JP2014238772A - Translation system, translation program, and translation method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a translation system suitable for reducing memory capacity to be used and processing load.SOLUTION: Morpheme translation processing acquires a morpheme from a character string to be translated, as a sub-character string, searches a translation information table 400 which stores a key morpheme and an optimal translation word in association with each other, for a key morpheme perfectly matches the sub-character string, acquires an optimal translation word corresponding to the extracted key morpheme from the translation information table 400 when the key morpheme is extracted, searches the translation information table 400 for a key morpheme including the sub-character string when the key morpheme is not extracted, and acquires an optimal translation word corresponding to a key morpheme having the highest occurrence frequency in the extracted key morphemes when a plurality of key morphemes are extracted, from the translation information table 400, to output the acquired optimal translation word as a translation word for the sub-character string.

Description

  The present invention relates to a system, a program, and a method for translating a character string to be translated in a first language into a second language for each morpheme, and in particular, a translation system and a translation suitable for reducing the used memory capacity and processing load. The present invention relates to a program and a translation method.

  Conventionally, for example, techniques described in Non-Patent Documents 1 and 2 are known as text translation applications.

“Google Translate for Android”, [online], May 8, 2013, Google Inc., [Search June 7, 2013], Internet <URL: https://play.google.com/store/ apps / details? id = com.google.android.apps.translate & hl = en> “ABBYY TextGrabber + Translator”, [online], June 3, 2013, ABBYY Software House, [Search June 7, 2013], Internet <URL: https://play.google.com/store/ apps / details? id = com.abbyy.mobile.textgrabber.full & hl = en>

  However, the technologies described in Non-Patent Documents 1 and 2 require the use of a library (dictionary) for each language in order to support a large number of languages, and thus have a problem of using a large amount of memory area in a mobile device. It was. In addition, since a complex algorithm is used to achieve high translation accuracy, there is a problem that the processing load on the mobile device is heavy. In recent years, although the performance of mobile devices has improved, the memory capacity and processing performance cannot be said to be sufficient yet, so the above problem is particularly remarkable in mobile devices.

  Therefore, the present invention has been made paying attention to such an unsolved problem of the conventional technique, and is suitable for reducing the memory capacity used and the processing load, as well as a translation program, and a translation program. It aims to provide a method.

  [Invention 1] In order to achieve the above object, a translation system according to Invention 1 is a translation system that translates a translation target character string composed of a first language into a second language for each morpheme. Morpheme acquisition means for acquiring a morpheme, and for a plurality of different key morphemes composed of the first language, the key morpheme, a translation obtained when the key morpheme is translated into the second language, and First key morpheme search means for searching for the key morpheme that completely matches the morpheme acquired by the morpheme acquisition means, from the translation information storage means for storing the key morpheme or the appearance frequency of the translated word in association with each other, and the first key A first translated word that obtains the translated word corresponding to the retrieved key morpheme from the translated information storage means when the key morpheme is found by the morpheme searching means; And a second key morpheme search unit that searches the translation information storage unit for the key morpheme including the morpheme acquired by the morpheme acquisition unit when the key morpheme is not searched by the acquisition unit and the first key morpheme search unit And when the plurality of key morphemes are searched for by the second key morpheme search means, the translation corresponding to the key morpheme having the highest appearance frequency among the plurality of searched key morphemes is the translation information. A second translation acquisition unit that acquires from the storage unit; and a translation output unit that outputs the translation acquired by the first translation acquisition unit or the second translation acquisition unit as a translation of the morpheme acquired by the morpheme acquisition unit. .

  If it is such a structure, a morpheme will be acquired from a translation object character string by a morpheme acquisition means, and the key morpheme which completely corresponds to the acquired morpheme will be searched from a translation information storage means by a 1st key morpheme search means. . Here, when the key morpheme is retrieved, the translation corresponding to the retrieved morpheme is acquired from the translation information storage unit by the first translation acquisition unit. Then, the translation output means outputs the acquired translation as a translation of the acquired morpheme.

  On the other hand, if the key morpheme is not found, the second key morpheme search unit searches for the key morpheme including the acquired morpheme from the translation information storage unit. Here, when a plurality of key morphemes are searched, the second translation acquisition unit acquires the translation corresponding to the key morpheme having the highest appearance frequency from the plurality of searched key morphemes from the translation information storage unit. Is done. Then, the translation output means outputs the acquired translation as a translation of the acquired morpheme.

