JPH0193861A - Electronic translator with japanese processing function - Google Patents

Abstract

PURPOSE:To facilitate an operation and to improve a translation accuracy by alarming the presence and the absence of a multiple parts is speech word at the time of an electronic translation and nuclear points in interpreting a linkage at the time of inputting Japanese to a user and urging to correct. CONSTITUTION:At the time of inputting a KANA (Japanese syllabary) character string from a keyboard 1, it is displayed on a display device 4. Then, at the time of pressing the KANA/KANJI (Chinese character) converting key of the keyboard 1 down, the KANA character string is converted to a KANJI and KANA mixed sentence based on the data of a dictionary 1 by a KANA/KANJI converting part 12a. At the same time, the linkage on an analysis according to the check of the presence and the absence of the multiple parts of speech word and the analysis of a syntax is decided by a morphological element analyzing part 19a to decide if there are plural interpretations on the linkages or not. Consequently, when there are the plural linkages, the relation of the linkages is alarmed and displayed. Then, the user reenters the KANA character string. Then, by depressing the translation converting key of the keyboard 1, a translation is executed in a translating part 19b based on the translation dictionary of the dictionary 21.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application Field) The present invention relates to an electronic translator equipped with a Japanese processing function using kana-kanji conversion input and a translation function for translating from Japanese to another language.

(Prior art) Traditionally, Japanese language processing machines (Japanese A Japanese language word processor) and an electronic translator, which receives kanji and kana mixed sentences inputted by an optical reader (OCR) or a Japanese input device and translates them into other languages, each exist separately.

That is, the schematic configurations of the Japanese word processor and the electronic translator are shown in FIGS. 9 and 10, respectively.

In Fig. 9, 1 is a keyboard (input device), 2 is a system main body consisting of a short-lived kanji conversion dictionary a and an editing/display/print control dictionary b, and 3 is a dictionary (storage device) that stores the Hagana-to-kanji conversion dictionary. , 4 is a display device such as a CRT display, 5 is a printer (printing device), and 6 is an external storage device.

In addition, in FIG. 10, the same numbers as in the previous book indicate the same members, and 7 is the 0CR18 that performs optical reading as an input device.
9 is a Japanese input device such as a word processor; 9 is a system body consisting of an analysis section 9a that performs morphological analysis of input character strings; and a translation section 9b that performs translation into another language SB; 10;
is a dictionary (storage device) that stores a translation dictionary.

In this way, in the past, the Japanese language processing machine and the electronic translator were separate and not connected to each other, so it was possible to interpret multi-part words and multiple dependencies in sentences input into the electronic translator. If there is a problem, it is not easy to check the input sentence after translation, make corrections, and re-input the input sentence, which may lead to a decrease in the accuracy of the translated sentence.

(Purpose of the Invention) The present invention solves the above-mentioned problems by combining a Japanese ZB processing function such as a Japanese R word processor with an electronic translation function, and also by controlling the presence/absence of multi-part speech words and modification during electronic translation. is an electronic computer equipped with Japanese processing functions that is easy to operate and improves translation accuracy by warning the user of ambiguous interpretations at the time of entering the text and prompting the user to make corrections. The purpose is to provide a translation machine.

(Structure of the Invention) The present invention includes an input device for inputting a kana character string and a function execution command, a storage device that stores a kana-kanji conversion dictionary and a translation dictionary, and a storage device that stores the input kana character string in the kana-kanji conversion dictionary. A kana-kanji conversion unit that converts into kanji and kana sentences, a control unit that controls the execution of editing, display, and printing functions according to function execution commands, and a display device that is driven by this control unit and displays character strings. A printing device that prints a character string, and a morpheme that determines whether or not the converted kanji-kana mixed sentence has multiple modifications, and displays a warning on the display device when there are multiple modifications. This is an electronic translation machine equipped with a Japanese language processing function, which includes an analysis section and a translation section that converts the above-mentioned kanji-kana mixed sentences into translated sentences based on the above-mentioned translation dictionary.

With this (jl configuration), when inputting Japanese, it is checked to see if there are multiple interpretations of dependencies in the input Japanese, and if the sentence is ambiguous due to multiple dependencies, it will be displayed to that effect. is displayed.The user can therefore modify the text to unambiguous Japanese at that point.

(Embodiment) FIG. 1 shows the configuration of an embodiment of the present invention. In the same figure,
1 Keyboard for entering ephemeral character strings and function execution commands (
4 is a display device, 5 is a printer (printing device), θ is an external storage device, 20 is the main body of this machine, and this main body dictionary 0 is a system body 12 as a Japanese language processing machine and a system body 12 as a translator. A system main body 19 is provided.

