JPS61150069A - Translating and editing device - Google Patents

Abstract

PURPOSE:To facilitate the collation of a mechanically translated sentence with its original, to lighten the burden imposed on an operator and to improve the translation efficiency by retrieving and displaying a corresponding word or paragraph between the original and translated sentence. CONSTITUTION:The word or paragraph for comprising the original, which is stored in an original memory part 1, is given to a translation processing part 2, and mechanically translated in sequence by referring to a translation dictionary 3. A word used registering part 3a given to the translation dictionary 3 corresponds the translated work (paragraph) available from translated in the translation processing part 2 to the word (paragraph) of the original having said translated word (paragraph) and registers it sequentially. With the use of the corresponded and registered word (paragraph) and its translated word (paragraph), the word (paragraph) corresponding to the original and the translated sentence is displayed in the collation processing of the original and translated sentence.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a translation editing device that can effectively support machine translation of an original text using a translation dictionary.

[Technical background of the invention and its problems]

2. Description of the Related Art Recently, the development of automatic translation devices, such as machine translation from English to Japanese or machine translation from Japanese to English, using computer systems has been actively progressed. This automatic translation is
Basically, a translation dictionary is used to find translated words or translated phrases corresponding to words or phrases in the original text, and these translated words or translated phrases are interconnected to find a translated text for the original text. In this case, it goes without saying that a translated sentence is constructed by concatenating the translated words or translated phrases in accordance with the grammatical rules of the original language and the translated word.

However, in general, even if a grammatically or linguistically correct translation is obtained by machine translating an original text that contains many ambiguities, the content may differ from the translation intention. For this reason, the reality is that a translator (operator) compares the machine-translated text with the original text and corrects any erroneous translations or parts that differ from the intended translation. Therefore, not only does this impose a considerable burden on the operator, but it is also difficult to expect an improvement in translation efficiency.

[Purpose of the invention]

The present invention has been made in consideration of these circumstances, and its purpose is to facilitate the process of comparing a machine-translated translation with its original text, reduce the burden on the operator, and improve the translation process. An object of the present invention is to provide a translation editing device that can improve efficiency.

[Summary of the invention]

The present invention provides a translation editing device that includes a translation processing unit that translates an input original text using a translation dictionary and obtains a translation of the original text, and displays the original text and the translated text, respectively. The correspondence between the translated word or phrase and the word or phrase in the original text from which this translated word or phrase was obtained is examined, and the original and translated words are determined based on the correspondence between the translated word or phrase and the word or phrase in the original text. The corresponding words or corresponding phrases between, for example, according to the attributes of the word and the translated word,
The system searches for the words or phrases in the translated text and the original text, including the conjugation changes, and displays the corresponding words or phrases in a way that allows them to be distinguished from other words or phrases in the original text and the translated text.

〔Effect of the invention〕

Thus, according to the present invention, it is possible to easily find corresponding words or phrases between the original text and the translated text during the translation process, and to display the corresponding words between the displayed original text and the translated text. While effectively comparing the translated word (translated phrase) with the sentence expression of the corresponding part in the original text, it is possible to easily and efficiently match the machine-translated translation with the original text.

Moreover, according to one corresponding word (corresponding word/phrase), you can easily search for words (phrases) that have the same correspondence in the above original and translated texts, including their conjugations, so that the translation can be unified throughout the text. Processing can be performed easily. Therefore, great practical effects such as efficient translation processing can be achieved without placing a large burden on the operator (H translator).

[Embodiments of the invention]

Hereinafter, one embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic configuration diagram of the embodiment device, and FIG. 2 is a diagram showing an example of display of an original text and its translation in the embodiment device.

The words or phrases constituting the original text stored in the original text storage unit 1 are given to the translation processing unit 2, and are sequentially machine translated by referring to the translation dictionary 3. In this case, it goes without saying that the translation process is performed according to the results of syntactic and grammatical analysis of the original text. The machine-translated translated words (translated phrases) are assembled syntactically according to predetermined grammatical rules and the like, and are stored in the translated sentence storage unit 4 as translated sentences for the original text.

The display unit 5 displays the above-mentioned original text to be machine translated in this way, its translated text, the plural translation candidates when a plurality of translation candidates exist for the word (phrase) to be translated, and the above-mentioned translation candidate obtained from the above-mentioned translation candidates. Words (phrases), etc. are displayed, for example, as shown in FIG. In FIG. 2, the first display area 5a is used to display the input original text (here, the English text), and the second display area 5b is used to display the translated text (here, the Japanese text). Also, the third display section 1g5
c is used to display the word (phrase) to be translated and its multiple translation candidates. An instruction to select and determine an appropriate translation candidate among the plurality of translation candidates displayed in the third display area 5C is performed via the instruction input section 6.

Therefore, the used word registration section 3a provided in the translation dictionary 3,
The translation obtained by translation processing in the translation processing unit 2 above!
! (translated phrase) and X in the original text from which the translated word (translated phrase) was obtained! (words/phrases) are registered sequentially in association with each other. Specifically, the address data in the translation dictionary 3 of the word used for translation and its part-of-speech information are stored in order, and the registration is performed. Using the words (words and phrases) and their translations registered in this way, the corresponding words (corresponding phrases) in the original and translated texts are displayed in the matching process between the original and the translated text, which will be described later. It will be done. Furthermore, the display of corresponding words (corresponding phrases) in the original text and translated text is as follows:
This is performed by searching for words (phrases) in the original text and translated text by the search unit 2a provided in the translation processing unit 2. This word (phrase) search is performed by pattern matching of character strings that constitute each word (phrase) in the original text and translated text, including the transformation of the word (phrase) due to changes in its usage.

