JP4135316B2 - Machine translation device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a machine translation device that translates an input first language into a second language with reference to a dictionary and enables voice input. In place Related.
[0002]
[Problems to be solved by the invention]
In recent years, various dictionary search software and translation software have been developed along with the improvement of the computer processing capability, and much useful software has been provided in terms of helping human translation. As such machine translation apparatuses, those using a personal computer system, or portable dedicated machines in which a display unit and a key operation unit are compactly provided in a portable main body have been provided.
[0003]
By the way, for example, when an interpreter who interprets a spoken word finds an unknown technical term, it may be possible to use the machine translation device as described above so as to assist the interpretation. However, the above-described conventional machine translation device translates text data input by a key operation or the like, and thus is not suitable for use when immediacy such as interpretation is required. It was.
[0004]
On the other hand, for example, in a personal computer system, the voice recognition function has been improved, and a part in which commands are inputted by voice has already been realized. Therefore, a configuration in which the above-described machine translation device and a speech recognition function are combined is conceivable.
[0005]
However, with a simple combination of a machine translation device and a speech recognition function, even if the user tries to input a command by voice while the translation software is running, the command word is translated into the second language. In order to avoid this problem, for example, it may be possible to turn on the translation key when inputting the first language to be translated by voice, but the user is forced to perform troublesome operations. At the same time, the speed of the translation process is inferior, and the convenience for the user is insufficient.
[0006]
The present invention has been made in view of the above circumstances, and an object of the present invention is to translate an input first language into a second language, which enables voice input and improves convenience. Dress Place In offer.
[0007]
[Means for Solving the Problems]
In order to achieve the above object, the machine translation apparatus according to claim 1 of the present invention. Is When the first language word or the control command for controlling the driving of the device is inputted by voice by the voice input means, any command word can be selected based on the judgment as to whether or not the input data is a specific command word. If the input data is identified as a control command, it is identified as data indicating a first language phrase to be translated. If processing corresponding to the control command is executed and the input data is identified as a word in the first language to be translated, whether the word in the first language is further translated into the second language and output. Alternatively, the output setting means sets whether to output the dictionary contents of the words in the first language, and the output setting means sets the words in the first language to be output after being translated into the second language. When the phrase of the first language is translated into the second language with reference to the dictionary and the result is output, and the output setting means is set to output the dictionary contents of the phrase of the first language. The dictionary contents corresponding to the words in the first language with reference to the dictionary are output in the input order by the voice input means, and if the words in the first language to be translated are unknown words not in the dictionary, The unknown word data is stored and saved in the unknown word data storage means as a dictionary registration candidate.
[0008]
According to this, it is possible to input the first language word to be translated and the control command by voice. At this time, the input data input by voice indicates the first language word to be translated. Whether it is data or data indicating a control command is automatically identified, and processing corresponding to the control command is executed, or translation or dictionary lookup is performed. In this case, since the phrase or control command of the first language to be translated is automatically identified, rapid processing is possible without forcing the user to perform troublesome operations for the distinction.
[0009]
Then, when the input data is data indicating a phrase of the first language to be translated, it is possible to set by the output setting means whether to output it after being translated into the second language or to output the dictionary contents. The user can select either desired output, which is easy to use and highly convenient. Further, there may be a case where the phrase of the first language to be translated is an unknown word that does not exist in the dictionary entry. In such a case, the unknown word data is stored in the unknown word data storage means as a dictionary registration candidate. Thus, the unknown word can be easily registered in the user dictionary later.
[0010]
Machine translation apparatus according to claim 2 of the present invention Is First, it is determined by the identification mode setting means whether or not the input data inputted by voice is a specific command word, and if it is any command word, it is identified as a control command, otherwise it is the first language to be translated A mode serving as an identification criterion is set from among a mode including a command priority mode for identifying a word and a translation priority mode for identifying all of the input data as a first language word to be translated, When a first language word or a control command for driving control of the apparatus is inputted by voice by means of voice input means, the first language to be translated based on the identification criteria set by the identification mode setting means If the input data is identified as a control command, the control command is identified. When corresponding processing is executed and the input data is identified as a word in the first language to be translated, the word in the first language is translated into the second language with reference to the dictionary, and the result And when the first language word to be translated is an unknown word not in the dictionary, the unknown word data is stored and saved in the unknown word data storage means as a dictionary registration candidate. Has characteristics.
