JP2002041517A - Machine translation apparatus, machine translation method, and computer-readable storage medium storing machine translation program - Google Patents

Abstract

(57) [Summary] [Problem] To enable efficient construction of a user dictionary by using dictionary information of a headword already registered. SOLUTION: When a duplication key is operated, a headword (English word) registered so far, its part of speech and its translation (Japanese translation) are displayed on a display device, and a word to be copied by the user is displayed. Is selected (S63), and the dictionary information registered in the selected headword is obtained (S6).
7), present it on the registration screen with the value entered. The user inputs an English word in the headword field (S6).
8) Furthermore, it is confirmed whether the duplication source dictionary information is appropriate for the word to be registered this time, and if necessary, it is corrected (S70).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a machine translation apparatus, a machine translation method, and a computer-readable storage medium storing a machine translation program, and more particularly, to easily construct a user dictionary to be referred to when performing translation work. The present invention relates to a machine translation apparatus, a machine translation method, and a computer-readable storage medium storing a machine translation program.

[0002]

2. Description of the Related Art Conventionally, in this type of machine translation system, in order to perform a translation operation, a plurality of dictionaries most suitable for the translated document are used as needed. In such a dictionary, a basic translation (system) in which general translations (second language) and grammars for a wide range of phrases (first language) are registered as being preset in the machine translation system.
There is a dictionary, which can be set in the machine translation system in advance or selectively set as necessary, as a specialized dictionary in which words and translations used only in certain specific fields are registered, There is a user dictionary that can be edited independently.

[0003] Various types of dictionaries and dictionaries are used in order to perform a translation operation even when a manual translation is used instead of a machine translation. There is also a need for general learning dictionaries with abundant examples, specialized fields dictionaries with bilingual technical terms, specialized books, magazines, and papers. These are not only those that are commercially available, but also use the knowledge of translators and a glossary of terms originally collected. This situation is the same in machine translation, and the specialized dictionary and user dictionary described above are required.

[0004] Since an actual translation contains a considerable proportion of terms and phrases used by individual users, only the information of the basic dictionary and the specialized dictionary prepared in advance in the translation system is provided. Is not enough,
Unregistered words exist. Further, even if a word is registered in the basic dictionary or the specialized dictionary, the translated word is not always the one most suitable for the user. There are some inconveniences, such as fixed phrases and unique expressions used by a specific user, that are not translated literally, but are not what you want.

For example, the Japanese notation "Kama" includes "Kama", "Kama" and "Kama". Also, even if you say "kiln" in English, there are "kiln" for baking pottery, "oven" for baking bread, "furnance" for furnace, and "iron pot" for kettle. "Ca (u) ld"
There are various "bites" such as "ron", "sickle" and "hook" of "sickle". Furthermore, in terms of the sewing machine, a part that accommodates a bobbin around which a bobbin thread is wound and forms a seam with the bobbin thread and the needle thread by supplementing the loop of the needle thread is referred to as a "kama". It is called "shuttle". The example of the sewing machine is quite limited in the field of text, but when translating this field, the user can never be satisfied because a translation other than "shuttle" would be wrong if it appeared. It can be seen that even the notation "kama" used casually has a limited meaning depending on the type of text and can be variously translated.

That is, in order to perform a translation process, it is necessary to select and use a dictionary including a translation word corresponding to the type of text to be translated. The basic dictionaries and specialized dictionaries usually provided by the system are also general ones. Therefore, at the same time as selecting a plurality of specialized dictionaries, what is also needed is a user dictionary that can limit the use of words and phrases most. In some cases, it may be necessary to register many terms in the user dictionary. If you don't have a user dictionary, you can't translate well. Under such circumstances, registration of the user dictionary by the user himself is indispensable.

[0007]

However, registering an electronic dictionary for a machine translation system or the like at once, in a large amount, and in a state in which information is gathered requires a great deal of user's labor. I do. Translates very short documents interactively using a machine translation system and corrects the translated text on the spot, or is not suitable for unknown words or translations that were not registered in the system dictionary in the machine translation system Sometimes things are registered in the user dictionary in some convenient way. in this way,
High translation accuracy can be expected due to the user's time and effort, and it is sufficient to build a user dictionary for the words that appear as needed. However, the most efficient use of a machine translation system is batch processing utilizing its speed. The 1993 edition of the Asia-Pacific Association for Machine Translation edited the research report, stating, "We are trying to increase the amount of processing by limiting the parts of speech to increase overall efficiency."

[0008] The success of machine translation depends on the small number of unknown words and the enrichment of user dictionaries and specialized dictionaries in specific fields. Therefore, a large number of user dictionaries most suitable for the document must be constructed at an early stage. However, a dictionary for a machine translation system must have various kinds of knowledge for the system to translate instead of humans as dictionary information, and a list of terms for humans on hand is immediately expanded into a user dictionary. Can not. Translation processing may fail due to inadequate information in the user dictionary.
It must be a complete, credible, and full-word user dictionary, not only with translated words. That is, in the user dictionary, it is necessary to register various additional information according to the part of speech in addition to the translated word, and depending on the part of speech, the number of items of the additional information is quite large. At the same time, since the user dictionary is a dictionary for translating a specific field or document as a feature of the user dictionary, it is necessary to register a large number of words having the same semantic attribute especially in the translation of manuals.

To meet such needs, various machine dictionary registration methods have been developed for each machine translation system. For example, it extracts unknown words that have not been registered in the system dictionary of the machine translation system, shows the list to the user, assigns translations, registers them as a user dictionary, and includes user-created headwords. In a text file, special symbols and delimiters are inserted according to a predetermined format, and a user dictionary conversion program is started to collectively register.
In addition, some systems have only a tool for registering, one word at a time, the unknown words appearing in the original text and the headwords that are determined to require a user dictionary.

However, in the case of the method of registering words one by one, it is quite troublesome to register all items from 1 for a headword to be newly registered. Cheap. Therefore, some systems may learn the value selected immediately before that, and then press the return key continuously when registering the next word, but since only the information of the previous headword can be used, It is not enough for labor saving.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problem, and efficiently constructs a user dictionary by using dictionary information of a headword already registered at the time of registration of the user dictionary. The purpose is to make it possible.

