0075]
In the above configuration, the input unit 31, the bilingual expression storage unit 32, the source language semantic distance calculating unit 33, the similar bilingual expression specifying unit 36, and the output unit 38 are the input unit 11, the bilingual expression in the embodiment of the first invention. It is the same as the storage unit 12, the source language semantic distance calculation unit 13, the similar bilingual expression specification unit 14, and the output unit 17. The translation unit 37 may include a bilingual expression application unit (161 in FIG. 1) and an additional translation word selection unit (162 in FIG. 1), as in the first embodiment of the present invention. Alternatively, the translation unit 37 may be the same as the translation unit of the first conventional example described above.
[0076]
When the source language expression has ambiguity in the source language expression, the source language semantic distance calculating means 33 calculates the source language semantic distance from the source language expression. Then, the meaning is specified using the target language meaning distance calculating means 34.
[0077]
Next, FIGS. 4 and 5 are flowcharts for explaining the operation of the second embodiment of the present invention. The operation of the second embodiment of the present invention will be described with reference to FIGS.
[0078]
First, for each bilingual expression in the bilingual expression storage unit 32 in advance, if the source language expression has an ambiguity in the meaning, the bilingual expression meaning specifying unit 35 causes the source language semantic distance calculating unit 33 and the target language semantic distance calculating unit. 34, the ambiguity of the meaning is eliminated from the translation given to the source language expression (step 401).
[0079]
FIG. 5 is a flowchart for explaining the details of step 401.
[0080]
For each bilingual expression (step 501), a word WE having a plurality of meanings in the source language expression is specified (step 502), and the word WE for each meaning mi of the word WE is determined using the source language semantic distance calculation unit 33. Synonyms are obtained, and a set S1i is created (step 503).
[0081]
Next, for each i, a translation of each word in S1i is obtained, and a set S2i is created (step 504).
[0082]
Next, the semantic distance between the translated word WJ given to the word WE in the bilingual expression of interest and the set S2i is determined (step 505), and the meaning corresponding to the set closest to WJ is the word in this bilingual expression. It means WE.
[0083]
Thereafter, when an input expression is input to the input unit (step 402), the similar bilingual expression specifying unit 36 controls the source language semantic distance calculation unit 33 to output the source expression most similar to the input expression from the bilingual expression storage unit 32. A bilingual expression for the language is specified (step 403).
[0084]
Here, since the ambiguity of the meaning of the source language expression of the bilingual expression has been eliminated by the bilingual expression meaning specifying unit in advance, a bilingual expression that is semantically different from the input expression is not specified.
[0085]
Next, the input expression is translated from the bilingual expression specified in step 403 (step 404).
[0086]
Finally, the translation result is output from the output unit (step 405).
[0087]
In this embodiment, each processing of the source language semantic distance calculation unit 33, the target language semantic distance calculation unit 34, the bilingual expression meaning specifying unit 35, the similar bilingual expression specifying unit 36, the translating unit 37, and the input unit The processing control of the output unit 31 and the output unit 38 is preferably realized by a program executed by a computer.
[0088]
[Example 2]
Next, a description will be given with reference to a specific operation example (Example 2). Hereinafter, the translation from English to Japanese in the second embodiment will be described.
[0089]
In the second embodiment of the present invention, the ambiguity of the meaning of the source language expression of each bilingual expression in the bilingual expression storage unit 32 of FIG. Hereinafter, it is assumed that the thesaurus illustrated in FIG. 10 is used for the processing in the source language semantic distance calculation unit 33, and the thesaurus illustrated in FIG. 11 is used for the processing in the target language semantic distance calculation unit 34.
[0090]
As an example, the following parallel expression is given.
[0091]
E2: I keep a mouse → I have a mouse.
[0092]
"Mouse" has the meaning of an animal mouse (m1, (05) in FIG. 10) and the meaning of a mouse of a personal computer peripheral device (m2, (06) in FIG. 10) (step 502).
[0093]
Using the source language semantic distance calculation unit, synonyms of these two meanings are obtained, and a set S1i is created (step 503).
[0094]
S11: (dog, cat, cow, rat)
S12: (keyboard, display, modem)
[0095]
Next, for each S1i, a translation of each word in S1i is obtained, and a set S2i is created (step 504).
[0096]
S21: (dog, cat, cow, rat)
S22: (Keyboard, display, modem)
[0097]
Finally, the semantic distance between the translation "rat" assigned to "mouse" in the bilingual expression and each of the sets S21 and S22 is determined (step 505).
[0098]
If the semantic distance between the target languages W1 and W2 is expressed as dist (W1, W2), the semantic distance between "mouse" and S21 and S22 is as follows using the thesaurus of FIG.
[0099]

[0100]

[0101]
As a result, since the distance from S21 is shorter, the meaning (04) of the animal rat, which is the meaning corresponding to S21, becomes the meaning of "mouse" in the bilingual expression.