[0011]
According to this, it is possible to input the first language word to be translated and the control command by voice. At this time, the input data input by voice is the first language word to be translated. Or data indicating a control command is automatically identified, and processing corresponding to the control command is executed or translated. In this case, since the phrase or control command of the first language to be translated is automatically identified, rapid processing is possible without forcing the user to perform troublesome operations for the distinction.
[0012]
And since the identification reference | standard at the time of identifying the input data can be set by the identification mode setting means, it becomes possible for a user to set a desired identification reference | standard according to a use etc., for example. Further, there may be a case where the phrase of the first language to be translated is an unknown word that does not exist in the dictionary entry. In such a case, the unknown word data is stored in the unknown word data storage means as a dictionary registration candidate. Thus, the unknown word can be easily registered in the user dictionary later.
[0013]
The machine translation apparatus according to claim 3 of the present invention Is First, it is determined by the identification mode setting means whether or not the input data inputted by voice is a specific command word, and if it is any command word, it is identified as a control command, otherwise it is the first language A mode serving as an identification criterion is set from among a mode including a command priority mode for identifying a word and a translation priority mode for identifying all of the input data as a first language word to be translated, When a first language word or a control command for driving control of the apparatus is inputted by voice by means of voice input means, the first language to be translated based on the identification criteria set by the identification mode setting means If the input data is identified as a control command, the control command is identified. When the corresponding process is executed and the input data is identified as a word in the first language to be translated, the word in the first language is further translated into the second language and output. If the output setting means sets whether to output the dictionary contents of the words in one language, and the output setting means sets the words in the first language to be translated into the second language, the first language Is translated into the second language with reference to the dictionary, and the result is output. If the output setting means is set to output the dictionary contents of the first language, the dictionary is referred to If the contents of the dictionary corresponding to the words in the first language are output in the order of input by the voice input means, and the words in the first language to be translated are unknown words that are not in the dictionary, the unknown word data is the dictionary registration candidate. As unknown word data storage Characterized in was configured to be stored stored in the stage.
[0014]
According to this, it is possible to input the first language word to be translated and the control command by voice. At this time, the input data input by voice is the first language word to be translated. Or data indicating a control command is automatically identified, and processing corresponding to the control command is executed or translated. In this case, since the phrase or control command of the first language to be translated is automatically identified, rapid processing is possible without forcing the user to perform troublesome operations for the distinction.
[0015]
And since the identification reference | standard at the time of identifying the input data can be set by the identification mode setting means, it becomes possible for a user to set a desired identification reference | standard according to a use etc., for example. At the same time, when the input data is data indicating a phrase of the first language to be translated, it can be set by the output setting means whether it is translated into the second language and outputted, or the dictionary contents are outputted. Therefore, the user can select either desired output, which is easy to use and highly convenient. Further, there may be a case where the phrase of the first language to be translated is an unknown word that does not exist in the dictionary entry. In such a case, the unknown word data is stored in the unknown word data storage means as a dictionary registration candidate. Thus, the unknown word can be easily registered in the user dictionary later.
[0016]
Further, when setting the identification standard by the identification mode setting means, the specific language or the phrase following the predetermined phrase is identified as a control command in the first language input by the voice input means, and there is no specific phrase or the predetermined phrase. A format designation mode for identifying the first language to be translated can be set. Four invention). According to this, by setting the format designation mode, it is possible to more reliably identify whether or not it is a control command when identifying input data.
[0017]
Then, when there is an unknown word, the unknown word may be stored in the unknown word data storage means by pairing the speech data and the translated word data (claims) Of 5 invention). According to this, unlike the case where the reading data and the translated word data are stored in pairs, it is possible to cope with the user's utterance, pronunciation, etc., and to store the unknown word data in a form that is easier to use. Can do.
[0018]
Furthermore, the translation result and the dictionary content may be displayed on the display means, and at that time, the output result of the translation result and the output content of the dictionary content may be displayed in different forms. 6's (Invention) Thus, the translation result and the contents of the dictionary can be output (displayed) in an easily understandable manner to the user in a state where they are distinguished.