The system main body 12 includes a kana-kanji conversion section 12a and a control section 12b for executing functions such as editing, displaying, and printing.
On the other hand, the system main body 19 determines the presence or absence of multipart speech words in the converted kanji kana wet sentence, and the interpretation of the same sentence.
It consists of a morphological analysis section 19a that determines whether there are multiple dependencies, and a translation section 19b that performs translation into Senkoku language.

In addition, as will be described in detail later, the above form (base element analysis unit 19
When it is determined in step a that a sentence containing kanji and kana has multiple modifications, the old one is displayed on the display device 4.

Reference numeral 21 denotes a common kana-kanji conversion/translation dictionary (hereinafter simply referred to as ``jiyo'') as a storage device in which a kana-kanji conversion dictionary and a translation dictionary are stored, and the dictionary is read from the main body 12.19 of both systems.

Next, the operation of the above configuration will be explained with reference to FIG. 2 showing a flowchart and FIG. 3 showing changes in the screen of the display device 4.

First, when a kana character string is input from the keyboard 1, which is an input device (step S1), it is displayed on the display device 4 as shown in FIG. 3(a).

Here, we will introduce "A new study guide"
An example of inputting the following is shown below. Next, when you press the Kana-Kanji conversion key on keyboard 1, the above kana character string will be converted to Kana-Kanji conversion section 1.
2a converts the sentence into a kanji-kana mixed sentence based on the data in the dictionary 21 (step S2>. At the same time, the morphological analysis unit 19a checks the rightness of the multi-part speech word and determines the analytical dependency by syntactic analysis. and here it is determined whether there are multiple interpretations of dependencies (
Steps 83, 34). As a result, if there are multiple dependencies, a warning will be displayed for the dependencies, and the third
The screen will look like the one shown in Figure (b). Here, ``Shin'' in the converted kanji-kana mixed sentence has a dependency relationship with both ``Institute 1'' and ``Director.'' Therefore, on the screen, the modification indicates the relationship with an arrow and also displays a message that there are multiple interpretations.

If the user looks at this screen and determines that the dependency is incorrect, the user re-enters the kana character string and repeats the same conversion and analysis steps (Steps 81 to S4); , enter the next parsed and kana string in the same way.

Figure 3 (C) shows the screen when the above re-input is performed. Here, by re-entering "Kenki Yujo's New Gocho" and pressing the Kana-Kanji conversion key,
The screen shown in FIG. 3(d) will appear. In this state, the dependency is uniquely determined and there is no ambiguity, so no warning is displayed.

Next, if the input sentence is at the end of the sentence (YES in step S6)
, by pressing the translation key on the keyboard 1, the translation into another language, for example English, is performed based on the translation dictionary in the dictionary 21 (step 87).

As a result, the translated text is displayed on the screen as shown in FIG. 3(e).

As mentioned above, when inputting a Kana character string and converting Kana to Kanji, interpretational dependencies are checked, so the translation results are accurate, improving translation accuracy, and after translation conversion, There is less wasteful operation work compared to the case where the user takes steps such as inputting a corrected kana character string.

It should be noted that from the screen in Figure 3(b), the user can select the relationship for the modification, and with that,
You can change the relationship between modifier numbers without re-entering the kana character string.
・It may be something that can be determined logically.

Next, two application examples of the present invention are shown in FIGS. 4, 5, and 6.
This will be explained with reference to FIGS.

FIG. 4 shows the first application example, in which the main body dictionary 2 of this machine includes, in addition to the part corresponding to the above-mentioned main body dictionary 0,
User dictionary registration dictionary 2a and translation dictionary check dictionary 2b
has. In addition, the dictionary 21 has a user dictionary 24 in its column. In this configuration, users can
When registering one word in the user dictionary 24, the same word is checked against the dictionary 21 owned by the system, and it is checked whether the same word has already been registered and whether it has been registered with a different part of speech. Users will be alerted if there is a problem.

To explain the above operation with reference to Fig. 5, the keyboard 1
When user dictionary 1 registration is specified from , the user dictionary registration dictionary 2a displays a setting screen on the display device 6. here,
Enter information such as Japanese, English, part of speech, etc. from keyboard 1. The dictionary search dictionary 2b searches the dictionary 21,
It is checked whether the same word exists in the dictionary 21 and whether it is recorded with a different part of speech. If there is no possibility of mistranslation, the input word is registered as is, but if there is a problem, the drawing A warning will be displayed.

An example of the screen at this time is shown in FIG. 5(a). In this example, the English word “m a de” is mistakenly translated into Japanese as “desk”.
, shows the case where "noun" is input as the part of speech. In this case, a message saying "made is already registered as a verb" is displayed.