[0102]
Here, it is assumed that the following bilingual expressions exist in the example storage unit 32.
[0103]
E2: I keep a mouse → I have a mouse
E3: I keep a knife → I have a knife
[0104]
It is also assumed that the following is given as an input expression.
[0105]
I2: I keep a keyboard.
[0106]
When the English expression I2 is input from the input unit 31, the similar bilingual expression specifying unit 36 specifies the bilingual expression for the source language most similar to the input from the bilingual expression storage unit 32.
[0107]
At this time, it is known that "mouse" of the source language expression in the bilingual expression E2 has the meaning of an animal mouse ((05) in FIG. 10), and "keyboard" (09) and "mouse" Since the distance of “mouse” (05) is longer than the distance of “keyboard” (09) and “knife” (08), the bilingual expression E3 is selected as the bilingual expression to be referred to unlike the first conventional example. Therefore, the translation of I2 can be translated as "I have a knife" by referring to the bilingual expression E3 without incorrectly translating "I have a keyboard" by referring to the bilingual expression E2. become.
[0108]
The specification of the meaning of the bilingual expression may be performed after the input expression is given, or may be performed in advance.
[0109]
As described above, in the present embodiment, by performing the semantic distance calculation on the target language side and previously eliminating the ambiguity of the meaning of the source language expression of each bilingual expression, the first conventional example correctly Even when the similar bilingual expression cannot be specified, there is an operational effect that the similar bilingual expression can be specified more accurately.
[0110]
[Embodiment 3]
The present invention is also applicable to a case where a similar rule is generated in advance from a translation rule in a rule-based translation method to increase the coverage of a translation rule for an input expression.
[0111]
FIG. 6 is a diagram showing a configuration of the third exemplary embodiment of the present invention. The third embodiment of the present invention will be described with reference to FIG. In the third embodiment of the present invention, a translation rule input unit 61 for inputting a translation rule, a source language semantic distance calculation unit 62 for calculating a semantic distance between source language expressions, and a semantic distance between target language expressions are set. Utilizing the target language semantic distance calculation unit 63, the source language semantic distance calculation unit 62, and the target language semantic distance calculation unit 63 to resolve the ambiguity of the meaning of the source language expression of the application condition of the input translation rule A rule meaning specifying unit 64 for obtaining a similar source language expression based on the specified meaning of the source language expression, and generating a similar rule by replacing the input translation rule with the similar source language expression. A translation unit for selecting a translated word for the replaced source language expression, and an output unit 67 for outputting the generated similarity rule.
[0112]
In the above configuration, the source language semantic distance calculator 62 and the target language semantic distance calculator 63 are the same as the source language semantic distance calculator 33 and the target language semantic distance calculator 34 in the second embodiment of the present invention.
[0113]
When the input source language expression has a specific structure, the translation rule input unit 61 inputs a translation rule defining how to translate the structure into the target language.
[0114]
FIG. 7 shows an example of a translation rule when the source language is English and the target language is Japanese.
[0115]
The rule meaning specifying unit 64 has a function similar to that of the bilingual expression meaning specifying unit 35 according to the second embodiment of the present invention. If the source language expression in the application condition of the input translation rule has a meaning ambiguity, From the translation given to the source language expression by the translation rule, the meaning is specified using the source language semantic distance calculation means 62 and the target language semantic distance calculation means 63.
[0116]
The similar rule generation unit 65 generates a rule in which the source language expression of the application condition is replaced with a similar expression based on the meaning of the source language specified by the rule meaning specification unit 64.
[0117]
FIG. 8 shows an example of the replacement rule.
[0118]
The rule translation word selection section 66 has the same function as the additional translation word selection section 162 in the first embodiment of the present invention, and converts the translation of the source language expression in the application condition of the replaced similar rule into the target language meaning. The selection is made using the distance calculation unit 63. Here, in the same manner as the additional translation word selection unit 162, of the translations of the source language expression, the translation word assigned to the expression in the application condition of the similar rule generation source translation rule corresponding to the expression and the target language meaning Select a translation that is closer in distance.
[0119]
Next, FIG. 9 is a flowchart for explaining the operation of the third embodiment of the present invention. The operation of the present embodiment will be described with reference to FIG.
[0120]
First, a translation rule is input from the translation rule input unit 61 (step 901). If the source language expression of the application condition of the input translation rule has an ambiguity in the meaning, the rule meaning specifying unit 64 specifies the source language meaning. Using the distance calculation unit 62 and the target language meaning distance calculation unit 63, the ambiguity of the meaning is eliminated from the translation given to the expression of the source language by the translation rule (step 902).
[0121]
The details of step 902 for resolving the ambiguity of the meaning are the same as those in the flowchart of FIG.