[0019]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, a first embodiment of the present invention will be described with reference to FIGS. In the first embodiment, the present invention is applied to a case where the present invention is realized by a general-purpose personal computer (abbreviated as personal computer) system. Further, here, an example is given in which translation is performed from Japanese as the first language to English as the second language.
[0020]
FIG. 2 shows an external configuration of a personal computer system 1 which is a machine translation apparatus according to the present embodiment, and FIG. 3 schematically shows an electrical configuration of a main part thereof by blocks. Here, as shown in FIG. 2, this personal computer system 1 comprises, for example, a liquid crystal display in a personal computer main body 2, and an original Japanese text, a translated English translation, and a control command inputted as described later. A display unit 3 which is a display means for displaying icons, etc., a keyboard 4 and a mouse 5 which are input operation devices for performing various settings such as various settings, input of original texts and instructions of control commands, voice input means As a microphone 6 (hereinafter simply referred to as a microphone 6).
[0021]
In the personal computer main body 2, as shown in FIG. 3, a control unit 7 that performs overall control, a hard disk device 8 that is an external storage device, a display control unit 9 that controls display of the display unit 3, and the microphone 6. A voice recognition unit 10 that recognizes a voice signal input from the CD-ROM drive 11 and a CD-ROM drive 11 are provided. The control unit 7 is mainly composed of a microcomputer, and is configured by connecting a CPU 12, a ROM 13, a RAM 14 as a memory, and an input / output interface 15 to each other via a bus 16 as is well known. The input / output interface 15 is connected to the display control unit 9, the keyboard 4, the mouse 5, the voice recognition unit 10, the CD-ROM drive 11, and the hard disk device 8.
[0022]
The CD-ROM drive 11 is set with a CD-ROM 17 as a recording medium on which a machine translation processing program to be described later is recorded, and the program is read. The hard disk device 8 stores a dictionary (dictionary data) for translation processing. In this case, the dictionary includes a basic dictionary 18, a specialized dictionary 19 edited for each specialized field, and a user dictionary 20 that is independently constructed by the user.
[0023]
FIGS. 4A, 4B, and 4C show examples of data structures of these dictionary data, that is, the basic dictionary 18, the specialized dictionary 19, and the user dictionary 20, respectively. These dictionary data are composed of, for example, dictionary information including headword information (reading data) in Japanese, part-of-speech information, translated word information in English, and syntax / semantic information such as field, utilization form, and usage. In the translation process, the user dictionary 20 is given the highest priority for the same headword, and then the dictionary is used (dictionary lookup) in the order of the specialized dictionary 19 and the basic dictionary 18. . The use of the user dictionary 20 and the specialized dictionary 19 depends on user settings.
[0024]
The control unit 7 executes various processes related to translation by executing a machine translation processing program read from the CD-ROM 17. Specifically, as will be described later in the description of the operation, the translation process for translating the input Japanese original sentence (input sentence) into English and outputting it (displayed on the display unit 3), the input Japanese phrase Dictionary content output (dictionary lookup) processing for outputting (displayed on the display unit 3) the dictionary content (for example, a plurality of English translations), input sentence editing processing for editing the input original sentence, and translation processing Processes such as a translation edit process for editing an English translation sentence and a dictionary edit process for constructing a user dictionary are executed.
[0025]
At this time, in this machine translation process, a text input mode in which Japanese original text is input by text data (character data), Japanese words to be translated (or dictionary looked), and drive control to be executed Control commands (for example, processes such as “file”, “help”, “close”, “edit”, “search”) can be executed in a voice input mode in which voice commands from the microphone 6 are received. .
[0026]
As will be described later in the description of the operation, in the voice input mode, the control unit 7 inputs Japanese speech from the microphone 6 and is recognized by the voice recognition unit 10 in Japanese words to be translated. If the input data is identified as a control command, the process is executed. If the input data is identified as a Japanese word to be translated, a translation process is performed. It is translated into English (or a dictionary) and output (displayed on the display unit 3). Therefore, the control unit 7 functions as an identification unit, a control command execution unit, a translation unit, and a dictionary content output unit.
