JPH0477964A - Machine translation device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention is suitable for application when translating a first language such as Japanese into a second language such as English, which is the translation target language. Regarding machine translation devices.

- In the layer details, if there is no factor to determine the subject in the input sentence to be translated, the desired subject is selected from the candidate subjects prepared in advance before translation is executed, or the desired subject is selected from the outside. A translation with the desired subject is generated by generating a translation result sentence by adopting either the set subject or the subject determined by the rules of generative grammar when the desired subject is selected or not set. This makes it easier to generate result sentences and allows efficient translation.

[Prior Art] Recently, there has been a desire to improve the efficiency of translation into different languages, and machine translation devices based on computer systems are being used.

In such a machine translation device, a first language such as Japanese in which a subject is not required may be translated into a second language such as English in which the subject is required. At this time, if a clue process regarding the subject is not set in advance for the input sentence to be translated in the first language, the subject determined by the rules of the generative grammar will be adopted and the translated sentence in the second language will be translated. generated. Alternatively, a translation result sentence in the second language in the passive form is generated and translated.

[Problems to be Solved by the Invention] In the conventional machine translation device described above, when translating from a first language to a second language, only the subject defined by the rules of generative grammar is applied, and the translation result is Because a sentence is generated or a translated sentence is generated in the passive form, it is difficult to obtain a translated sentence with the desired subject, which makes post-editing difficult and reduces translation efficiency. are doing.

The present invention has been made in view of the above, and an object of the present invention is to obtain a translation result sentence with a desired subject when translating from a first language in which the subject is not essential to a second language in which the subject is essential. Therefore, it is an object of the present invention to provide an excellent machine translation device that can translate efficiently.

[Means for Solving the Problems] In order to solve the above problems, the machine translation device of the present invention, as shown in the claim correspondence diagram of FIG. The machine translation device generates a translation result sentence in a second language that does not include a sentence, in which a selection and designation means (A) selects and designates a subject in the second language from a plurality of subjects prepared in advance.

Further, in the setting means (B), when the desired subject is not selected by the selection and designation means (A), the subject of the second language is input and set. (A), when the subject is not specified by the setting means (B), the subject determined by the rules of the generative grammar is adopted.

Furthermore, the resetting means (D) directly sets the subject and additional information.

[Operation] In the machine translation device of the present invention configured as described above, when there is no factor determining the subject in the input sentence to be translated,
A desired subject selected externally from candidate subjects prepared in advance before execution of translation, a subject set externally, or a rule of generative grammar when the desired subject is not selected or set. One of the determined subjects is adopted to generate a translation result sentence. Thereby, a translation result sentence with a desired subject can be easily generated.

[Example] Next, an example of the machine translation device according to the present invention will be described in detail below with reference to the accompanying drawings.

FIG. 2 is a configuration diagram showing an embodiment, FIG. 3 is a flowchart for explaining the operation, and FIGS. 4(a) and 4(b) are screen diagrams for explaining the operation.

First, the configuration will be explained. In FIG. 2, 10 is a machine translation device such as a computer system, and this machine translation device 10 has an input section 12 consisting of a keyboard, etc., and a CRT.
A display section 14 such as a printer, etc., and a printing section 16 such as a scanning printer are provided. Furthermore, a data processing section 18 which is a computer main body to which an input section 12, a display section 14, and a printing section 16 are connected.
has.

The data processing section 18 includes an editing section 20 and a translation section 22, which are function combining means, as well as an analysis dictionary 24, a conversion dictionary 26, and a generation dictionary 28 stored on a hard disk or optical disk.

The translation unit 22 includes a morphological analysis unit 30, which is a functional combination means related to a well-known transfer method, and a syntactic analysis unit 32.
Syntax conversion section 34, syntax generation section 36, and morpheme generation section 3
8 is provided.

Here, the basic operation in the above configuration will be explained.

The translation target sentence input from the input unit 12 is edited into an appropriate form by the editing unit 20 and then input to the translation unit 22. The sentence to be translated inputted to the translation unit 22 is divided into words by the morphological analysis unit 30 with reference to the analysis dictionary 24, and the parts of speech and conjugations are clarified. Subsequently, the parsing section 32 refers to the parsing dictionary 24 to add dictionary information, and based on this, parsing is performed to clarify the structure and function. Thereafter, the translation target sentence is input to the syntax conversion unit 34, and the translation target sentence is converted into a translation target sentence with reference to the conversion dictionary 26. Then, the translation target sentence converted into the translation target sentence by the syntax converter 34 is generated into a correct syntax by the syntax generator 36 based on the dictionary information of the generator dictionary 28 and the grammar rules of the translation target sentence. Thereafter, the morpheme generation unit 38 generates a translation result sentence. Further, the translation result sentence is freely edited in the editing section 20 and displayed as an image on the display section 14, or output as a hard copy from the printing section 16.

Next, the operation when generating a translated sentence using any subject will be explained with reference to the flowchart of FIG.

In step (hereinafter referred to only as S) 100, a sentence to be translated "I'm going to school." is input from the input unit 12.