  Here, the translation output means outputs the translation by, for example, display, printing, voice output, writing to a storage device or storage medium, transmission to an external terminal, vibrations such as vibration, heat generation, and other methods. Can do. Therefore, the output includes at least display, printing, audio output, writing (including file generation), transmission, vibration, and heat generation. Hereinafter, the concept of output is the same.

  In addition, the key morpheme, the translation word, and the appearance frequency are stored in association with each other. (1) For example, the key morpheme, the translation word, and the appearance frequency are stored in the same record, for example, (2) ) For example, providing a table for associating key morphemes and intermediate information and a table for associating and registering translation words, appearance frequencies, and intermediate information. It is. That is, any data structure can be adopted as long as the translated word or appearance frequency can be traced from the key morpheme. Hereinafter, the concept of associating and storing information is the same.

  The translation information storage means stores key morphemes, translations and appearance frequencies at any time and at any time, and may store key morphemes, translations and appearance frequencies in advance, The key morpheme, the translated word, and the appearance frequency may be stored by inputting from the outside or the like during the operation of the system without previously storing the key morpheme, the translated word, and the appearance frequency.

  Further, the present system may be realized as a single device, terminal, or other device, or may be realized as a network system in which a plurality of devices, terminals, or other devices are communicably connected. In the latter case, each component may belong to any one of a plurality of devices and the like as long as they are connected so as to communicate with each other. Hereinafter, the same applies to the translation system of the invention 2.

  [Invention 2] Furthermore, the translation system of Invention 2 is a translation system that translates a translation target character string in a first language into a second language for each morpheme, and acquires a morpheme from the translation target character string. And a plurality of different key morphemes composed of the first language, and an optimum translated word that is most recommended as a candidate of a translated word among a plurality of translated words obtained when the key morpheme and the key morpheme are translated into the second language Obtained by the morpheme acquisition unit from the translation information storage unit that stores the next candidate translation other than the optimum translation among the plurality of translations and the appearance frequency of the key morpheme or the optimum translation in a predetermined document. A first key morpheme search means for searching for the key morpheme that completely matches the morpheme, and the key morpheme searched by the first key morpheme search means If the key morpheme is not retrieved by the first translation acquisition unit that acquires the optimal translation corresponding to the searched key morpheme from the translation information storage unit, and the first key morpheme search unit, A second key morpheme search unit that searches the translation information storage unit for the key morpheme including the morpheme acquired by the acquisition unit; and a plurality of the key morpheme searched by the second key morpheme search unit; A second translation acquisition unit that acquires the optimal translation corresponding to the key morpheme having the highest appearance frequency from the translation information storage unit, and the first translation acquisition unit or the second translation A translation output unit that outputs the optimal translation acquired by the acquisition unit as a translation of the morpheme acquired by the morpheme acquisition unit; and the first translation acquisition unit or the second translation acquisition unit The next candidate translation output means for outputting the next candidate translation corresponding to the optimum translation, and the next candidate translation output by the next candidate translation output means for the optimum translation specified among the optimum translations obtained in the step Next candidate translation selecting means for selecting one of the above, and a translation changing means for changing the designated optimum translation to the next candidate translation selected by the next candidate translation selecting means.

  If it is such a structure, a morpheme will be acquired from a translation object character string by a morpheme acquisition means, and the key morpheme which completely corresponds to the acquired morpheme will be searched from a translation information storage means by a 1st key morpheme search means. . Here, when the key morpheme is searched, the optimal translation corresponding to the searched morpheme is acquired from the translation information storage unit by the first translation acquisition unit. And the acquired optimal translation is output as a translation of the acquired morpheme by a translation output means.

  On the other hand, if the key morpheme is not found, the second key morpheme search unit searches for the key morpheme including the acquired morpheme from the translation information storage unit. Here, when a plurality of key morphemes are retrieved, the second translation acquisition unit obtains the optimal translation corresponding to the key morpheme having the highest appearance frequency from the translation information storage unit. To be acquired. And the acquired optimal translation is output as a translation of the acquired morpheme by a translation output means.