[0122]
Next, the similar rule generation unit 65 uses the source language semantic distance calculation unit 62 to convert the source language expression in the application condition of the input translation rule to a source language expression that is semantically similar. Generate and output (step 903).
[0123]
The rule translation word selection unit 66 receives the similarity rule, and determines a translation word for the replaced source language expression in the similarity rule using the target language semantic distance calculation unit 63 (step 904).
[0124]
Finally, the translation result is output from the output unit 67 (step 905).
[0125]
In this embodiment, a translation rule input unit 61, a source language semantic distance calculation unit 62, a target language semantic distance calculation unit 63, a rule meaning specification unit 64, a similar rule generation unit 65, and a rule translation word selection unit Each process of 66 and the process control of the output unit 67 are preferably realized by a program executed by a computer.
[0126]
[Example 3]
Next, a specific operation example (third embodiment) will be described. In the third embodiment, in the third embodiment, a translation rule from English to Japanese is input, and the structure of the translation rule in FIG. 8 is simply represented as follows. .
[0127]
R1: (feed (OBJ mouse)) → (feeding (in rats))
[0128]
When R1 is input to the translation rule input unit 61 in FIG. 6, first, the source language expression “mouse” in the application conditions of R1 has ambiguity of meaning of mouse (personal computer peripheral device) and mouse (animal). The rule meaning specifying unit 64 resolves ambiguity in the same manner as the translation expression meaning specifying unit 35 in FIG. 3, and the meaning of “mouse” in this translation rule is the meaning of the animal rat ((05) in FIG. 10). Is specified. Next, a similar expression of "mouse" of the meaning of the animal is obtained using the source language semantic distance calculation unit 62, and the source language expression in the application condition of the input translation rule is replaced.
[0129]
(Feed (OBJ mouse)) → (feeding (in rats))
⇒ (feed (OBJ cat)) → (Feed (Ni ???))
⇒ (feed (OBJ pig)) → (Feed (Ni ???))
⇒ (feed (OBJ seal)) → (Feed (Ni ???))
...
[0130]
At this time, by specifying the meaning by the rule meaning specifying unit 64 in FIG. 6, the following erroneous similarity is obtained by using a mouse ((06) in FIG. 10) of a personal computer peripheral device, which is another meaning of “mouse”. Prevents rules from being generated.
[0131]
× (feed (OBJ keyboard)) → (Feed (Ni ???))
× (feed (OBJ modem)) → (Feed (Ni ???))
[0132]
Next, a translated word corresponding to the replaced source language expression in the application conditions of the translation rule is selected. The selection method is similar to the additional translation word selection unit 162 of the second embodiment of the invention, in which the translation word (“rat”) given to the replacement unit (“mouse”) of the original translation rule and the semantic distance of the target language are different. Select a translation that is closer.
[0133]
For example, considering the similar rule replaced with "seal", "seal" has two translations, "seal" and "seal". For example, referring to the thesaurus in FIG. 11, a mouse (04) and a seal ( Since the distance between the mouse (04) and the seal (05) is shorter than the distance at 06), "seal" is selected as a translation.
[0134]
(Feed (OBJ mouse)) → (feeding (in rats))
⇒ (feed (OBJ cat)) → (feeding (cat))
⇒ (feed (OBJ pig)) → (Feed (Pig)
⇒ (feed (OBJ seal)) → (Feed food (to seal))
...
[0135]
In this embodiment, a configuration having only one of the rule meaning specifying unit 64 and the rule translation word selecting unit 66 may be employed.
[0136]
As described above, in the present embodiment, when a rule similar to a translation rule is automatically generated as a pre-process at the time of performing translation or before performing translation, the semantic distance calculation on the target language side is performed and the source of each rule is determined in advance. By eliminating the ambiguity of the meaning of the linguistic expressions, it is possible to suppress the generation of erroneous similar rules, and to improve the translation of the source language expressions that have been replaced from the original translation rules when the similar rules were generated. It has the effect of being able to select accurately.
[0137]
[Embodiment 4]
A fourth embodiment of the present invention will be described. The present invention is also applicable to a case where a similar translation rule is dynamically generated at the time of execution when there is no translation rule whose application condition matches the input expression in the rule-driven translation system. This embodiment corresponds to using a translation rule as a bilingual expression in the first and second embodiments of the present invention.
[0138]
FIG. 12 is a diagram showing a configuration of the fourth exemplary embodiment of the present invention. A fourth embodiment of the present invention will be described with reference to FIG. The fourth embodiment of the present invention includes an input unit 131 for inputting an input expression of a source language, a rule storage unit 132 for storing translation rules, and a source language semantic distance calculation for calculating a semantic distance between the source language expressions. A target language semantic distance calculator 134 for calculating a semantic distance between the target language expressions, a source language semantic distance calculator 133, and a target language semantic distance calculator 134, and A rule meaning specifying unit 135 for resolving the ambiguity of the meaning of the source language expression of the application condition of each translation rule, a similar rule specifying unit 136 for specifying a rule similar to the input expression based on the specified meaning, A translation unit 137 for performing translation and an output unit 138 for outputting the converted target language expression are provided. The translation unit 137 further includes a rule application unit 1371 that translates according to the specified similarity rule, and an additional translation word selection unit 1372 that selects a translation word of a part whose translation has not been determined by the rule application unit 1371.