[0027]
At this time, in this embodiment, any one of a command priority mode, a translation priority mode, and a format designation mode is used as an identification criterion when identifying whether the input data is a Japanese word to be translated or a control command. These modes can be set (selected) in advance by operating the user's keyboard 4 (or mouse 5). Note that the command priority mode is initialized without performing any setting operation. Therefore, the keyboard 4 and the mouse 5 function as identification mode setting means.
[0028]
Among them, the command priority mode first determines whether the input data is a first language (specific word / phrase) in a specific format, that is, a command word such as “file”, “help”, “close”, Any command word is identified as a control command, otherwise it is identified as a first language phrase to be translated. In the translation priority mode, all input data is identified as a phrase of the first language to be translated. Therefore, in order to execute a control command in this mode, a command input by key operation is required. . In the format designation mode, a phrase that is input after a predetermined word in a specific format is identified as a control command. For example, a command word is input after “## (sharp sharp)” that is a predetermined word. For example, it is identified as a control command, and if there is no predetermined phrase, it is identified as a phrase in the first language to be translated.
[0029]
In this embodiment, when the input data is identified as a Japanese word to be translated, the translation result is basically output (displayed), but the user can use the keyboard 4 (or mouse 5). By instructing the output of the dictionary contents by operating, the dictionary contents of the input phrase of the first language are output. Accordingly, the keyboard 4 and mouse 5 also function as output setting means.
[0030]
In this case, although not shown, the display unit 3 is provided with, for example, an area (window) that is located on the left side and displays input data, and an area (window) that displays a translation (translated word) on the right side thereof. However, the display of the dictionary content is performed by providing a window that is different from the above, so that the translated content and the dictionary content are displayed in different forms. It has become. When outputting the contents of the dictionary, the display is performed in the order of input data.
[0031]
Further, when a Japanese word inputted by voice is an unknown word that does not exist in the dictionary entry, the unknown word data is stored and saved in the RAM 14 as a dictionary registration candidate to be registered in the user dictionary 20. It is like that. At this time, the RAM 14 is provided with a data storage file and functions as an unknown word data storage means.
[0032]
Various known methods can be used as a method of speech recognition by the speech recognition unit 10 described above. For example, a feature parameter obtained by analyzing a voice signal input from the microphone 6 every predetermined period (4 to 20 ms) and extracting a spectrum included in the voice by a bandpass filter bank is stored in a standard for voice recognition dictionary. A method of outputting a standard pattern having a minimum distance as a recognition result as compared with a pattern, a method of using a sound hidden Markov model, or the like can be employed.
[0033]
Next, the operation of the above configuration will be described with reference to FIGS. First, the flowchart of FIG. 5 shows an outline of the translation / editing processing procedure in the text input mode during the execution of the machine translation processing program by the control unit 7. That is, in step S1, the original text in Japanese is input. This input is performed by the user operating the keyboard 4 to input a Japanese original text, or may be performed by reading an already created original text file. When the original Japanese text is input, the input text is displayed on the display unit 3 in step S2.
[0034]
Here, when the user wants to edit the input sentence, the user instructs to edit the input sentence by the key input operation of the keyboard 4 or the operation of the mouse 5 (Yes in step S3). Then, in step S4, an input sentence editing process is executed. In this input sentence editing process, the input sentence displayed on the display unit 3 can be edited such as deletion, addition, correction, etc., based on the operation of the user's keyboard 4 and mouse 5.
[0035]
When the input (or edited) original sentence is to be translated, the user instructs the input sentence editing by the key input operation of the keyboard 4 or the mouse 5 (Yes in step S5). Then, a translation process for translating the input sentence into an English translation sentence is performed (step S6). This translation processing will be described later (FIG. 6). When the translation process is performed, the input sentence and the translated sentence are stored in the RAM 14 in a state of being associated with each other in step S7, and the input sentence and the translated sentence are displayed on the display unit 3 in step S8. .