[0012]

According to a first aspect of the present invention, there is provided a machine translation apparatus for translating an original sentence in a first language into a translation in a second language with reference to a dictionary. A user dictionary in which a headword in a first language, a translation in a second language corresponding to the headword, and auxiliary information used for translation processing of the headword are registered as one headword group, and the user dictionary And a dictionary registering means for storing a new headword group in the user dictionary. In the case where a new headword group is registered in the user dictionary by the dictionary registration means, one heading is selected from the headword groups already registered in the user dictionary. A selection means for selecting a word group and a new entry word group registration screen are displayed. Registration screen display control means for displaying an input field so that the input field can be input, wherein the dictionary registration means, a headword newly input in the headword input field, a translation word newly input in the translation word input field, The supplementary information of the headword group selected by the selection means is registered in a user dictionary.

[0013] In the machine translation apparatus according to this configuration, from among the headword groups that are already registered in the user dictionary, those having attached information that matches the attached information of the headword to be newly registered are selected. Since the additional information of the selected headword group is newly registered as the additional information of the headword to be newly registered, it is not usually necessary to input the auxiliary information covering a large number of items from scratch, and the input can be saved.

[0014] In the machine translation apparatus according to the second aspect of the present invention,
In addition to the configuration of the machine translation apparatus according to claim 1, the user dictionary stores the part of speech information of the headword, and the registration screen display control means controls the part of speech information of the headword group selected by the selection means. , A registration screen for a headword group based on the part of speech information is displayed.

In the machine translation apparatus according to this configuration, in addition to the operation of the machine translation apparatus according to claim 1, since the user dictionary stores part-of-speech information, a screen for inputting different auxiliary information for each part of speech. Is prepared, and the user inputs the information in accordance therewith, whereby the necessary additional information is accumulated for each part of speech, and a detailed user dictionary can be constructed.

According to a third aspect of the invention, there is provided a machine translation apparatus.
In addition to the configuration of the machine translation apparatus according to claim 1 or 2, the registration screen display control means displays a headword entry field and its translation term entry field as blank so that they can be entered.

In the machine translating apparatus according to this configuration, in addition to the operation of the machine translating apparatus according to claim 1 or 2, a registered head having attached information that matches the attached information of a headword to be newly registered is provided. Even if a word group is selected, it is usually necessary to newly enter a headword and its translated word. Therefore, by leaving these blank and accepting the input, the user's labor is further reduced and the input is reduced. be able to.

According to the machine translation apparatus of the invention according to claim 4,
In addition to the configuration of the machine translation apparatus according to any one of claims 1 to 3, the registration screen display control means inputs the attached information of the headword group selected by the selecting means in an attached information input field. It is characterized in that it is displayed in a state in which it is held.

In the machine translation device according to this configuration, in addition to the operation of the machine translation device according to any one of claims 1 to 3, the additional information of the selected registered headword group is newly registered. Users can register it as an adjunct to the dashi to check if there are any problems,
A highly accurate user dictionary can be constructed.

In the machine translation apparatus according to the fifth aspect,
In addition to the configuration of the machine translation device according to any one of claims 1 to 4, in the registration screen, the content of an input field for additional information is:
It can be changed.

In the machine translation apparatus according to this configuration, in addition to the operation of the machine translation apparatus according to any one of claims 1 to 4, a registered machine having attached information that coincides with attached information of a headword to be newly registered. Even if there is no headword group, it is possible to select a registered headword group with similar auxiliary information, modify a part of the headword group, and register it. Can be saved.

According to the machine translation apparatus of the invention according to claim 6,
In addition to the configuration of the machine translation device according to any one of claims 1 to 5, the registration screen display control means displays a plurality of related registration screens when one headword group is registered. .

In the machine translation device according to this configuration, in addition to the operation of the machine translation device according to any one of claims 1 to 5, when the supplementary information of the headword to be newly registered covers a large number of items, a plurality of items are added. By accepting the input separately on the registration screen, the user dictionary can be constructed easily and easily.

According to the machine translation method of the invention according to claim 7,
A machine translation method for translating an original sentence in a first language into a translation in a second language with reference to a dictionary, comprising: a headword in the first language; a translation in the second language corresponding to the headword; and translation of the headword. A registration step of registering the auxiliary information used for the processing as one headword group in the user dictionary; and, when registering a new headword group in the user dictionary, one of the headword groups already registered in the user dictionary. A selection step of receiving selection of one headword group, a display step of displaying a registration screen of a new headword group, and displaying a headword input field and a translation word input field on the registration screen so that the headword input field and its translation word input field can be input. A headword entered in the headword input field, a translation word newly input in the translation word input field, and a headword group selected by the selection means Characterized in that the secondary information is composed of a replication registration step of registering the user dictionary.

In the machine translation method according to this configuration, from among the headword groups that have already been registered in the user dictionary, those that have attached information that matches the attached information of the headword to be newly registered are selected, Since the additional information of the selected headword group is newly registered as the additional information of the headword to be newly registered, it is not usually necessary to input the auxiliary information covering a large number of items from scratch, and the input can be saved.

In the machine translation method according to the present invention,
In addition to the configuration of the machine translation method according to claim 7, the registration step registers part-of-speech information of the headword in a user dictionary, and the display step performs a part-of-speech of a headword group selected by the selection step. On the basis of the information, a registration screen of a headword group based on the part of speech information is displayed.

In the machine translation method according to this configuration, in addition to the operation of the machine translation method according to claim 7, part of speech information is registered in a user dictionary, and different attached information is input for each part of speech based on the part of speech information. Screen is prepared, and the user can enter
Necessary attached information is accumulated for each part of speech, and a detailed user dictionary can be constructed.

In the machine translation method according to the ninth aspect,
In addition to the configuration of the machine translation method according to claim 7 or 8, the display step displays a headword entry field and its translation term entry field as blank so that they can be entered.

[0029] In the machine translation method according to this configuration, in addition to the operation of the machine translation method according to claim 7 or 8, a registered head having additional information that matches the additional information of the headword to be newly registered. Even if a word group is selected, it is usually necessary to newly enter a headword and its translated word. Therefore, by leaving these blank and accepting the input, the user's labor is further reduced and the input is reduced. be able to.

According to a tenth aspect of the present invention, in addition to the configuration of the machine translation method according to any one of the seventh to ninth aspects, the display step includes the step of: It is characterized in that it is displayed in a state where the attached information of the selected headword group has been input.

In the machine translation method according to this configuration, in addition to the operation of the machine translation method according to any one of claims 7 to 9, additional information of the selected registered headword group to be newly registered is added. Users can register it as an adjunct to the dashi to check if there are any problems,
A highly accurate user dictionary can be constructed.

According to an eleventh aspect of the present invention, in addition to the configuration of the machine translation method according to any one of the seventh to tenth aspects, the contents of the input field of the auxiliary information displayed on the registration screen are changed. And a step of receiving the information.