  When any one of the acquired optimal translations is designated, the next candidate translation output means outputs a next candidate translation corresponding to the designated optimal translation. Then, when one of the output next candidate translated words is selected by the next candidate translated word, the designated optimum translated word is changed to the selected next candidate translated word.

  Here, the translation information storage means stores the key morpheme, the optimum translation, the next candidate translation and the appearance frequency at any time and at any time, and stores the key morpheme, the optimum translation, the next candidate translation and the appearance frequency in advance. The key morpheme, the optimal translation, the next candidate translation, and the appearance frequency may be stored in advance without the key morpheme, the optimal translation, the next candidate translation, and the appearance frequency being stored. And the appearance frequency may be stored.

  [Invention 3] Further, the translation system of Invention 3 is the translation system of Invention 2, wherein the translation information storage means includes, for a plurality of different key morphemes, the key morpheme, the part of speech of the key morpheme, the optimal translation, Part-of-speech utilization generation means for storing the next candidate translation and the appearance frequency in association with each other and generating all part-of-speech utilizations of the key morpheme based on the part-of-speech corresponding to the key morpheme retrieved by the first key morpheme retrieval unit And, based on the part-of-speech utilization generated by the part-of-speech utilization generation unit, the part-of-speech search for the part-of-speech utilization matching the character string after the morpheme acquired by the morpheme acquisition unit in the translation target character string A utilization retrieval unit; and a part of speech utilization output unit that outputs the part of speech utilization retrieved by the part of speech utilization retrieval unit.

  With this configuration, the part-of-speech utilization generation unit generates all the part-of-speech utilizations of the key morpheme based on the part-of-speech corresponding to the searched key morpheme, and the part-of-speech utilization retrieval unit generates the part-of-speech utilization Among them, the part-of-speech utilization that matches the character string after the acquired morpheme in the character string to be translated is searched from the character string to be translated. The part-of-speech utilization retrieved is output by the part-of-speech utilization output means.

  Here, the translation information storage means stores the part of speech of the key morpheme by any means and at any time. The part of speech of the key morpheme may be stored in advance, or the part of speech of the key morpheme may be stored. The part of speech of the key morpheme may be stored by external input or the like during the operation of the system without being stored in advance.

  [Invention 4] On the other hand, in order to achieve the above object, the translation program of Invention 4 is a translation program for executing a process of translating a character string to be translated in the first language into the second language for each morpheme. A morpheme acquisition step of acquiring a morpheme from the translation target character string, and a translation obtained by translating the key morpheme and the key morpheme into the second language with respect to a plurality of different key morphemes of the first language And a first key morpheme that searches for the key morpheme that completely matches the morpheme acquired in the morpheme acquisition step from the translation information storage unit that stores the key morpheme or the appearance frequency of the translated word in a predetermined document in association with each other. If the key morpheme is retrieved in the retrieval step and the first key morpheme retrieval step, before the corresponding key morpheme is retrieved If the key morpheme is not searched in the first translated word acquisition step and the first key morpheme search step to acquire the translated word from the translation information storage unit, the key morpheme including the morpheme acquired in the morpheme acquisition step is When a plurality of the key morphemes are searched for in the second key morpheme search step searched from the translation information storage means and the second key morpheme search step, the appearance frequency is the highest among the plurality of searched key morphemes. The second translation acquisition step of acquiring the translation corresponding to the high key morpheme from the translation information storage means, and the translation acquired in the first translation acquisition step or the second translation acquisition step is acquired in the morpheme acquisition step. A program for causing a computer to execute processing including a translated word output step for outputting a translated morpheme word is there.

  With such a configuration, when the program is read by the computer and the computer executes processing in accordance with the read program, an operation equivalent to that of the translation system of aspect 1 is obtained.