[0036]
Further, when the user wants to edit the output translation, the user instructs the translation editing by the key input operation of the keyboard 4 or the operation of the mouse 5 (Yes in step S9). Then, in step S10, editing / editing processing such as deletion, addition, correction, etc. is performed on the translation displayed on the display unit 3. If the user wants to end the translation process, the user ends the process (Yes in step S11), and the process ends. For example, if an instruction for other processing such as editing of the user dictionary is given (No in step S11), the corresponding processing is executed (step S12). Although not shown, as a matter of course, as another process, an input sentence and a translated sentence can be stored in a recording medium such as a floppy disk or printed.
[0037]
Here, FIG. 6 shows the procedure of the translation process described above (details of the process in step S6 in FIG. 5). That is, first, in step S21, a morphological analysis process is performed in which a dictionary is searched for words constituting an input sentence by using predetermined morphological analysis rules, the words are extracted, and search information such as a translation and a part of speech is added to the morphological analysis result. Is executed. In step S22, based on the result of the morphological analysis process, an analysis process for analyzing the Japanese structure is executed using a predetermined analysis grammar.
[0038]
In the next step S23, using a predetermined conversion grammar, a Japanese structure analyzed by the analysis process is converted into an English structure, and a conversion process for replacing individual words with English is executed. Subsequently, in step S24, using a predetermined generation grammar, the English word order is determined from the English structure converted by the conversion process, an article peculiar to English is generated, the subject required for the English sentence The generation process of generating an output sentence composed of an English word string is executed by processing tense, aspect, personality, and number feature.
[0039]
Then, in step S25, using a predetermined morpheme generation rule, an inflected form such as plural nouns, past verbs, and past participle is generated for an English word string generated by the generation process. The morpheme generation for obtaining the translated sentence is executed, and the translation process is completed.
[0040]
Now, the flowchart of FIG. 1 shows a processing procedure in the voice input mode during the execution of the machine translation processing program by the control unit 7. Here, although not shown, the identification standard is set in advance (identification setting step), and the command priority mode is set (the setting operation by the user is not performed and the initial setting is maintained. ).
[0041]
In this voice input mode, as described above, Japanese words to be translated (words and phrases) and control commands such as “file”, “help”, “close”, “edit”, “search”, It is adapted to be received by voice input from the microphone 6 (voice input process), and in step S31, it waits for voice input by the user. When there is a voice input (Yes in step S31), the input voice is recognized by the voice recognition unit 10 and converted into reading data (step S32).
[0042]
In the next step S33, it is determined whether or not the input data (reading data) is a control command, that is, whether it is in a command word (identification step). If the input data is identified as a control command (Yes in step S33), processing corresponding to the control command is executed in step S34 (control command execution step).
[0043]
On the other hand, if it is determined that the input data is not a control command (No in step S33), it is determined that the input data is a Japanese phrase to be translated, and the read data is determined in the next step S35. Is matched with a dictionary entry. If the reading data is present in the dictionary entry (Yes in step S36), it is determined in step S37 whether to translate and output or to output the dictionary contents (dictionary lookup).
[0044]
Here, as described above, if the user does not perform any operation, the translation is performed as it is, and the dictionary is retrieved by performing the key operation by the user (output setting process). In step S38, the input data is translated, and the result (English translation (translation)) is displayed on the display unit 3 (translation process). If dictionary content output (dictionary lookup) is set, the dictionary contents are displayed on the display unit 3 in the order of input in step S39 (dictionary content output step).
[0045]
On the other hand, when the input data (reading data) input by speech is an unknown word that does not exist in the dictionary data entry (No in step S36), the unknown word is obtained in step S40. The reading data is written and stored in the data storage file of the RAM 14 as a dictionary registration candidate (unknown word data storage step). In the present embodiment, even if the input data is present in the dictionary data entry word and translated, the input data (and the translated word data) is stored as a dictionary registration candidate in the RAM 14 by, for example, a user operation. Can be stored and saved in the data saving file (step S41).
[0046]
Thus, in the voice input mode, the user can input a Japanese word or control command to be translated into English from the microphone 6 by voice. Then, it is automatically identified whether it is a Japanese word or phrase to be translated or a control command, and translation (dictionary lookup) or command execution is performed.