In the machine translation method according to this configuration, in addition to the operation of the machine translation method according to any one of claims 7 to 10, the registered machine having the additional information that matches the additional information of the headword to be newly registered is registered. Even if there is no headword group, it is possible to select a registered headword group with similar auxiliary information, modify a part of the headword group, and register it. Can be saved.

According to a twelfth aspect of the present invention, a computer-readable storage medium storing a machine translation program stores a machine translation program for translating an original sentence in a first language by a computer into a translation in a second language with reference to a dictionary. Computer-readable storage medium, wherein the headword in the first language, the translation in the second language corresponding to the headword, and the auxiliary information used in the translation process of the headword, as one headword group A registration step of registering in the user dictionary, and a selection step of receiving a selection of one headword group from the headword groups already registered in the user dictionary when registering a new headword group in the user dictionary. The entry screen for the entry group is displayed, and the entry screen A display step of displaying the translation word input field as an empty field so that the translation word can be input, a headword newly input in the headword input field, a translation word newly input in the translation word input field, and a translation selected by the selection means. And causing the computer to execute a copy registration step of registering the supplementary information of the dashi word group and the user dictionary.

In the computer-readable storage medium storing the machine translation program according to this configuration, a headword group that has already been registered in the user dictionary matches with the additional information of the headword to be newly registered. Select the one with the attached information and register the attached information of the selected headword group newly as it is as the attached information of the headword to be newly registered.Therefore, it is not necessary to input the attached information over many items from the beginning, Labor saving of input is achieved.

According to a thirteenth aspect of the present invention, in the computer readable storage medium storing the machine translation program, in addition to the configuration of the computer readable storage medium storing the machine translation program according to the twelfth aspect, the registration step Registers the part-of-speech information of the headword in a user dictionary, and based on the part-of-speech information of the headword group selected by the selection step, displays the headword group registration screen based on the part-of-speech information. It is characterized by displaying.

In the computer-readable storage medium storing the machine translation program according to this configuration, in addition to the operation of the computer-readable storage medium storing the machine translation program according to claim 12, the part-of-speech information is stored in the user dictionary. A screen for registering and inputting additional information different for each part of speech based on the part of speech information is prepared.
Necessary attached information is accumulated for each part of speech, and a detailed user dictionary can be constructed.

According to a fourteenth aspect of the present invention, there is provided a computer readable storage medium storing the machine translation program, wherein the computer readable storage medium stores the machine translation program according to the twelfth or thirteenth aspect. The display step is characterized in that the headword entry field and its translated word entry field are displayed as blank so that they can be entered.

In the computer-readable storage medium storing the machine translation program according to this configuration, in addition to the function of the computer-readable storage medium storing the machine translation program according to claim 12, a new registration is desired. Even if you select a registered headword group that has ancillary information that matches the ancillary information of the headword, the headword and its translated word usually need to be newly entered. By accepting,
The labor of the user can be reduced by further reducing the user's labor.

According to a fifteenth aspect of the present invention, there is provided a computer-readable storage medium storing a machine translation program according to any one of the twelfth to fourteenth aspects. In addition, the display step is characterized in that the display is performed in a state where the auxiliary information of the headword group selected by the selection means is input in the input field of the auxiliary information.

A computer-readable storage medium storing a machine translation program according to this configuration has a computer-readable storage medium storing a machine translation program according to any one of claims 12 to 14, and further includes a computer-readable storage medium. To create a high-accuracy user dictionary that allows users to confirm whether there is no problem by registering additional information of registered registered headword groups as additional information of headwords that they want to register. Can be.

According to a sixteenth aspect of the present invention, there is provided a computer readable storage medium storing a machine translation program according to any one of the twelfth to fifteenth aspects. In addition, the method further includes a step of receiving a change in the content of the input field for the additional information displayed on the registration screen.

A computer-readable storage medium storing a machine translation program according to this configuration has a new function in addition to the function of the computer-readable storage medium storing the machine translation program according to any one of claims 12 to 15. Even if there is no registered headword group that has ancillary information that matches the ancillary information of the headword that you want to register, select a registered headword group that has similar ancillary information, and Correction and registration can be performed, and the labor of the user can be further reduced, and the labor of input can be saved.

[0044]

BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. First, FIG. 1 is a block diagram showing a schematic configuration of an English-Japanese machine translation apparatus according to an embodiment to which the present invention is applied.

As shown in FIG. 1, the machine translation apparatus of the present embodiment is used for an input device 21 such as a keyboard for inputting English sentences to be translated and various operation commands, and for translation processing. An external storage device 24 composed of a readable / writable hard disk or optical disk storing various dictionaries (a system dictionary and a user dictionary), and a CRT for displaying the input English original and the translated Japanese translation as a translation result , A printing device 23 such as a laser printer for printing and outputting a translation of Japanese as a translation result, and a data input from an input device 21 and an external storage device 24. And a CPU 10 for performing control such as displaying the result on the display device 22 and outputting the result from the printing device 23, and various programs. And a ROM 40, and a RAM50 Metropolitan storing information and dictionary information input as needed.

Here, the input device 21 has a character key for inputting characters, an input sentence edit key for starting various processes, a translation execution key, a dictionary edit key, an output sentence edit key, an end key, and a print key. Various function keys such as an execution key, a cursor key, an insert key, a delete key, an enter key, and a next candidate selection as an operating tool for editing an input sentence or an output sentence, or operating a screen displayed on the display device 22 or the like. A key and the like are provided.

The external storage device 24 stores, as dictionaries used in the translation processing, a system dictionary whose contents are stored in advance by the translation device and a user dictionary that is independently constructed by the user. Note that a plurality of dictionaries may exist simultaneously for each user or for each field of use, for example. Here, one use is made for convenience. The system dictionary is write-protected to prevent reorganization by the user.

FIG. 2 and FIG. 3 each show an example of the data structure of the dictionary. FIG. 2 shows the data configuration of the system dictionary, and FIG. 3 shows the data configuration of the user dictionary. Each dictionary includes headings 60 and 70, translations 62 and 72 for headings 60 and 70, parts of speech 64 and 74,
It is composed of various dictionary information 66 and 76 that differ depending on the part of speech. As will be described later, the dictionary information represents information on the use of the translated word (in this case, Japanese), wording, semantic classification, and the like, and is indispensable at the time of translation.

Next, the control unit of the machine translation apparatus according to the present embodiment comprises a CPU 10, a ROM 40, a RAM 50, an input / output port 20, and a bus line connecting these units.