However, if the user registers words in the user dictionary 24 without any check function, there is a possibility that words that are important for electronic translation, such as verbs, adjectives, and prepositions, may be mistakenly recorded as a different part of speech. This was the basis for mistranslation during translation, but such problems can be prevented. In other words, if there is no incorrect user dictionary registration, the translation will be correct as shown in Figure 5(b), but if the incorrect user dictionary registration is made, the translation will be correct as shown in Figure 5(b). Mistranslations occur as shown in C). In this embodiment, this problem of mistranslation is eliminated, and together with the function that checks whether there are multiple interpretational dependencies mentioned above, it is possible to further improve usability and translation accuracy. .

FIG. 6 shows a second application example, in which the main body dictionary 5 of this machine corresponds to the above-mentioned main body dictionary 0, and this main body dictionary 5
And, l! A dictionary switching dictionary 6 is interposed between the i21. This dictionary switching dictionary 6 is used to switch translations from Japanese to English or American, for example.

In other words, some languages have different expressions in English and primitive languages, but by storing different translations in the dictionary 21 and specifying one of the modes in the switching dictionary 6 at the time of translation, , the bilingual translation is automatically switched.

An example of Jizuke 21 for translating from Japanese to English or American is shown in FIG. Now, in English mode, the translation of ``1st basement'' or 11th floor'' in Japanese is ``grand''.
floor”, and in Yoneichi it is “first floor”.
In addition, the translation of ``2nd floor'' in Japanese is ``first floor'' in English and 5th floor'' in American language. In this way, some of the 8 nods with different bilingual translations are translated in either the English or American mode specified by the dictionary 1 switching dictionary 6. In this way, accurate translation into the target language is achieved. Note that if neither mode is selected, the configuration may be configured so that translation is performed in the mode with the highest frequency.

FIG. 8 shows a screen in which the way the translation is done differs depending on the mode switching described above. In the figure (a), the English mode is specified, and in the figure (b), the American mode is specified, respectively.
Indicates the state in which translation has been executed.

In this way, it is possible to save labor in operation compared to a system that does not have a mode switch and that translates into either ``-dan'' or ``--ritsu'' and then corrects it later as necessary.

(Effects of the Invention) As described above, according to the present invention, when a Japanese language processing machine such as a Japanese word processor is combined with an electronic translation machine, and when Japanese is input and converted into a kanji-kana mixed sentence, the morphological The analysis determines whether or not the same sentence has multiple dependencies in terms of interpretation, and if there are multiple dependencies, a warning is displayed, so accurate translations can be obtained.
In addition to improving the accuracy of electronic translation, you can also
Compared to the case of re-entering Japanese, work efficiency is improved and operability can be improved.

[Brief explanation of the drawing]

Fig. 1 is a block diagram of an electronic translation machine equipped with a Japanese language processing function according to an embodiment of the present invention, Fig. 2 is a flowchart showing the operation of the machine, and Fig. 3 shows changes in the display screen of the machine. 4 is a block configuration diagram of an electronic translation machine showing a first application example of the present invention, FIG. 5 is a diagram showing various display screens of the same machine, and FIG. 6 is a diagram showing a second application example of the present invention. FIG. 7 is a diagram showing a part of the dictionary in the electronic translator, FIG. 8 is a diagram showing the display screen of the electronic translator,
FIG. 9 is a block diagram of a conventional Japanese word processor, and FIG. 10 is a block diagram of a conventional electronic translator. 1...Keyboard (input device), 4...Display device,
5... Printer (printing device), 12a... Kana-Kanji converter, 12b... Edit/display/mark 1ri1 control unit,
19a... Morphological analysis section, 19b... If translation section 1
? 0...Main part, 21...Kana-Kanji conversion/translation common dictionary. Patent Applicant: Brother Industries, Ltd. Agent: Patent Attorney: Etsushi Kotani, Patent Attorney: Chief 1) Seido: Patent Attorney: Ino Itaya No. 11YI Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8

Claims (1)

[Claims] 1. An input device for inputting kana character strings and function execution commands, a storage device storing a kana-kanji conversion dictionary and a translation dictionary, and converting the input kana character strings into kanji-kana mixed sentences based on the above-mentioned kana-kanji conversion dictionary. A kana-kanji conversion unit, a control unit that controls execution of each function of editing, display, and printing according to function execution commands, a display device that is driven by this control unit to display a character string, and a printing device that prints the same character string. ,
a morphological analysis unit that determines whether or not the converted kanji-kana-mixed sentence has multiple modifications, and when there are multiple modifications, displays a warning on the display device; An electronic translation machine equipped with a Japanese language processing function, characterized in that it is comprised of a translation section that converts a sentence into a translated sentence based on the above-mentioned translation dictionary.