  [Invention 5] On the other hand, in order to achieve the above object, the translation method of Invention 5 is a translation method for translating a character string to be translated in the first language into the second language for each morpheme, and the morpheme acquisition means includes The morpheme acquisition step of acquiring a morpheme from the character string to be translated, and the first key morpheme search means, for a plurality of different key morpheme consisting of the first language, the key morpheme, the key morpheme to the second language The key morpheme that completely matches the morpheme acquired in the morpheme acquisition step from the translation information storage means that stores the translation obtained in the translation and the key morpheme in the predetermined document or the appearance frequency of the translation in association with each other. When the first key morpheme search step and the first translation acquisition unit search for the key morpheme in the first key morpheme search step, A first translation acquisition step for acquiring the translation corresponding to the issued key morpheme from the translation information storage unit; and a second key morpheme search unit that does not search for the key morpheme in the first key morpheme search step. A second key morpheme search step for searching for the key morpheme including the morpheme acquired in the morpheme acquisition step from the translation information storage unit, and a second translated word acquisition unit includes a plurality of the keys in the second key morpheme search step. A second translated word acquisition step of acquiring the translated word corresponding to the key morpheme having the highest appearance frequency from the translated information storage means when the morpheme is retrieved; The morpheme obtained by the means in the morpheme acquisition step is the translation acquired in the first translation acquisition step or the second translation acquisition step. A translated word output step for outputting as a translated word.

  As described above, according to the translation system of the invention 1, the translation program of the invention 4, or the translation method of the invention 5, it is sufficient to have translation information storage means for storing key morphemes, translated words, and appearance frequencies in association with each other. Therefore, it is not necessary to use a large number of libraries, and the used memory capacity can be reduced as compared with the conventional case. Moreover, since it is a simple algorithm of translating a character string to be translated into a second language for each morpheme, the processing load can be reduced as compared with the conventional case. Further, in addition to the translation by the first key morpheme search unit and the first translation acquisition unit, the translation by the second key morpheme search unit and the second translation acquisition unit in consideration of the appearance frequency of the key morpheme or translation is performed. A certain translation accuracy can be ensured while adopting a simple algorithm.

  Furthermore, according to the translation system of the invention 2, there is no need to use a large number of libraries because there is only need for translation information storage means for storing key morphemes, optimal translations, next candidate translations, and appearance frequencies in association with each other. As compared with this, the memory capacity used can be reduced. Moreover, since it is a simple algorithm of translating a character string to be translated into a second language for each morpheme, the processing load can be reduced as compared with the conventional case. Furthermore, in addition to the translation by the first key morpheme search unit and the first translation acquisition unit, the translation by the second key morpheme search unit and the second translation acquisition unit in consideration of the appearance frequency of the key morpheme or the optimal translation is performed. A certain translation accuracy can be secured while adopting a simple algorithm. Furthermore, since the optimal translation can be changed to the next candidate translation, usability can be improved.

  Furthermore, according to the translation system of the invention 3, since the morpheme part-of-speech utilization in the character string to be translated can be obtained, the translation accuracy can be improved.

1 is a diagram illustrating an external configuration of a mobile device 100. FIG. 2 is a functional block diagram of the mobile device 100. FIG. It is a figure which shows the data structure of the translation information table. It is a figure which shows the data structure of the Chinese character correction information table. It is a flowchart which shows imaging | photography translation processing. It is a flowchart which shows a morpheme translation process. It is a flowchart which shows a similar key morpheme search process. It is a sequence chart which shows the operation | movement which translates a translation object character string. It is a sequence chart which shows the operation | movement which translates a translation object character string. It is a screen which changes an optimal translation into a next candidate translation.

  Embodiments of the present invention will be described below. 1 to 10 are diagrams showing this embodiment.

  This embodiment shows a case where a character string to be translated in Japanese is translated into English for each morpheme. For example, the translation target string “Foreigners love Japanese food” is “foreigner” “(object)” “Japan” ““ 's ”” “cooking” “(subject)” “favourite (is ) ". Thus, the processing load is reduced by adopting a simple algorithm of translating for each morpheme.

First, the configuration of mobile device 100 according to the present embodiment will be described.
FIG. 1 is a diagram illustrating an external configuration of the mobile device 100. FIG. 2A is a front view of the mobile device 100, and FIG. 2B is a side view of the mobile device 100. FIG.

FIG. 2 is a functional block diagram of the mobile device 100.
As shown in FIG. 1, the mobile device 100 includes a cabinet 10 that includes a front surface and a back surface. A touch panel 21 is provided on the front surface of the cabinet 10. The touch panel 21 includes a display unit 28 that displays an image and a detection unit 29 that is superimposed on the display unit 28.