The display control program 4 is stored in the ROM 40.
1, character string input program 42, translation program 43,
User dictionary registration program 44, copy registration (dictionary copy)
The program 45 is stored. Also, the RAM 50
Is composed of an input character string storage area 51, a display character string storage area 52, a translation system dictionary storage area 53, and a user dictionary storage area 54.

In the CPU 10, the ROM 4
0 by the translation program 43 stored in the input device 2
In addition to translating the English original sentence input from No. 1 into a Japanese translated sentence, translation processing and editing for editing the original sentence and translated sentence are performed. The user dictionary registration program 44 executes a dictionary editing process for constructing a user dictionary to be used at the time of translation, a duplication registration (dictionary copy) program 45 executes a user dictionary registration process using dictionary information for existing headings, and the like. .

In the RAM 50, the English original sentence input from the input unit is stored in the input character string storage area 51, and the Japanese translation translated and edited by the translation editing process is stored in the display character string storage area 52. In addition, in the user dictionary registration process, the dictionary information of the user dictionary is stored in the user dictionary storage area 54.

Next, the translation editing process executed by the control unit 10 will be described with reference to the flowcharts shown in FIGS. Each step of the flowchart is abbreviated as “S” below.

As shown in FIG. 4, when the translation editing process is executed, first, in S11, the user inputs the input device 2
One character key or the like is operated to wait for the input of an English original (first language) to be translated. Here, the input of the original text may be performed by reading an already created original text file. Then, when an English original sentence (hereinafter, input sentence) is input, the process proceeds to S12, and the input sentence is displayed in a predetermined area of the display device 22.

In the following S13, it is determined whether or not the input sentence editing key provided on the input device 21 has been operated, and if not, the flow proceeds to S15. If the input sentence editing key has been operated, the flow shifts to S14, where the input sentence editing process is executed. When the input sentence editing process is executed,
The user can directly correct an input sentence displayed in a predetermined area of the display device 22 using a cursor key, a delete key, an insert key, a character key, or the like provided on the input device 21.

Next, in S15, it is determined whether or not the translation execution key provided on the input device 21 has been operated, and if not, the process proceeds to S19. If the translation execution key has been operated, the process proceeds to S16, where a translation process (to be described in detail later) for translating an English input sentence into a Japanese (second language) translated sentence (hereinafter, output sentence) is executed. At S17, the output sentence obtained by the translation process is paired with the input sentence and
In step S18, an input sentence and an output sentence are displayed in predetermined areas provided on the display device 22, respectively. The translation process may be executed collectively for all the input sentences, or a range may be designated for each sentence, and only the sentences in the designated range may be executed.

In S19, it is determined whether or not the user dictionary display function key provided on the input device 21 has been operated.
If the operation has not been performed, the process proceeds to S21. If the user dictionary display function key has been operated, a list of the user dictionaries is displayed in S25 and the process proceeds to S20 to execute a user dictionary editing process. When the user dictionary editing process is executed, the user can correct or delete the contents of the information registered in the user dictionary, or register a new necessary phrase.

In S21, it is determined whether or not the output sentence edit key provided on the input device 21 has been operated, and if not, the flow proceeds to S23. If the output sentence editing key has been operated, the flow shifts to S22 to execute the output sentence editing process. When the output sentence editing process is executed, the user executes the S16 translation process to convert the output sentence (Japanese translation) displayed in a predetermined area of the display device 22 into the input sentence of S14. As with the editing process, the input device 2
1 can be directly corrected using a cursor key, a delete key, an insert key, a character key, etc.

In S23, it is determined whether or not the end key provided on the input device 21 has been operated, and if not, the process proceeds to S24. For example, the translated result is output to the printing device 23 and printed. A key operation such as a print execution key to be performed is detected, and a process corresponding to the operated key is executed.

Then, S14, S18, S20, S
After the processing of S22 and S24, the process returns to S13, and these processings are repeatedly executed. If it is determined in S23 that the end key has been operated, the present process is terminated.

That is, in the translation editing process, after the English original is input, the input sentence editing process, the translation process, the user dictionary editing process, and the output sentence editing process are performed in accordance with the operation of various keys provided on the input device 21. Are performed. Then, if the translation process is executed by operating the translation execution key, the obtained output sentence (Japanese translation)
If the input text is incorrect or unnatural, operate the input text edit key to execute the input text edit process and re-edit the input English original text, or display the user dictionary Use the key to execute the user dictionary editing process, change the contents of the user dictionary, and then operate the translation execution key again to re-translate, or operate the output sentence editing key to execute the output sentence editing process. Execute and directly edit the Japanese translation that is the output sentence. If the translation result desired by the user can be obtained by repeating such processing, the print key is operated and the translation result is transmitted to the printing apparatus 2.
After the output from No. 3, the end key is operated to end the processing.

Next, a description will be given of a translation process executed in S16 to translate an input sentence (English original sentence) into an output sentence (Japanese translated sentence) with reference to the flowchart shown in FIG.

First, in S52, using a predetermined morphological analysis rule, a dictionary is searched for words constituting an input sentence, and words are cut out, and search information such as translated words and parts of speech are added to the cut out words. Execute morphological analysis processing. Here, regarding the dictionary search for the same word, the user dictionary has the highest priority. Next, the dictionaries are used in the order of the system dictionaries. In S53, the result of the morphological analysis process is received, and a syntax analysis process for analyzing the English structure is executed using a predetermined analysis grammar stored in the ROM 40.

In step S54, the English structure analyzed by the analysis process is converted into the Japanese structure using the predetermined conversion grammar stored in the ROM 40, and each word is replaced with the Japanese language. Execute the translation process. In S55, using a predetermined generation grammar stored in the ROM 40, the word order of the Japanese language is determined from the Japanese structure converted by the conversion process, and the tense, aspect, person, numerical feature, and the like are determined. Processing is performed to execute candidate generation processing for generating an output sentence composed of a Japanese word string. Then, in S56, an inflected form such as a past tense of a verb is generated for a Japanese word string generated by the generation process using a predetermined morpheme generation rule stored in the ROM 20 and output. The morpheme generation process for obtaining the sentence is executed, and the translation process ends.

Next, the user dictionary editing process executed in step S20 in which the user corrects the contents of the user dictionary or registers a new necessary word / phrase will be described with reference to the flowchart shown in FIG. 6 and FIGS. Explanation will be given along the screen display shown.

In the user dictionary editing process, the user selects a word to be edited after inputting a user dictionary display function key to display a user dictionary screen 80 for displaying a list of user dictionary words as shown in FIG. Start from (S61)
Is done. In FIG. 8, initially, a word at a predetermined position (eg,
The cursor is located at the top word (word) and is highlighted. This highlighted word means that it is in a selected state as a target of editing processing, and the cursor position is changed by operating a cursor key or a mouse to select a word as a target of editing processing again (S
62).