[0139]
In the above configuration, the input unit 131, the source language semantic distance calculation unit 133, the target language semantic distance calculation unit 134, and the output unit 138 are the input unit 11, the source language semantic distance calculation unit 13, and the source language semantic distance calculation unit 13 in the first embodiment. This is the same as the target language meaning distance calculation unit 15 and the output unit 17.
[0140]
The rule meaning specifying unit 135 is the same as the rule meaning specifying unit 64 in the third embodiment.
[0141]
When the source language expression has a specific syntactic structure, the translation rule storage unit 132 stores a translation rule that defines how to translate the structure in the target language.
[0142]
FIG. 7 shows an example of a translation rule when the source language is English and the target language is Japanese.
[0143]
The similarity rule specifying unit 136 corresponds to the similar bilingual expression specification unit 14 in the first embodiment, and performs input based on the specified meaning of the source language expression in the application condition of the translation rule by the rule meaning specifying unit 135. Identify translation rules applicable to source language expressions that are similar to the expression.
[0144]
The rule application unit 1371 corresponds to the bilingual expression application unit 161 in the embodiment of the first invention, and performs translation according to the specified translation rule.
[0145]
The additional translation word selection unit 1372 corresponds to the additional translation word selection unit 162 in the first embodiment of the present invention, and selects a translation word of a part for which a translation has not been determined by the rule application unit 1371. Here, in a similar manner to the additional translation word selection unit 162, of the translations of the portion, the translation word assigned to the expression in the application condition of the similar rule generator translation rule corresponding to the portion and the target language semantic distance are determined. Select a translation that is closer.
[0146]
FIG. 13 is a flowchart for explaining the operation of the fourth embodiment of the present invention. The operation of the fourth exemplary embodiment of the present invention will be described with reference to FIGS.
[0147]
For each translation rule in the rule storage unit 132 in advance, if the source language expression in the application condition has ambiguity in the meaning, the rule meaning specifying unit 135 causes the source language semantic distance calculation unit 133 and the target language semantic distance calculation unit Using 134, the ambiguity of the meaning is eliminated from the translation given to the source language expression (step 1401).
[0148]
The details of step 1401 are the same as those of step 902 (see FIG. 9) in the third embodiment of the present invention, and thus the details will be omitted.
[0149]
Thereafter, when the input expression is input to the input unit (step 1402), the similar rule specifying unit 136 controls the source language semantic distance calculation unit 133 to convert the source language expression similar to the input expression from the rule storage unit 132. An applicable bilingual expression is specified (step 1403). Here, since the ambiguity of the meaning of the source language expression in the rule application condition has been eliminated by the rule meaning specifying unit 136 in advance, the rule applied to the source language expression that is semantically different from the input expression is specified. Never.
[0150]
Next, the input expression is translated from the bilingual expression specified in step 1403 (step 1404), and finally, the translation result is output from the output unit (step 1406).
[0151]
In this embodiment, the source language semantic distance calculation unit 133, the target language semantic distance calculation unit 134, the rule meaning specification unit 135, the similar rule specification unit 136, the rule application unit 1371, and the additional translation word selection unit 1372 Each process and the process control of the input unit 131 and the output unit 138 are preferably realized by a program executed by a computer.
[0152]
[Example 4]
Next, a specific operation example (Example 4) will be described. In the fourth embodiment, an operation of translating from English into Japanese in the fourth embodiment will be described.
[0153]
In the fourth embodiment of the present invention, the ambiguity of the meaning of the source language expression in the application condition of each rule in the rule storage unit 132 is removed in advance.
[0154]
As an example, the following rules are given.
[0155]
R2: (keep (OBJ mouse)) → (Keep (the mouse))
[0156]
“Mouse” has the meaning of an animal mouse (m1) and the meaning of a mouse of a personal computer peripheral device (m2). In R2, since the mouse is given as a translation of “mouse”, the present invention is practiced. In the same manner as in Steps 502 to 505 of Example 2, it is found that the meaning of “mouse” of R2 is the meaning of “mouse” of the animal ((05) in FIG. 10).
[0157]
Here, it is assumed that the following translation rules exist in the rule storage unit 132.
[0158]
R2: (keep (OBJ mouse)) → (Keep (the mouse))
R3: (keep (OBJ knife)) → (has (knife))
[0159]
It is also assumed that the following is given as an input expression.