  The display unit 28 includes a liquid crystal panel and a panel backlight that illuminates the liquid crystal panel. The liquid crystal panel has a display surface 28a for displaying an image, and the display surface 28a appears outside. A detection unit 29 is provided on the display surface 28a. Instead of the liquid crystal panel, an organic EL (Electro Luminescence) panel or other display element can be used.

  The detection unit 29 includes a touch sensor or the like and is formed in a transparent sheet shape. The display surface 28 a can be seen through the detection unit 29. The detection unit 29 includes a first transparent electrode, a second transparent electrode, and a cover that are provided in a matrix. The detection unit 29 detects a change in capacitance between the first transparent electrode and the second transparent electrode, thereby detecting a position on the display surface 28a touched by the user (hereinafter referred to as “input position”). The position signal corresponding to the input position is output to the control unit 27. Note that the detection unit 29 is not limited to an electrostatic capacitance type, and an ultrasonic type, a pressure sensitive type, a resistance film type, a light detection type, or the like can also be used.

  The user touching the display surface 28a means that the user touches, slides, taps, flicks, etc. the display surface 28a with a contact member such as a finger or a pen (hereinafter simply referred to as “finger”). is there. Further, touching the display surface 28a actually means touching a region where an image of the display surface 28a is reflected on the surface of the cover covering the detection unit 29. “Slide” refers to an operation in which a user moves a finger while touching the display surface 28a. The “tap” is an operation in which a user touches a certain point on the display surface 28a so that the user taps the display surface 28a with a finger and releases the finger in a short time. The “flick” is an operation in which the user quickly flicks the display surface 28a with a finger, and refers to an operation in which the finger is moved by a predetermined distance or more in a short time while the finger is in contact with the display surface 28a. “Pinch out” refers to an operation in which the user touches the display surface 28a with two fingers and spreads the finger while touching the display surface 28a. “Pinch-in” refers to an operation in which the user touches the display surface 28a with two fingers and narrows the finger while touching the display surface 28a.

  A microphone 11 and a speaker 12 are provided on the front surface of the cabinet 10. The user can make a call by capturing the sound from the speaker 12 with his / her ear and emitting the sound to the microphone 11.

  A lens window (not shown) of the photographing unit 24 is provided on the back surface of the cabinet 10. An image of the subject is taken into the photographing unit 24 from the lens window.

  As shown in FIG. 2, the mobile device 100 includes a touch panel 21, a communication unit 22, a position information acquisition unit 23, a photographing unit 24, a storage unit 25, an output unit 26, and a control unit 27.

  The display unit 28 displays a GUI (Graphical User Interface) or the like used for a user operation under the control of the control unit 27. The detection unit 29 outputs coordinate information indicating the input position to the control unit 27. When inputting text, for example, when a text box or the like is tapped, a keyboard image (GUI) is displayed on the display unit 28, and the characters can be input by tapping characters arranged in the keyboard image. . In addition, text can be input by performing character recognition processing based on a photographed image photographed by the photographing unit 24 or performing speech recognition processing based on speech input by the microphone 11.

  The communication unit 22 performs data communication according to the wireless communication standard used in the mobile phone, and requests transmission of information from a server or the like via a network such as the Internet. The communication unit 22 receives information transmitted from a server or the like and outputs the information to the control unit 27.

  The position information acquisition unit 23 acquires position information specifying the current position of the mobile device 100 using a GPS (Global Positioning System), a wireless LAN (Local Area Network), or the like. For example, the position information acquisition unit 23 uses radio LAN radio waves transmitted from at least three base stations installed at different locations, and uses a triangulation method based on the position information of those base stations, The current position of the mobile device 100 is specified.

  The imaging unit 24 includes an imaging element such as a CCD (Charge Coupled Device) sensor or a CMOS (Complementary Metal Oxide Semiconductor) sensor, and takes a picture according to a user's shutter operation. When the photographing unit 24 takes a photograph, the control unit 27 stores the photographed photographed image in the storage unit 25.

  The storage unit 25 includes a nonvolatile semiconductor memory (for example, an EEPROM (Electronically Erasable and Programmable Read Only Memory)). The storage unit 25 stores a program (application software) to be executed by the control unit 27 and a photographed image photographed by the photographing unit 24.