When the object to be edited is determined by the above processing, it is determined in the next step whether or not to make a copy registration (S63). Here, if the duplicate registration is not performed, the process proceeds to the next step to determine whether or not to perform the new registration (S6).
4). In the case of new registration, as shown in FIG. 7, a user dictionary screen 80 is displayed, a menu registration button 81 is operated, and a part-of-speech selection screen 82 of an English word (disambiguation word) to be registered is presented by a pull-down menu. (S7
1). When the part of speech is selected, a registration screen 83 for the designated part of speech as shown in FIG. 9 is presented, and waits for the input of an English word to be found (S72). FIG.
12 is a registration screen 83 when a noun is selected. This means that, for new registration, the headword column 101 and the translated word column 102 are blank, and each item of the dictionary information is presented in a state where none is selected. And, by the user,
(English word) is input (S72), a translated word (Japanese translation) is input (S73), and dictionary information is input (S7).
4). This dictionary information is additional information used for the translation processing.

Here, the dictionary information for nouns includes “person”, “organization”, “animal”, “plant”, “parts”, “tools”, “food”, “vehicles”, “semantic classification information”. “Documents”, “Materials”, “Money”, “Generations”, “
Phenomenon "," action "," unit "," property "," time ","
There are sub-names such as “Place name”, “Place”, “Intellectual product”, “Emotion”, and “Event”, and the applicable item is selected from these items. In addition, there are specifics of “countable”, “countable”, and “always plural” as countable / countable information.
One of them is selected. Furthermore, there are sub-types of "rule" and "irregular" as plural types, and one of them is selected. In the case of "rule", "-s", "
−es ”,“ y → es ”,“ f → ves ”and“ fe →
ves ". One of the sub-classes of nouns is" normal noun "and" proper noun ", and either one is selected. A counting method (for example, in the case of a person, "person") is inputted and registered. As described above, with respect to nouns, the dictionary information includes only five information units of semantic classification, countable / uncountable, plural types, subclassification, and counting. However, for different parts of speech, the dictionary information is a combination of information units different from the nouns. The details of this information unit become dictionary information for each heading.

When the user operates the registration key,
These input information are registered in the user dictionary (S75),
The display returns to the user dictionary list display screen as shown in FIG.

On the other hand, in the case of copy registration, when the copy key is operated, a copy registration process is performed on the word selected in S61 or S62. For example, as shown in FIG.
When the word "tumoto" is selected, the dictionary information registered in the headword "Matsumoto" is obtained (S67), and the dictionary information is presented on the registration screen 84 as shown in FIG. . In this registration screen, all dictionary information of "Matsumoto" is displayed, and the headword column 101, the translated word column 102, and the main noun column 103 are blank to prompt the user to input. Then, as shown in FIG. 11, the user inputs an English word into the heading word field 101 of the registration screen 84 (S68), and inputs a Japanese translation into the translation word field 102 (S69). Here, the headword column 101, the translated word column 102, and the main noun column are displayed in blank spaces, but may be displayed in a state where the information of the duplication source is input so as to be correctable. For example, when registering derivative words, etc.,
This saves input and has an effect that the source of the copy can be easily understood at a glance.

Further, the user confirms whether the copy source dictionary information is appropriate for the word to be registered this time, and corrects it if necessary (S70). Here, headword column 1
As a screen display for inputting only 01 and the translation field 102, the screen may be simplified without displaying dictionary information. Further, the dictionary information may be displayed so that the user can confirm it, but the correction may not be accepted. In these cases, it is necessary to select a duplication source whose dictionary information completely matches, and some of the corrections cannot be made.

In the example of FIG. 11, "kobayashi" is entered in the entry word field 101 and "Kobayashi" in the translation word field 102 of the registration screen 84, and the dictionary information may be the same as that of "Matsumoto" of the copy source. After confirming these, the registration key is operated. When the user operates the registration key, the input information is registered in the user dictionary (S75), and the display returns to the user dictionary list display screen as shown in FIG.

As described above, when the copy registration is selected, the user does not need to input the details of the dictionary information each time, and can register a new word in the user dictionary in a short time. In addition, since it is possible to select not only the word registered immediately before but also a word registered before, a word to be a copy source, it is possible to select a word whose dictionary information completely matches, In addition, if the dictionary information can be modified, even if there is no exact match, it is possible to select words that can be dealt with by modifying only a part of the dictionary information. As the number increases, the input can be saved to a considerable extent.

By the way, when it is determined whether or not to perform new registration (S64), if new registration is not performed, it is determined whether or not an end key has been input (S65). Here, if there is an input of the end key, the user dictionary editing process ends. On the other hand, if there is no input of the end key, processing of keys other than registration, for example, deletion of already registered words and change of registered contents are performed (S66), and a list of user dictionaries as shown in FIG. 8 is displayed again. Return to the screen.

In the above description, the case where a noun is selected in S71 is described. However, since the dictionary information held by the part of speech is different, the displayed registration screen is also different as follows. First, FIG. 12 shows a registration screen 85 when a verb is selected in S71. The verb has many details in the dictionary information, and the required details differ depending on the type. Therefore, the verb is displayed on two pages, and an appropriate second page is displayed based on the value selected on the first page. . FIG. 12 shows the registration screen 85 for the first page, and FIGS. 13 and 14 show the registration screens 86 and 87 for the second page.

On the registration screen 85 of the first page shown in FIG.
Enter the items required for all verbs. Headword column 13
1 (S68 / S72), enter the plural word type (rule / irregular) as the dictionary information, delete the present participle (-ing, e, delete consonant, ie → yin)
g), tangen (-s, -es, y → ies), English syntax
(Intransitive, transitive, self / transitive, dual object (to),
Input a dual object (for) (S70 / S7)
4). When the "next" key is operated after inputting all of these items, when the self / transitive verb is selected in the English syntax, a registration screen 87 on the second page shown in FIG. Fig. 1 when an item other than the transitive verb is selected
A second page registration screen 86 shown in FIG. 3 is displayed.

On the second page registration screen 86 shown in FIG.
A Japanese translation is entered in the translation word column 132 (S69 / S73),
The inflected form is entered in the inflected form field 134 and the modifier required for Japanese translation is further entered in the modifier phrase field 133. Further, as the dictionary information, a Japanese syntax and the presence or absence of intention are selected and input (S70 / S74). On the registration screen 87 of the second page shown in FIG. 14, similar items are input for the intransitive verb and the transitive verb (S70 / S74).
As described above, in the case of a verb, the step of inputting a translated word (S69 / S73) is performed in the step of inputting dictionary information (S7).
0 / S74).