[0160]
I2: I keep a keyboard
[0161]
When the English expression I2 is input from the input unit 131, the similar rule specifying unit 136 determines the translation rule having the highest similarity among the translation rules applicable to the source language expression similar to the input from the rule storage unit 132. Identify. At this time, it is known that “mouse” in the source language expression in the above rule R1 has the meaning of an animal mouse ((05) in FIG. 10), and “mouse” (05) and “keyboard” (09) ) Is longer than the distance between “knife” (08) and “keyboard” (09).
[0162]
Therefore,
R2 is used by mistake,
"I have a keyboard" is not translated,
R3 is used,
"I have a keyboard"
Is translated.
[0163]
It is also assumed that the following is given as an input expression.
[0164]
I1: I serve a dish
[0165]
When the English expression is input from the input unit 131, the similar rule specifying unit 136 specifies the following rule in the same manner as the above method.
[0166]
R4: (serve (OBJ pie)) → (put out (pie))
[0167]
In the rule applying unit 1371, the translation of "serve ~" is determined from the rule R3 as "serving ~".
[0168]
When there are two translations of “dish” as “translation” of “dish”, the additional translation selection unit 1372 performs the same operation as the additional translation selection unit 162 (see FIG. 1) in the first embodiment. , Select a translation. "Cooking" is selected as a translation.
[0169]
Further, the translation of "I" is determined as "I" from another translation rule or word dictionary. As a result, from the output unit 138,
"I serve food"
Is output.
[0170]
In the present embodiment, the meaning of the source language expression in the application condition of the translation rule may be specified after the input expression is given, or may be specified in advance. Further, in this embodiment, a configuration having only one of the rule meaning specifying unit 135 and the additional translation word selecting unit 1372 may be adopted.
[0171]
As described above, in the present embodiment, when there is no translation rule applicable to the input, and when a similar translation rule is used instead, the semantic distance calculation on the target language side is performed and each rule is determined in advance. By eliminating the ambiguity of the meaning of the source language expression in the application conditions of, the use of incorrect translation rules is suppressed, and the original translation rules replaced by the original translation rules are used. This has the effect that the translation of the linguistic expression can be selected more accurately.
[0172]
Although the present invention has been described with reference to each of the above embodiments, the present invention is not limited only to the configuration of the above embodiment, but falls within the scope of the claims of the present invention. Needless to say, various changes and modifications that can be made by those skilled in the art are included.
[0173]
【The invention's effect】
As described above, according to the present invention, in an input expression in which a translation is not directly determined from a bilingual expression, which may not have been correctly determined in the first conventional example described in [Prior Art]. By providing a constraint using the semantic distance of the target language with respect to a part of the target language, it is possible to determine translation words more accurately.
[0174]
According to the present invention, by calculating the semantic distance on the target language side and eliminating the ambiguity of the meaning of the source language expression of each bilingual expression in advance, the similar bilingual expression can be correctly corrected in the first conventional example. In the case where the similar bilingual expression cannot be specified, the similar bilingual expression can be specified more accurately.
[0175]
According to the present invention, when a similar rule of a translation rule is automatically generated as a pre-process at the time of performing translation or before performing translation, a semantic distance calculation is performed on the target language side, and the meaning of the source language expression of each rule is determined in advance. By eliminating the ambiguity of, the generation of erroneous similar rules is suppressed, and the translation of the source language expression replaced from the original translation rule at the time of generating similar rules is more accurately selected. The effect is that it can be done.
[0176]
According to the present invention, when there is no translation rule applicable to the input and a similar translation rule is used instead, a semantic distance calculation on the target language side is performed to determine in advance the meaning of the source language expression in the application condition of each rule. By eliminating ambiguity, it is possible to suppress the use of incorrect translation rules, and to more accurately select the translated word of the source language expression of the part replaced from the original translation rule used The effect is as follows.
[Brief description of the drawings]
FIG. 1 is a diagram showing a configuration of a first exemplary embodiment of the present invention.
FIG. 2 is a flowchart illustrating a machine translation processing procedure according to the first embodiment of the present invention.
FIG. 3 is a diagram showing a configuration of a second exemplary embodiment of the present invention.
FIG. 4 is a flowchart illustrating a machine translation processing procedure according to the second embodiment of the present invention.
FIG. 5 is a flowchart showing a processing procedure of step 401 (subroutine) in FIG. 4;
FIG. 6 is a diagram illustrating a configuration of a similar rule generation device according to a third embodiment of the present invention.
FIG. 7 is a diagram illustrating an example of a translation rule.
FIG. 8 is a diagram illustrating an example of generation of a similar rule from a translation rule.
FIG. 9 is a flowchart illustrating a procedure for generating a similar rule according to the third embodiment of the present invention.
FIG. 10 is a diagram showing an example of a source language English thesaurus when translating from English to Japanese.
FIG. 11 is a diagram showing an example of a Japanese thesaurus of a target language when translating from English to Japanese.