Thus, there is now an original text in English for a device configured in this way, such as rMany economists forecast.
that the market in Japan
Assume that a Japanese translation is given as ``will pick up, J'' and obtained by machine translation as ``Many economists predict that Japan's economy will recover.'' Since the translated word "forecast" is unnatural, the operator needs to correct it by using the translated text displayed in the display area 5b inside the display unit 5. The corresponding translated word in the translated text is indicated by, for example, moving the cursor 5x. Upon receiving this instruction information, the search section 2a refers to the used word registration section 3a, searches the address in the translation dictionary 3 in which the word (phrase) in the original text from which the translated word was obtained is registered, and 3) is read out from the translation dictionary 3 as shown in FIG. 3, for example. In this case, a word (phrase) in the original text is used as a heading RA, and its part-of-speech information B and its translated word C are read out. Note that only the part of speech information B used for translation is read out.

Accordingly, the search unit 2a performs pattern matching between the designated translation word and the translation NC read out from the translation dictionary 3.
The original word from which the translated word was obtained is known from the entry word A, and for example, "f
The character string orecastj is searched for in the original text.

This search is performed by pattern matching of strings. In this case, the word (fOrecastJ) is conjugated to j rorecasts Jr forecas.
tedJ j forecasting j, etc., the search for the corresponding word is performed by longest matching or matching using normalized expression. The R (word/phrase) in the original text searched in this way is displayed as another word, for example, in high brightness display, underlined display, blanking display, etc., corresponding to the specified word (phrase) in the translated text. (phrase)
is displayed in an identifiable manner.

Note that there is no problem with words (phrases) that have regular changes in usage, but for words that have irregular changes in usage, such as r goJ, the matching method described above may not be enough to search for them. It is enough.

Therefore, in order to deal with such a case, for example, as shown in Fig. 4, a conjugation form showing irregular conjugation changes is prepared in the translation dictionary 2, and the above-mentioned search process is performed including this conjugation form. Let's do it.

Specifically, if we obtain the translated sentence ``pain went'' for the original sentence f T he pain went, J, and search for the corresponding word in the original sentence due to the unnaturalness of the translated word ``went,'' From the word "gone" we get the heading F aoJ, u A, and by combining its conjugations, we get r gojrgo.
esj rooina J fwentJ l”go
The original text is searched using neJ as a search target. In this way, it is possible to search for corresponding words even for words that exhibit irregular changes in usage, and it is possible to display corresponding words (phrases) in the original text and the translated text in a corresponding manner.

As described above, according to this device, the correspondence between the original text and the translated text is 1.
(phrases) can be displayed correspondingly in the displayed original and translated texts, making it possible to perform matching and correction processing while reliably confirming the parts to be matched during translation matching processing. . Therefore, the burden on the operator can be reduced, and translation correction can be carried out efficiently by appropriately instructing the verification points, which in turn can have great practical effects, such as improving translation processing efficiency. It can be played.

Note that the present invention is not limited to the embodiments described above. Although an example of translation from an English sentence to a Japanese sentence is given here, the reverse may be used. Furthermore, the present invention is not limited to translation between English and Japanese, but can be similarly applied to translation processing between so-called foreign languages. In addition, in the example, we explained the search for the corresponding word (phrase) in the original text for the translated word (phrase), but vice versa,
In other words, it may be possible to search for corresponding words (phrases) in the translated text for words (phrases) in the original text and display them in correspondence. Furthermore, it is also possible to configure the search section 2a independently of the translation processing section 2. In short, the present invention can be implemented with various modifications without departing from the gist thereof.

[Brief explanation of drawings]

FIG. 1 is a schematic configuration diagram of a device according to an embodiment of the present invention, FIG. 2 is a diagram showing an example of displaying an original text and a translated text in the embodiment device, and FIG.
The figure shows the dictionary contents of used words read out from the translation dictionary, and FIG. 4 shows an example of the structure of the translation dictionary that takes into account irregular changes in the usage of words. 1... Original text storage unit, 2... Translation processing unit, 2a...
Used word registration section, 3... Translation dictionary, 3a... Search section,
4... Translation storage section, 5... Display section, 6... Instruction input section. Applicant's agent Patent attorney Takehiko Suzue Figure 1

Claims (2)

[Claims] (1) In a translation editing device equipped with a translation processing unit that translates an input original text using a translation dictionary and obtains a translation of the original text, the translated word or translated phrase in the translated text and this translated word or translated phrase are obtained. searching for corresponding words or phrases in the original text and the translated text based on the correspondence between the translated word or phrase and the word or phrase in the original text; A translation editing device characterized by comprising means for displaying corresponding words or phrases in the original text and the translated text so that they can be distinguished from other words and phrases. (2) The search means for corresponding words or phrases includes changes in the conjugation of the translated word or phrase in the translated text and changes in the conjugation of the word or phrase in the original text, according to the attributes of the word and the translated word.
The translation editing device according to claim 1, which searches for corresponding words or corresponding phrases in the original text and the translated text.