FIG. 15 shows an adjective dictionary registration screen 88. Items common to the other parts of speech include a headword field 141 and a translated word field 142, and as dictionary information items, a modifier word field 143 and Japanese syntax options are prepared as in the case of verbs. As dictionary information items peculiar to adjectives, the presence / absence of comparability, comparative / superlative patterns, and English syntax are selected. As in the case of the noun, the user inputs an English word to be registered in the headword field 141 (S68 / S72), and
The Japanese translation is input to 42 (S69 / S73). In the case of new registration, dictionary information is newly input (S74), and in the case of duplicate registration, the presented dictionary information is registered as it is or after the user confirms and corrects (S70), the registration is performed. Registered by key.

FIG. 16 shows a dictionary registration screen 89 for adverbs. Items common to other parts of speech include an English word input field 151 serving as a heading and a Japanese translation field 152. Items of dictionary information include, as in the case of adjectives, the presence or absence of comparability, An advanced pattern is selected, and as a dictionary information item peculiar to an adverb, whether the subject word is a verb, an adjective, or an adverb is selected. As in the case of the noun, the user inputs an English word to be registered in the headword field 151 (S68 / S72), and inputs a Japanese translation in the translated word field 152 (S69 / S73). In the case of new registration, dictionary information is newly input (S74). In the case of duplicate registration, the presented dictionary information is registered as it is or after the user confirms and corrects it (S70), It is registered with a registration key.

FIG. 17 shows a dictionary registration screen 90 for the initial connective.
It is shown. English word input field 16 to be headline
1. There is no particular item for inputting dictionary information other than the Japanese translation column 162, and if there is a matter to be particularly noted, it is made to correspond in the memo column. The user inputs an English word to be registered in the headword field 161 (S68 / S72), and inputs a Japanese translation in the translated word field 162 (S69 / S73).

FIG. 18 shows a dictionary registration screen 91 for classifiers. In addition to the English word input field 171 and the Japanese translation field 172, which are headings, as dictionary information, it is only necessary to select whether or not a hyphen precedes an English word. I correspond. As in the case of the noun, the user inputs an English word to be registered in the headword field 171 (S68 / S72), and the translated word field 172.
Is input (S69 / S73). In the case of new registration, dictionary information is newly input (S74). In the case of duplicate registration, the presented dictionary information is registered as it is or after the user confirms and corrects it (S70), It is registered with a registration key.

FIG. 19 shows a preposition dictionary registration screen 92. In addition to the English word input field 181 serving as a headline and the Japanese translation field 182, as the dictionary information, the user is allowed to select whether the qualified word is a declinable word or the noun immediately before. Thing (form, location,
Time, content, and partner). As in the case of the noun, the user inputs an English word to be registered in the headword field 181 (S68 / S72), and the translation field 1
A Japanese translation is input to 82 (S69 / S73). In the case of new registration, dictionary information is newly input (S74). In the case of duplicate registration, the presented dictionary information is registered as it is or after the user confirms and corrects it (S70), It is registered with a registration key.

Next, a second embodiment to which the present invention is applied will be described with reference to FIG. FIG. 20 is a block diagram showing a schematic configuration in a case where machine translation is performed on a general-purpose computer using a CD-ROM which is a storage medium storing the English-Japanese machine translation program of the present embodiment.

As shown in FIG. 20, in the present embodiment, various programs necessary for machine translation such as a translation program 243, a user dictionary registration program 244, a copy registration (dictionary copy) program 245, and a translation process. Various dictionaries (system dictionary 246 and user dictionary 247) to be used are stored in the CD-ROM 240.
The CD-ROM 240 is set in the CD drive 225 and read by the computer. The computer used in the present embodiment includes an input device 221 composed of a keyboard or the like for inputting English sentences to be translated and various operation instructions, and various dictionaries and various programs used for the translation processing. And a read / write hard disk 224 copied and stored from a PC, a display device 222 composed of a CRT or the like for displaying the input English original text and the translated Japanese text as the translation result, and the Japanese translation text as the translation result. A printing device 223 such as a laser printer for printing and outputting a translation and the like, and a translation process based on data input from the input device 221 and the external storage device 224, and displaying and printing the result on the display device 222 C for performing control such as output from the device 223
It comprises a PU 210 and a RAM 250 for storing input information and dictionary information as needed.

Here, the input device 221 has a character key for inputting characters, an input sentence edit key for starting various processes, a translation execution key, a dictionary edit key, an output sentence edit key, an end key, and a print key. Various function keys such as an execution key, a cursor key, an insert key, a delete key, an enter key, and a next candidate selection as an operating tool for editing an input sentence or an output sentence, or operating a screen displayed on the display device 222, or the like. A key and the like are provided.

The CD-ROM 240 stores a system dictionary 246 prepared in advance and a user dictionary 247 independently constructed by a user as dictionaries used for translation processing. Note that a plurality of dictionaries may exist simultaneously for each user or for each field of use, for example. Here, one use is made for convenience. Further, the system dictionary 246 is write-protected to prevent reorganization by the user. These dictionaries are directly stored in the CD-ROM 240 during the translation process.
Alternatively, it may be used by reading from the server, or may be copied to the hard disk 224 in advance and used. The data structures of the system dictionary 246 and the user dictionary 247 are the same as those in the first embodiment, and a description thereof will not be repeated.

The RAM 250 stores the input character string storage area 2
51, a display character string storage area 252, a translation system dictionary storage area 253, and a user dictionary storage area 254. Then, in the CPU 210, C
The translation program 243 stored in the D-ROM 240 translates an English original sentence input from the input device 221 into a Japanese translation, and also performs translation processing and editing for editing the original and translation. The user dictionary registration program 244 executes a dictionary editing process for constructing a user dictionary to be used at the time of translation, and a duplication registration (dictionary copy) program 245 executes a user dictionary registration process using dictionary information for existing headings. .

In the RAM 250, the English original sentence input from the input unit is stored in the input character string storage area 251 and the Japanese translation translated and edited by the translation and editing processing is stored in the display character string storage area 252. In addition, in the user dictionary registration process, the dictionary information of the user dictionary is stored in the user dictionary storage area 254.

The operations of the translation editing process, the translation process, and the user dictionary editing process in the present embodiment are the same as those in the first embodiment, and therefore, the description is omitted.