FIG. 12 is a diagram showing a configuration of a fourth embodiment of the present invention.
FIG. 13 is a flowchart illustrating a machine translation processing procedure according to the fourth embodiment of the present invention.
FIG. 14 is a diagram showing a configuration of a conventional machine translation device.
[Explanation of symbols]
11 Input section
12 Bilingual expression storage
13 Source language semantic distance calculator
14 Similar bilingual expression specifying unit
15 Target language semantic distance calculator
16 Translator
161 Bilingual expression application section
162 Additional translation word selection section
17 Output section

Claims (20)

In a translation device that inputs a source language expression, converts it into a target language expression, and outputs it,
A translation device, comprising: a target language semantic distance calculation unit configured to calculate a semantic distance between the target language expressions. In a translation device that inputs a source language expression, converts it into a target language expression, and outputs it,
A bilingual expression storage unit storing a bilingual expression composed of a set of a source language expression and a target language expression corresponding to the source language expression;
Source language semantic distance calculation means for calculating the semantic distance between the source language expressions,
A target language semantic distance calculating means for calculating a semantic distance between the target language expressions,
Controlling the source language semantic distance calculation means, and specifying a bilingual expression most similar to the input expression;
Using a bilingual expression specified by the similar bilingual expression specifying means, a bilingual expression applying means for translating the input expression,
Controlling the target language semantic distance calculation means, an additional translation word selection means for selecting a translation word of a part whose translation is not determined by the bilingual expression application means,
A translation device, comprising: In a translation device that inputs a source language expression, converts it into a target language expression, and outputs it,
Source language semantic distance calculation means for calculating the semantic distance between the source language expressions,
A target language semantic distance calculating means for calculating a semantic distance between the target language expressions,
Given a source language expression and a target language expression corresponding to the source language expression, the source language expression is calculated using the source language semantic distance calculation means and the target language semantic distance calculation means. Means for resolving ambiguity in the meaning of
A translation device, comprising: In a translation device that inputs a source language expression, converts it into a target language expression, and outputs it,
A bilingual expression storage unit storing a bilingual expression composed of a set of a source language expression and a target language expression corresponding to the source language expression;
Source language semantic distance calculation means for calculating the semantic distance between the source language expressions,
A target language semantic distance calculating means for calculating a semantic distance between the target language expressions,
From the target language expression corresponding to the source language expression in the bilingual expression stored in the bilingual expression storage means, the source language semantic distance calculating means and the target language semantic distance calculating means are used to calculate the source of the bilingual expression. A rule semantic identification means for resolving the ambiguity of the meaning of the linguistic expression,
Controlling the source language semantic distance calculation means, based on the meaning of the source language expression of the bilingual expression specified by the rule meaning specifying means, a bilingual expression specifying means for specifying a bilingual expression most similar to the input expression,
Using the bilingual expression specified by the bilingual expression specifying means, translating means for translating the input expression,
A translation device, comprising: In a translation device that inputs a source language expression, converts it into a target language expression, and outputs it,
A bilingual expression storage unit storing a bilingual expression composed of a set of a source language expression and a target language expression corresponding to the source language expression;
Source language semantic distance calculation means for calculating the semantic distance between the source language expressions,
A target language semantic distance calculating means for calculating a semantic distance between the target language expressions,
Controlling the source language semantic distance calculating means and the target language semantic distance calculating means to resolve the ambiguity of the meaning of the source language expression of each bilingual expression stored in the bilingual expression storage means; Means,
Controlling the source language semantic distance calculation means, and specifying a bilingual expression most similar to the input expression of the source language;
Translation means for translating the input expression based on the bilingual expression specified by the similar bilingual expression specifying means,
Output means for outputting the converted target language expression;
A translation device, comprising: In a translation device that inputs a source language expression, converts it into a target language expression, and outputs it,
When the input source language expression has a predetermined specific structure, translation rule input means for inputting a translation rule that determines how to translate the structure into the target language,
Source language semantic distance calculation means for calculating the semantic distance between the source language expressions,
A target language semantic distance calculating means for calculating a semantic distance between the target language expressions,
If the source language expression in the application condition of the translation rule input from the translation rule input means has ambiguity in meaning, the source language semantic distance calculation is performed based on the translation given to the source language expression in the translation rule. Using the means and the target language semantic distance calculating means, to specify the meaning, and rule meaning specifying means for eliminating the ambiguity of the meaning of the source language expression of the application condition of the input translation rule,
A similar rule generation means for obtaining a similar source language expression based on the meaning of the source language expression specified by the rule meaning specifying means, and replacing the input translation rule with the similar source language expression to generate a similar rule; ,
Rule translation word selection means for selecting a translation word for the replaced source language expression,
Output means for outputting the generated similarity rule;
A translation device, comprising: Input means for inputting a source language expression;
Rule storage means for storing translation rules,