[0047]
For this reason, a special operation for distinguishing between a word to be translated and a control command is not necessary, and the user is not forced to perform a troublesome operation, and the speed of processing can be improved. Therefore, Japanese words and phrases can be translated easily and quickly. For example, an interpreter who interprets spoken language can be used for assisting interpretation when an unknown technical term appears. It becomes.
[0048]
When the input data is data indicating a phrase of the first language to be translated, the user can set whether to output the dictionary or the contents of the dictionary after translation into the second language. Can select either desired output, making it easy to use and highly convenient. At the same time, an identification criterion for identifying whether the command is a control command can also be set (selected) from among the three modes by the user, so that a desired identification criterion can be set.
[0049]
In this case, although a detailed description is omitted, a format designation mode for identifying a control command is set by inputting a command word after inputting a predetermined word (for example, “## (sharp sharp)”). This makes it possible to more reliably identify whether the control command is a word in the first language to be translated. If it is determined that the user does not perform voice input of the control command, if the translation priority mode is set, translation processing can be performed immediately on all input data, and processing time from voice input to output (Responsiveness) can be improved.
[0050]
Moreover, when the Japanese word to be translated is an unknown word that does not exist in the dictionary data entry, the input data can be stored and saved in the data storage file of the RAM 14 as a dictionary registration candidate. The user can easily perform operations such as registering the unknown word in the user dictionary 20 by calling the data storage file in the RAM 14.
[0051]
As described above, according to the present embodiment, input Japanese is translated into English, voice input is possible, and input data input by voice is a Japanese phrase to be translated. Since it is configured to automatically identify whether the command is a control command, it is possible to easily and quickly translate Japanese words and phrases, and to improve the convenience for the user.
[0052]
In addition, in this embodiment, in particular, when the input data is Japanese to be translated, the user can set whether to translate or look up a dictionary, and at the same time, for identifying whether it is a control command or not Since the user can also set the identification criteria, it is easy to use and highly convenient. Furthermore, if the Japanese word to be translated is an unknown word that does not exist in the dictionary entry, the unknown word data is stored in the dictionary. Since it is configured to store and save as registration candidates, it is possible to obtain advantages such as easy registration of the user dictionary 20.
[0053]
FIG. 7 shows an external configuration of the machine translation device 21 according to the second embodiment of the present invention. This machine translation device 21 is different from the one using the personal computer system as in the first embodiment, and is a portable dedicated machine that translates the input first language (Japanese) into the second language (English). It is configured as. This machine translation device 21 is an LCD for displaying an input Japanese original sentence (input sentence) and an English translation sentence on a surface portion of a main body case 22 having a thin rectangular shape having a portable size. And a key operation unit 24 having a number of keys for a user to perform necessary operations.
[0054]
The key operation unit 24 includes a large number of character input keys 24a for inputting original text, a return key 24b, a menu key 24c for setting functions, a cancel key 24d, a command priority mode and a translation priority mode. A selector switch 24e as an identification mode setting means for switching, a dictionary lookup key 24f as an output setting means for instructing dictionary lookup, and the like are provided. The main body case 22 is connected with a microphone 25 as voice input means, and for example, a charging terminal 26 connected to a commercial power source, data for exchanging data with the outside (such as a personal computer). A transmission connector 27 and the like are also provided. If this connector 27 is used, the data of the specialized dictionary 19 can be taken in from a personal computer or the like in advance.
[0055]
Although not shown, the main body case 22 includes a control unit configured mainly by a microcomputer to control the entire apparatus, various processing circuits such as voice recognition processing and data input / output processing, and a secondary power source. A battery or the like is provided. At this time, a machine translation processing program, dictionary data for translation processing, and the like are stored in the memory (ROM) of the control unit (microcomputer). The RAM of the control unit is provided with a data storage file for storing and storing unknown word data and the like.
[0056]
In this case, by operating the menu key 24c, as in the first embodiment, a text input mode in which Japanese original text is input by text data (character data), and Japanese words and phrases to be translated are executed. It is possible to execute a voice input mode in which a control command for drive control is received by voice input from the microphone 25.