In this embodiment, the CD-ROM 2
Copy registration (dictionary copy) program 245 stored in 40
By doing so, it is possible to perform duplicate registration in user dictionary registration, so that the user does not need to input details of dictionary information one by one, and can register a new word in the user dictionary in a short time. In addition, since it is possible to select not only the word registered immediately before but also a word registered before, a word to be a copy source, it is possible to select a word whose dictionary information completely matches, In addition, if the dictionary information can be modified, even if there is no exact match, it is possible to select words that can be dealt with by modifying only a part of the dictionary information. As the number increases, the input can be saved to a considerable extent.

In this embodiment, various programs and dictionaries used for translation processing are stored on the CD-ROM, but may be distributed via a network such as the Internet.

[0093]

As is apparent from the above description, claim 1
In the machine translation device of the invention according to the invention, from among the headword groups already registered in the user dictionary, those having additional information matching the additional information of the headword to be newly registered are selected, and the selected word Since the additional information of the dashi group is newly registered as the additional information of the headword to be newly registered, it is not necessary to input the additional information covering a large number of items from the beginning, and this has an effect that the input can be saved.

In the machine translation apparatus according to the second aspect of the present invention,
In addition to the effects of the machine translation device according to claim 1, since the user dictionary stores part-of-speech information, a screen for inputting additional information that is different for each part-of-speech is prepared. In addition, necessary auxiliary information is accumulated for each part of speech, and a detailed user dictionary can be constructed.

In the machine translation apparatus according to the third aspect,
In addition to the effects of the machine translation device according to claim 1 or 2, even if a registered headword group having attached information that matches the attached information of the headword to be newly registered is selected, Usually, it is necessary to enter new translations, so by leaving these blank and accepting input,
The labor of the user can be reduced by further reducing the user's labor.

In the machine translation apparatus according to the fourth aspect of the present invention,
In addition to the effects of the machine translation device according to any one of claims 1 to 3, there is no problem in registering the attached information of the selected registered headword group as the additional information of the headword to be newly registered. The user can check whether or not the user dictionary can be constructed.

In the machine translation apparatus according to the fifth aspect,
In addition to the effects of the machine translation apparatus according to any one of claims 1 to 4, even if there is no registered headword group having attached information that matches the attached information of the headword to be newly registered, similarity is obtained. The registered headword group having the attached information can be selected, a part of the headword group can be corrected and registered, and the labor of the user can be further reduced and the input can be saved.

[0098] In the machine translation apparatus according to claim 6,
In addition to the effects of the machine translation device according to any one of claims 1 to 5, when the supplementary information of the headword to be newly registered covers a large number of items, the input is divided into a plurality of registration screens and the input is understood. A user dictionary can be constructed easily and easily.

According to the machine translation method of the invention according to claim 7,
From the headword groups that are already registered in the user dictionary, select the one that has ancillary information that matches the auxiliary information of the headword that you want to newly register, and newly add the auxiliary information of the selected headword group. Since the heading information to be registered is directly registered as the auxiliary information, it is not necessary to input the auxiliary information covering a large number of items from the beginning, so that there is an effect that the input can be saved.

In the machine translation method according to the present invention,
In addition to the effect of the machine translation method according to claim 7, a screen is provided for registering part-of-speech information in a user dictionary and allowing the user to input different attached information for each part of speech based on the part-of-speech information. Therefore, by inputting, necessary auxiliary information is accumulated for each part of speech, and a detailed user dictionary can be constructed.

In the machine translation method according to the ninth aspect,
In addition to the effects of the machine translation method according to claim 7 or 8, even if a registered headword group having attached information that matches the attached information of the headword to be newly registered is selected, Usually, it is necessary to enter new translations, so by leaving these blank and accepting input,
The labor of the user can be reduced by further reducing the user's labor.

According to a tenth aspect of the present invention, in addition to the effects of the machine translation method according to any of the seventh to ninth aspects, additional information on the selected registered headword group is newly added. The user can register whether or not there is any problem by registering the information as a headword to be registered, and a highly accurate user dictionary can be constructed.

According to an eleventh aspect of the present invention, in addition to the effects of the machine translation method according to any one of the seventh to tenth aspects, additional information that matches the additional information of a headword to be newly registered is added. Even if there is no registered headword group that has a registered headword group that has similar supplementary information, the registered headword group can be selected and part of it can be corrected and registered, further reducing the user's labor. As a result, the input can be saved.

In the computer-readable storage medium storing the machine translation program according to the twelfth aspect of the present invention, the headword information to be newly registered is selected from the headword groups already registered in the user dictionary. It is necessary to input the appendix information covering many items from the beginning to select the one with the appendix information that matches with and to register the appendix information of the selected headword group as the appendix information of the headword to be newly registered. Therefore, there is an effect that the input can be saved.

According to a thirteenth aspect of the present invention, a computer readable storage medium storing a machine translation program has the effect of a computer readable storage medium storing a machine translation program according to the twelfth aspect, and has a user dictionary. Since a screen for registering part-of-speech information and allowing the user to input different attached information for each part of speech based on the part-of-speech information is prepared, by inputting according to the user, the attached information required for each part of speech is accumulated, You can build a detailed user dictionary.

According to the fourteenth aspect of the present invention, a computer readable storage medium storing a machine translation program has a new effect in addition to the effect of the computer readable storage medium storing the machine translation program according to the twelfth or thirteenth aspect. Even if you select a registered headword group that has ancillary information that matches the annex information of the headword that you want to register, leave the headword blank and its translation blank, since it is usually necessary to enter a new one. By accepting the input, the labor of the user can be further reduced and the labor of the input can be reduced.

According to a fifteenth aspect of the present invention, there is provided a computer readable storage medium storing a machine translation program according to any one of the twelfth to fourteenth aspects. In addition, the user can confirm whether there is no problem by registering the supplementary information of the selected registered headword group as the supplementary information of the headword to be newly registered. Can be built.

According to a sixteenth aspect of the present invention, a computer readable storage medium storing a machine translation program provides an effect of the computer readable storage medium storing the machine translation program according to any one of the twelfth to fifteenth aspects. In addition, even if there is no registered headword group that has ancillary information that matches the additional information of the headword to be newly registered, select a registered headword group that has similar auxiliary information and select A part of the information can be corrected and registered, and the labor of the user can be further reduced, and the input can be saved.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an overall configuration of a machine translation device according to a first embodiment.

FIG. 2 is an explanatory diagram showing a configuration of system dictionary data stored in an external storage device.

FIG. 3 is an explanatory diagram showing a configuration of user dictionary data stored in an external storage device.

FIG. 4 is a flowchart illustrating a translation editing process.

FIG. 5 is a flowchart illustrating a flow of a translation process.