Source language semantic distance calculation means for calculating the semantic distance between the source language expressions,
A target language semantic distance calculating means for calculating a semantic distance between the target language expressions,
A rule meaning for eliminating the ambiguity of the meaning of the source language expression of the application condition of each translation rule stored in the rule storage means, using the source language semantic distance calculation means and the target language semantic distance calculation means; Identification means,
Based on the meaning specified by the rule meaning specifying means, similar rule specifying means for specifying a rule similar to the input source language expression,
A translation unit including a rule application unit that performs translation in accordance with the similar rule specified by the similar rule identification unit, and an additional translation word selection unit that selects a translation word of a part whose translation has not been determined by the similar rule application unit. ,
Output means for outputting a translation result;
A translation device, comprising: The bilingual expression meaning specifying means,
Means for inputting a bilingual expression from the bilingual expression storage unit,
Means for identifying words having a plurality of meanings in the source language expression of the bilingual expression;
Means for obtaining a synonym of the word for each meaning of the word having a plurality of meanings in the source language expression, using the means for calculating the semantic distance between the source language expressions, and creating a first set;
Means for determining a translation of each word in the plurality of first sets and creating a second set;
Means for determining a semantic distance between the translated word given to the word in the bilingual expression of interest and the second set;
Means for taking the meaning corresponding to the second set having the smallest distance from the translated word as the meaning of the word in the bilingual expression;
The translation device according to claim 5, comprising: The translation apparatus according to claim 7, comprising only one of the rule meaning specifying unit and the additional translation word selecting unit. In a translation method in which a translation process of inputting a source language expression, converting it into a target language expression and outputting the same using a computer including an input unit and an output unit,
Inputting a source language expression from the input unit;
By controlling the process of calculating the semantic distance between the source language expressions, the input from the bilingual expression storage unit storing a bilingual expression consisting of a set of the source language expression and the target language expression corresponding to the source language expression, Identifying a bilingual expression for the source language expression that is most similar to the rendered expression;
Translating the input expression in the source language using the specified bilingual expression;
Controlling the process of calculating the semantic distance between the target language expressions, and determining the translation of the word in the input expression whose translation was not determined in the step of performing the translation;
Outputting a translation result from the output unit;
A translation method, comprising: In a translation method in which a translation process of inputting a source language expression, converting it into a target language expression and outputting the same using a computer including an input unit and an output unit,
Referring to a bilingual expression storage unit storing a bilingual expression composed of a set of a source language expression and a target language expression corresponding to the source language expression, for each bilingual expression stored in the bilingual expression storage unit, When there is ambiguity in the meaning of the linguistic expression, the processing is performed on the semantic distance between the source language expressions and the processing on the semantic distance between the target language expressions. A step of previously eliminating the ambiguity of the meaning from the translation;
Inputting a source language expression from the input unit;
Controlling a process of calculating a semantic distance between source language expressions, and specifying a bilingual expression (hereinafter, referred to as “most similar bilingual expression”) for the source language expression most similar to the input expression from the bilingual expression storage unit When,
Translating the input source language expression using the identified most similar bilingual expression;
Outputting a translation result from the output unit;
A translation method, comprising: In a translation method in which a translation process of inputting a source language expression, converting it into a target language expression and outputting the same using a computer including an input unit and an output unit,
When the source language expression has a predetermined specific structure, a step of inputting a translation rule that defines how to translate the structure into the target language;
When there is ambiguity of meaning in the source language expression of the application condition of the input translation rule, a process of calculating a semantic distance between the source language expressions and a process of calculating a semantic distance between the target language expressions are performed. Using the translation rules to resolve the ambiguity of meaning from the translation given to the source language expression,
Using a process of calculating a semantic distance between source language expressions, generating a similar rule in which the source language expression in the application condition of the input translation rule is replaced with a semantically similar source language expression,
Determining a translation for the replaced source language expression in the similarity rules;
Outputting a translation result from an output unit;
A translation method, comprising: In a translation method in which a translation process of inputting a source language expression, converting it into a target language expression and outputting the same using a computer including an input unit and an output unit,
For each translation rule, if there is ambiguity in the meaning of the source language expressions in the applicable conditions, a process of calculating the semantic distance between the source language expressions and a process of calculating the semantic distance between the target language expressions are used. Disambiguating the meaning from the translation given to the source language expression;
Inputting a source language expression from the input unit;
Controlling the process of calculating the semantic distance between the source language expressions, and identifying a bilingual expression applicable to the source language expression similar to the input expression from the rule storage unit;
Translating the input expression from the identified bilingual expression;
Outputting a translation result from the output unit;
A translation method, comprising: The step of eliminating the ambiguity of the meaning in advance,
Inputting a bilingual expression from the bilingual expression storage unit;
Identifying words having a plurality of meanings in the source language expression of the bilingual expression;