Proceed to the next step 8101 and proceed to the contents shown in Figure 4 (a), that is, [Determine the default subject. ] [Please enter your number. ] r(1)itJ r(2) the3rJ r(3) heJ r(4)N "(n) Direct human power" is displayed as an image.

Next, the process advances to 5102 and the operator inputs the r(1) i i , r(2) theyJ, r(
3) heJ, r(4) I ``--r(n) Direct manual labor'' is specified.

Next, in step 5103, the morphological analysis unit 30 and syntactic analysis unit 32 of the translation unit 22 perform analysis.

Furthermore, in 5104, the syntax conversion unit 34 converts the conversion dictionary 2
The conversion process is executed with reference to 6.

Then, in 5ios, the sentence to be translated is ``I'm going to school.''

” is judged whether there is a clue regarding the subject or not.

Here, there is no clue regarding the subject in the sentence to be translated, ``Go to school.'' Here, it is impossible to supplement the appropriate subject using the grammar rules of the translation unit 22. If this decision is YES, it is assumed that the translation can be generated as is, and the process proceeds to generation processing in step 5111. If NO, proceed to the next step 8106. Here, it is assumed that r(3] heJ is specified. Then, in 5106, it is determined whether a subject has been specified. If no, the following 8107
The process of setting the subject (r(4)N) designated as the default value is executed, and it is determined that the translation can be generated, and the process proceeds to the process of generating 5ill. If Yes, 510
The process advances to step 8, where it is determined whether or not the content has been specified from the options shown in FIG. 4(a). In case of No,
The process advances to step 5109, where the subject entered here is set as the default value, as shown on the display screen of FIG. 4(b). Then proceed to 5iti. If Yes, the designated subject is determined as the default and the process proceeds to 5111. In this 5111, the operator is the subject r(3) heJ
rl g for the sentence to be translated by selecting ``I'm going to school.''

A translation result sentence in English of the translation target language, ``to 5choolJ'', is generated. For reference, if r(3) heJ is selected, rHe goesto 5cho
It becomes olJ.

Next, when translating the target sentence of a technical geek into English, ``It turns slowly.'' Again, there are no clues about the subject in the sentence. In this case, the operator must understand that the subjects of the sentences to be translated in the technical manual are often parts, etc., as shown in Figure 4 (a).
) instead of the subject in the options shown in [rThe part turns sl
.

We would like to obtain a translated sentence w 1 y J .

Here, first, the above 5108, 5I09. SL ],
I will be executed. Here, select the desired subject rthe pa on the selection display screen shown in FIG. 4(b).
rt'', the above rThe pa
The translation result sentence rt turns sl□ w i y J is obtained.

As described above, when there is no factor that determines the subject in the input sentence to be translated, candidate subjects r(1) itJ , r(2) they prepared in advance before translation are executed.
J, r(a) heJ, r(4) I J --
The desired subject "(3) he" specified externally from "r(Ill direct human power)", or the externally set subject rthe partJ, or the generative grammar when the desired subject is not selected or set. A translation result sentence is generated by adopting one of the subjects determined according to the rule.Thereby, a translation result sentence with a desired subject can be easily generated.

In addition, in the above embodiment, the first language is Japanese,
Furthermore, although the explanation has been made by applying English as a second language, this is not limited to this, and it is also possible to obtain similar effects by applying it to other languages that have similar language forms, such as French and Spanish. Included in invention.

[Effects of the Invention] As can be understood from the above explanation, according to the machine translation device of the present invention, when there is no factor for determining the subject in the input sentence to be translated, candidates prepared in advance before translation are the desired subject selected from the outside, or
The translation result sentence is generated by selecting either the subject set from the outside or the subject determined by the rules of generative grammar when the desired subject is not selected or set, so the desired subject is This makes it easier to generate translation result sentences with the following functions, which has the effect of improving translation efficiency.

[Brief explanation of drawings]

Fig. 1 is a diagram corresponding to the claims of the machine translation device of the present invention, Fig. 2 is a configuration diagram showing an embodiment of the machine translation device of the present invention, and Fig. 3 is an explanation of the operation of the embodiment shown in Fig. 2. FIGS. 4(a) and 4(b) are screen diagrams used to explain the operation of the embodiment shown in FIG. 1. 12... Input section, 14... Display section, 16.
...Printing section, 18...Data processing section, 20.
...Editorial department, 22...Translation department, 24...Analysis dictionary, 26...Conversion dictionary, 28...Generation dictionary,
30... Morphological analysis unit, 32... Syntactic analysis unit,
34... Syntax conversion unit, 36... Syntax generation unit,
38...morpheme generation unit.

Claims (2)

[Claims] (1) A machine translation device that generates a translation result sentence from a first language into a second language in which there are no factors that determine the subject, and where the subject of the second language is generated from a plurality of subjects prepared in advance. a selection designating means for selecting and specifying a subject; a setting means for inputting and setting a subject in a second language when the desired subject is not selected by the selection and designation means; A machine translation device comprising: a rule subject adoption means for adopting a subject determined by a rule of a generative grammar when the subject is not specified; (2) The machine translation device according to claim 1, further comprising a reset means for directly setting the subject and additional information.