[0057]
Furthermore, in the voice input mode, the command priority mode and the translation priority mode can be switched by operating the changeover switch 24e. Among them, in the translation priority mode, the input data that is input from the microphone 25 and recognized by the speech is translated as a Japanese word to be translated (the original word and the translated word are displayed on the display unit 23). It has become. Therefore, in order to execute a command in this mode, the command input by the key operation unit 24 is required.
[0058]
In the command priority mode, as in the first embodiment, the control unit controls whether the input data that has been voice-input from the microphone 25 and voice-recognized is a Japanese word to be translated. If the input data is identified as a control command, the process is executed. If the input data is identified as a Japanese word to be translated, it is translated into English. Are output (displayed on the display unit 23).
[0059]
At this time, translation is normally performed and the result is output, but when the dictionary lookup key 24f is turned on by the user, the contents of the dictionary are output (displayed). Further, when a Japanese word inputted by speech is an unknown word that does not exist in the dictionary data entry, the unknown word data is stored in the RAM data storage file as a dictionary registration candidate to be registered in the user dictionary. Memorize and save. Further, this unknown word information or the like can be taken into the personal computer through the connector 27.
[0060]
Therefore, also in the second embodiment, as in the first embodiment, the input Japanese is translated into English, enabling voice input and input data input by voice. Has a configuration that automatically identifies whether a word is a Japanese word to be translated or a control command, so Japanese words can be translated easily and quickly, improving the convenience for the user There are excellent effects such as being able to. Moreover, in the present embodiment, since the apparatus is configured as a portable type, it is possible to obtain an advantage of excellent portability when interpreting.
[0061]
In each of the above embodiments, the unknown word is stored in the data storage file in the RAM, and the reading data is stored. However, the unknown word is stored as a pair of the voice data and the translated word data. In this way, it is possible to cope with a habit of a user's utterance, pronunciation, etc., and to make data easier to use. In each of the above-described embodiments, whether to output the translation result or the dictionary contents is set by key operation. However, output setting may be performed in advance by mode setting.
[0062]
In addition, the recording medium on which the machine translation processing program is recorded is not limited to the CD-ROM 17, and various media such as a magnetic disk, a semiconductor memory, a CD-R, an MD, an MO, a PD, a DVD, and a magnetic tape can be used. Moreover, not only the translation from Japanese to English, but also the translation from English to Japanese, the first language and the second language can be applied to the languages of all over the world. The present invention can be implemented with appropriate modifications within a range not departing from the gist of the invention.
[0063]
【The invention's effect】
As is apparent from the above description, according to the present invention, the input data input by the voice input means is identified whether it is a phrase of the first language to be translated or a control command, and the input data Is identified as a control command, the process is executed. If the input data is identified as a first language phrase to be translated, the input data is translated into a second language and output. When the word of the language is an unknown word that does not exist in the dictionary data entry, the unknown word data is stored and stored in the memory as a dictionary registration candidate. There is an excellent effect of improving the convenience.
[Brief description of the drawings]
FIG. 1 shows a first embodiment of the present invention, and is a flowchart showing a processing procedure at the time of voice input
FIG. 2 is an external view of a machine translation device
FIG. 3 is a block diagram schematically showing an electrical configuration of the machine translation apparatus.
FIG. 4 is a diagram schematically showing the structure of dictionary data.
FIG. 5 is a flowchart showing a translation / editing procedure in the text input mode.
FIG. 6 is a flowchart showing the procedure of translation processing.
FIG. 7 is a perspective view showing an appearance of a machine translation apparatus according to the second embodiment of the present invention.
[Explanation of symbols]
In the drawings, 1 is a personal computer system (machine translation device), 2 is a personal computer body, 3 and 23 are display units (display means), 4 is a keyboard (identification mode setting means and output setting means), and 6 and 25 are microphones ( (Speech input means), 7 is a control section (identification means, control command execution means, translation means, dictionary content output means), 10 is a speech recognition section, 11 is a CD-ROM drive, 12 is a CPU, 13 is a ROM, RAM (unknown word data storage means), 17 is a CD-ROM (recording medium), 20 is a user dictionary, 21 is a machine translation device, 22 is a main body case, 24 is a key operation unit, and 24e is a changeover switch (identification mode setting means). ) And 24f indicate dictionary lookup keys (output setting means).