Generating a first set of synonyms of the word for each of the meanings of the word having a plurality of meanings in the source language expression by using a process of calculating a semantic distance between the source language expressions; When,
Determining a translation for each word in the plurality of first sets and creating a second set;
Calculating a semantic distance between the translated word given to the word in the bilingual expression of interest and the second set;
Making the meaning corresponding to the second set having the smallest distance from the translation word the meaning of the word in the bilingual expression;
The translation method according to claim 11, comprising: In a program that causes a computer including an input unit and an output unit to execute a translation process of inputting a source language expression, converting the input language expression into a target language expression, and outputting
Inputting a source language expression from the input unit;
Processing for calculating the semantic distance between the source language expressions;
Processing for calculating the semantic distance between the target language expressions;
By controlling the process of calculating the semantic distance between the source language expressions, a bilingual expression storage unit storing a bilingual expression consisting of a set of a source language expression and a target language expression corresponding to the source language expression, A process of identifying a bilingual expression for a source language expression most similar to the input expression;
A process of translating the input expression using the specified bilingual expression;
Controlling a process of calculating a semantic distance between the target language expressions, and determining a translation of a word in the input expression whose translation has not been determined in the process of translating the input expression;
Outputting a translation result from the output unit;
For causing the computer to execute. In a program that causes a computer including an input unit and an output unit to execute a translation process of inputting a source language expression, converting the input language expression into a target language expression, and outputting
Processing for calculating the semantic distance between the source language expressions;
Processing for calculating the semantic distance between the target language expressions;
Referring to a bilingual expression storage unit storing a bilingual expression composed of a set of a source language expression and a target language expression corresponding to the source language expression, for each bilingual expression stored in the bilingual expression storage unit, When there is ambiguity in the meaning of the expression of the source language expression, the process of calculating the semantic distance between the source language expressions and the process of calculating the semantic distance between the target language expressions are applied to the source language expression. Processing to resolve the ambiguity of the meaning from the translation in advance,
Inputting a source language expression from the input unit;
By controlling the processing of calculating the semantic distance between the source language expressions, a bilingual expression (hereinafter referred to as “most similar bilingual expression”) for the source language expression most similar to the input expression is specified from the bilingual expression storage unit. Processing,
Using the specified most similar bilingual expression, a process of translating the input source language expression,
Outputting a translation result from the output unit;
For causing the computer to execute. In a program that causes a computer including an input unit and an output unit to execute a translation process of inputting a source language expression, converting the input language expression into a target language expression, and outputting
Processing for calculating the semantic distance between the source language expressions;
Processing for calculating the semantic distance between the target language expressions;
When the source language expression has a predetermined specific structure, a process of inputting a translation rule that defines how to translate the structure into the target language;
When the source language expressions of the application conditions of the input translation rules have ambiguity in meaning, a process of calculating a semantic distance between the source language expressions and a process of calculating a semantic distance between the target language expressions Processing to remove the ambiguity of the meaning from the translation given to the source language expression by the translation rule, and
Using a process of calculating a semantic distance between the source language expressions, generating a similar rule in which the source language expressions in the application conditions of the input translation rules are replaced with semantically similar source language expressions. Processing,
Determining a translation for the replaced source language expression in the similarity rules;
A process of outputting a translation result from an output unit;
For causing the computer to execute. In a program that causes a computer including an input unit and an output unit to execute a translation process of inputting a source language expression, converting the input language expression into a target language expression, and outputting
Regarding the translation rules, when there is ambiguity in the meaning of the source language expressions in the applicable conditions, the process of calculating the semantic distance between the source language expressions and the process of calculating the semantic distance between the target language expressions are used. Processing to resolve meaning ambiguity from the translation given to the source language expression,
Inputting a source language expression from the input unit;
A process of controlling a process of calculating a semantic distance between source language expressions, and specifying a bilingual expression applicable to a source language expression similar to the input expression from the rule storage unit;
Translating the input source language expression from the specified bilingual expression;
Outputting the translation result from the output unit to the computer. The program according to claim 16,
The process of eliminating the ambiguity of the meaning in advance,
Inputting the bilingual expression storage unit bilingual expression;
A process of specifying a word having a plurality of meanings in the source language expression of the bilingual expression;
A process of calculating a synonym of the word for each of the meanings of the word having a plurality of meanings in the source language expression by using a process of calculating a semantic distance between the source language expressions, and creating a first set When,
A process of obtaining a translation of each word in the plurality of first sets and creating a second set;
A process of calculating a semantic distance between a translated word given to a word in the bilingual expression of interest and the second set;
A process of setting the meaning corresponding to the second set having the smallest distance to the translated word as the meaning of the word in the bilingual expression;
For causing the computer to execute. A computer-readable recording medium on which the program according to any one of claims 15 to 19 is recorded.

Images