  The control unit 27 is configured by bus-connecting a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), and the like.

Next, the data structure of the storage unit 25 will be described.
FIG. 3 is a diagram illustrating a data structure of the translation information table 400.

  As shown in FIG. 3, the storage unit 25 stores a translation information table 400 for translating the translation target character string into English. Thus, by using only one translation information table 400 for translation, the used memory capacity of the RAM is reduced.

  In the translation information table 400, one record is registered for each ID. Each record has a field 402 for registering a uniquely assigned ID, a field 404 for registering a key morpheme as a key, a field 406 for registering a key morpheme reading, and a field 408 for registering a part of speech of a key morpheme. A field 412 for registering the most suitable translation as a candidate for translation among a plurality of translations obtained when the key morpheme is translated into English, and a plurality of translations obtained when the key morpheme is translated into English It includes a field 410 for registering next candidate translations other than the optimal translation and a field 414 for registering the appearance frequency of key morphemes in a predetermined Japanese document. Here, the part of speech of the field 408 is, for example, “n” is a noun, “v” is a verb inflection, “inf” is a verb infinitive, “adj” is an adjective, “aux” is an auxiliary verb, and “adv” is an adverb. , “Conj” represents a conjunction, and “prep” represents a preposition.

  In the example of FIG. 3, the record in the seventh row contains “Retained” as the key morpheme, “Mochinaos” as the reading, “v5s, vt” as the part of speech, “to recover” as the optimal translation, “To recover / totally / to improve / to pick up / (P) /” is registered as a candidate translation, and “161” is registered as an appearance frequency. This is about the key morpheme “Takashi”, the reading is “Mochinaos”, the part of speech is the verb conjugation, the optimal translation is “to recover”, the next candidate translation is “to rally”, “to improve”, “to pick up” It shows that the frequency is 161 times.

FIG. 4 is a diagram showing a data structure of the kanji correction information table 420.
As shown in FIG. 4, the storage unit 25 further stores a kanji correction information table 420 for correcting a misrecognized character string in the translation target character string to an appropriate kanji.

  Each record of the Chinese character correction information table 420 includes a field 424 for registering a correction target character string, and a field 422 for registering a Chinese character for correcting the correction target character string included in the translation target character string. .

  In the example of FIG. 4, “ne” and “sada” are registered as the correction target character strings, and “と し て” is registered as the corrected kanji character in the first row field. This indicates that if two consecutive character strings “ne” and “sada” are included in the character string to be translated, they should be corrected to the Chinese character “禎”.

Next, processing executed by the mobile device 100 will be described.
In the mobile device 100, the CPU activates a program stored in the ROM or the storage unit 25, and executes the photographing / translation processing shown in the flowchart of FIG. 5 according to the program.

FIG. 5 is a flowchart showing the photographing / translation processing.
When the imaging / translation process is executed by the mobile device 100, as shown in FIG. 5, first, the process proceeds to step S100.

  In step S100, a photographed image photographed by the photographing unit 24 is read from the storage unit 25, and the process proceeds to step S102 to execute a character recognition process for recognizing characters included in the read photographed image. For this character recognition processing, for example, a known character recognition technology employed in OCR (Optical Character Recognition) processing or the like can be employed.

  Next, the process proceeds to step S104, and a morpheme translation process is performed in which the translation target character string recognized in the character recognition process in step S102 is translated into English for each morpheme. The optimal translation acquired in the morpheme translation process corresponding to each morpheme is displayed on the display unit 28, and the process proceeds to step S108.

  In step S108, it is determined whether any one of the displayed optimum translations has been specified. If it is determined that any optimum translation has been designated (YES), the process proceeds to step S110 and is designated. The next candidate translation corresponding to the optimum translation is displayed on the display unit 28, and the process proceeds to step S112.

  In step S112, it is determined whether any one of the displayed next candidate translations is selected. If it is determined that any next candidate translation has been selected (YES), the process proceeds to step S114. The designated optimal translation is changed to the selected next candidate translation, and the series of processes is terminated.

  On the other hand, if it is determined in step S112 that no next candidate translation is selected (NO), the process proceeds to step S108.