Claims (6)

A machine translation device that translates an input first language into a second language with reference to a dictionary,
Voice input means for voice input of words or phrases of the first language or a control command for drive control of the device;
It is determined whether or not the input data by the voice input means is a specific command word, and if it is any command word, it is identified as data indicating a control command. An identification means for identifying data indicating a language phrase;
When the identification means identifies that the input data is a control command, a control command execution means for executing processing corresponding to the control command;
When the identification means identifies that the input data is a phrase of the first language to be translated, the phrase of the first language is translated into the second language and output, or the phrase of the first language Output setting means for setting whether to output the dictionary contents of
When it is set by this output setting means that the first language word is translated into the second language and output, the first language word is translated into the second language with reference to the dictionary, and the result is output. Translation means,
When it is set by the output setting means to output the dictionary contents of the words in the first language, the dictionary contents corresponding to the words in the first language are output in the input order by the voice input means with reference to the dictionary. Dictionary content output means;
A machine translation device comprising: unknown word data storage means for storing and storing unknown word data as a dictionary registration candidate when a phrase of the first language to be translated is an unknown word not in the dictionary. A machine translation device that translates an input first language into a second language with reference to a dictionary,
Voice input means for voice input of words or phrases of the first language or a control command for drive control of the device;
Identification means for identifying whether the input data by the voice input means is data indicating a phrase of the first language to be translated or data indicating a control command;
The mode used as an identification standard when the identification means performs identification is first determined whether the input data is a specific command word, and if it is any command word, it is identified as a control command. An identification set from a command priority mode for identifying a phrase in the first language to be translated, and a mode including a translation priority mode for identifying all the input data as a phrase in the first language to be translated Mode setting means;
When the input means identifies the input data as a control command, the control means executing means for executing processing corresponding to the control command;
If the identification means identifies that the input data is a phrase in the first language to be translated, the phrase in the first language is translated into a second language with reference to the dictionary, and the result is Translation means to output;
A machine translation device comprising: unknown word data storage means for storing and storing unknown word data as a dictionary registration candidate when a phrase of the first language to be translated is an unknown word not in the dictionary. A machine translation device that translates an input first language into a second language with reference to a dictionary,
Voice input means for voice input of words or phrases of the first language or a control command for drive control of the device;
Identification means for identifying whether the input data by the voice input means is data indicating a phrase of the first language to be translated or data indicating a control command;
The mode used as an identification standard when the identification means performs identification is first determined whether the input data is a specific command word, and if it is any command word, it is identified as a control command. An identification set from a command priority mode for identifying a phrase in the first language to be translated, and a mode including a translation priority mode for identifying all the input data as a phrase in the first language to be translated Mode setting means;
When the input means identifies the input data as a control command, the control means executing means for executing processing corresponding to the control command;
When the identification means identifies that the input data is a phrase of the first language to be translated, the phrase of the first language is translated into the second language and output, or the phrase of the first language Output setting means for setting whether to output the dictionary contents of
When it is set by this output setting means that the first language word is translated into the second language and output, the first language word is translated into the second language with reference to the dictionary, and the result is output. Translation means,
When it is set by the output setting means to output the dictionary contents of the words in the first language, the dictionary contents corresponding to the words in the first language are output in the input order by the voice input means with reference to the dictionary. Dictionary content output means;
A machine translation device comprising: unknown word data storage means for storing and storing unknown word data as a dictionary registration candidate when a phrase of the first language to be translated is an unknown word not in the dictionary.   The identification mode setting unit identifies, as a control command, a specific word or a phrase following the predetermined word from the first language input by the voice input unit, and if there is no specific word or the predetermined word, the first language to be translated The machine translation apparatus according to claim 2 or 3, wherein a format designation mode for identifying as follows is settable.   5. The machine translation apparatus according to claim 1, wherein the unknown word data storage unit stores the unknown word in pairs of the voice data and the translated word data.   A display means for displaying the output contents of the translation means and the dictionary content output means is provided, and the display means displays the output contents of the translation means and the output contents of the dictionary content output means in different forms. Item 6. The machine translation device according to any one of Items 1 to 5.

Images