FIG. 6 is a flowchart illustrating a flow of a user dictionary editing process.

FIG. 7 is an explanatory diagram showing a new user dictionary registration screen.

FIG. 8 is an explanatory diagram showing a user dictionary copy registration screen.

FIG. 9 is an explanatory diagram showing a new registration screen for a noun in a user dictionary.

FIG. 10 is an explanatory diagram showing a copy registration screen of a noun in a user dictionary.

FIG. 11 is an explanatory diagram showing a copy registration screen of a noun in a user dictionary.

FIG. 12 is an explanatory diagram showing a first page of a registration screen of a verb in a user dictionary.

FIG. 13 is an explanatory diagram showing a second page of a verb registration screen of the user dictionary.

FIG. 14 is an explanatory diagram showing a second page of a registration screen of a verb in a user dictionary.

FIG. 15 is an explanatory diagram showing an adjective registration screen of a user dictionary.

FIG. 16 is an explanatory diagram showing a registration screen of an adverb of a user dictionary.

FIG. 17 is an explanatory diagram showing a registration screen of a head connective in a user dictionary.

FIG. 18 is an explanatory diagram showing a screen for registering a classifier in a user dictionary.

FIG. 19 is an explanatory diagram showing a preposition registration screen of a user dictionary.

FIG. 20 is a block diagram illustrating an overall configuration of a second embodiment.

[Explanation of symbols]

 Reference Signs List 10 CPU 21 Input device 22 Display device 24 External storage device 40 ROM 45 Copy registration (dictionary copy) program 50 RAM 80 User dictionary screen 81 Menu registration button 82 Selection screen 83 Noun dictionary registration screen 84 Noun dictionary registration screen 85 Verb Dictionary registration screen 88 Adjective dictionary registration screen 89 Adverb dictionary registration screen 90 Prepositional conjunction dictionary registration screen 91 Classifier dictionary registration screen 92 Preposition dictionary registration screen 210 CPU 221 Input device 222 Display device 224 Hard disk 240 CD-ROM 245 Copy registration (dictionary copy) program 250 RAM

Claims (16)

[Claims] 1. A machine translator for translating an original sentence in a first language into a translated sentence in a second language with reference to a dictionary, comprising: a headword in the first language; a translated word in the second language corresponding to the headword; A user dictionary in which auxiliary information used for translation processing of the headword is registered as one headword group; a dictionary registration unit that stores a new headword group in the user dictionary; When registering a word group in the user dictionary, selecting means for selecting one headword group from the headword groups already registered in the user dictionary, and displaying a new headword group registration screen and registering the same Registration screen display control means for displaying, on a screen, a headword entry field and a translation word entry field so as to be inputtable, the dictionary registration means , A new entry word input entry word entered in the column, and the newly entered in the translation entry field translation,
A machine translation apparatus for registering, in a user dictionary, information associated with a headword group selected by said selection means. 2. The user dictionary stores the part of speech information of the headword, and the registration screen display control means is based on the part of speech information of the headword group selected by the selection means, based on the part of speech information. 2. The machine translation apparatus according to claim 1, wherein a registration screen of a headword group is displayed. 3. The machine translation apparatus according to claim 1, wherein the registration screen display control means displays a headword entry field and its translation term entry field as blank fields so that they can be entered. 4. The registration screen display control means displays the attached information of the headword group selected by the selection means in an input field of the attached information in a state where it is input. 3. The machine translation device according to any one of 3. 5. The content of an input field for ancillary information on the registration screen can be changed.
A machine translation device according to any one of the above. 6. The machine translation apparatus according to claim 1, wherein said registration screen display control means displays a plurality of related registration screens when one headword group is registered. . 7. A machine translation method for translating an original sentence in a first language into a translated sentence in a second language with reference to a dictionary, comprising: a headword in the first language; a translated word in the second language corresponding to the headword; A registration step of registering the auxiliary information used for the translation processing of the headword in the user dictionary as one headword group, and a new headword group registered in the user dictionary when the new headword group is registered in the user dictionary. A selection process for accepting the selection of one headword group from the headword group, and displaying a registration screen for a new headword group, and displaying a headword input field and a translation word input field on the registration screen so as to allow input. A step, a headword newly input in the headword input field, a translation word newly input in the translation word input field, and the selected word are selected by the selection means. Machine translation method, characterized in that it is composed of a duplicate registration step of registering a supplementary information of soup language groups to a user dictionary. 8. The registration step registers the part-of-speech information of the headword in a user dictionary, and the display step determines the part-of-speech information based on the part-of-speech information of the headword group selected by the selection step. The machine translation method according to claim 7, wherein a registration screen of a headword group based on the display screen is displayed. 9. The machine translation apparatus according to claim 7, wherein in the display step, a headword entry field and its translated word entry field are displayed as blank so that they can be entered. 10. The display step according to claim 7, wherein the display step displays the attached information of the headword group selected by the selecting means in an attached information input field. A machine translation device according to any one of the above. 11. The machine translation method according to claim 7, further comprising a step of accepting a change in the content of an input field for additional information displayed on the registration screen. 12. A computer-readable storage medium storing a machine translation program for translating an original sentence in a first language by a computer into a translation in a second language with reference to a dictionary, comprising: a headword in the first language; A registration step of registering the translation in the second language corresponding to the headword and the additional information used for the translation processing of the headword in the user dictionary as one headword group, and a new headword group in the user dictionary. When registering, a selection step of accepting selection of one headword group from headword groups already registered in the user dictionary, displaying a new headword group registration screen, and entering a headword on the registration screen Field and its translation word entry field are displayed as blank so that they can be entered, and a new entry is made in the entry field Machine translation that causes a computer to execute a registered entry word, a translation word newly entered in a translation word input field, and a supplementary word of the headword group selected by the selection means in a user dictionary. A computer-readable storage medium storing a program. 13. The part-of-speech information of the headword is registered in a user dictionary in the registration step, and the part-of-speech information is registered based on the part-of-speech information of the headword group selected by the selection step. 13. A computer-readable storage medium storing a machine translation program according to claim 12, wherein a registration screen of a headword group based on the program is displayed. 14. The computer-readable program storing a machine translation program according to claim 12, wherein the display step displays a headword entry field and its translation term entry field as blank fields so that they can be entered. Storage medium. 15. The display step according to claim 12, wherein the attached information of the headword group selected by the selection means is input in an attached information input field. A computer-readable storage medium storing the machine translation program according to any one of the above. 16. A computer storing a machine translation program according to claim 12, further comprising a step of receiving a change in the content of an input field for additional information displayed on said registration screen. A readable storage medium.