  On the other hand, if it is determined in step S108 that no optimal translation is designated (NO), the process waits in step S108.

Next, the morpheme translation process in step S104 will be described.
FIG. 6 is a flowchart showing the morpheme translation process.

  When the morpheme translation process is executed in step S104, the process first proceeds to step S200 as shown in FIG.

  In step S200, the character string to be translated is acquired from the character recognition process in step S102, and the process proceeds to step S202 to refer to the kanji correction information table 420 to appropriately identify the erroneously recognized character string in the character string to be translated. The character is corrected to kanji, and the process proceeds to step S204.

  In step S204, the value of the variable ind for storing the start position of the morpheme acquired in the translation target character string is set to “0”, the process proceeds to step S206, and the value of the variable ind is the length of the translation target character string. If it is determined that the length is smaller than the length of the character string to be translated (YES), the process proceeds to step S208.

  In step S208, the minimum value of the variable n is calculated. Specifically, (1) the number of characters of the longest key morpheme in the translation information table 400, (2) the number of characters from the position indicated by the value of the variable ind in the translation target character string to the end of the translation target character string, (3 ) The number of characters from the position indicated by the value of the variable ind in the character string to be translated to immediately before the next punctuation mark is calculated, and the minimum value among the calculated number of characters is set as the variable n.

  Next, the process proceeds to step S210, a character string corresponding to the length of the variable n is acquired as a sub-character string from the position indicated by the value of the variable ind in the translation target character string, and the process proceeds to step S212.

  In step S212, a key morpheme that completely matches the sub-character string is searched from the translation information table 400, the process proceeds to step S214, and it is determined whether or not the corresponding key morpheme has been searched for. When it is determined that it has been (YES), the process proceeds to step S216.

  In step S216, the part of speech corresponding to the searched key morpheme is acquired from the translation information table 400, and based on the acquired part of speech, all the part of speech utilization of the searched key morpheme is generated, and the process proceeds to step S218.

  In step S218, based on the generated part-of-speech use, the longest part-of-speech use that matches the character string after the sub-character string in the character string to be translated is searched from the character string to be translated. It is determined whether or not the utilization has been searched. If it is determined that the corresponding part of speech utilization has been searched (YES), the process proceeds to step S222.

  In step S222, the retrieved part-of-speech utilization character string is stored in an array, the process proceeds to step S224, a value corresponding to the retrieved part-of-speech utilization length is added to the variable ind, and the process proceeds to step S226.

  In step S226, the optimal translation and the next candidate translation corresponding to the key morpheme searched in step S212 are acquired from the translation information table 400, and the process proceeds to step S228, where the searched key morpheme and the acquired part of speech, optimal translation and The next candidate translated word is stored in the array, and the process proceeds to step S206.

  On the other hand, if it is determined in step S220 that the corresponding part-of-speech use is not searched (NO), the process proceeds to step S230, and a value corresponding to the length of the key morpheme searched in step S212 is added to the variable ind. Then, the process proceeds to step S226.

  On the other hand, if it is determined in step S214 that the corresponding key morpheme is not found (NO), the process proceeds to step S232, “1” is subtracted from the value of the variable n, and the process proceeds to step S234. It is determined whether or not the length of the character string is “0”, and if it is determined that the length of the sub-character string is not “0” (NO), the process proceeds to step S210.

  On the other hand, when it is determined in step S234 that the length of the sub character string is “0” (YES), the process proceeds to step S236.

  In step S236, a similar key morpheme search process for searching for a similar key morpheme that partially matches a character string around the sub character string is executed, and the process proceeds to step S238 to determine whether or not the corresponding key morpheme has been found. If it is determined that the corresponding key morpheme has been found (YES), the process proceeds to step S240.

  In step S240, when a plurality of key morphemes are retrieved, a key morpheme having the highest appearance frequency among the retrieved plurality of key morphemes is acquired from the translation information table 400, and the target character string in the translation target character string is obtained. Replace with the acquired key morpheme. When only one key morpheme is retrieved, the target character string in the translation target character string is replaced with the retrieved key morpheme. Here, the target character string refers to a sub-character string that is a source of searching for a key morpheme having the highest appearance frequency in the translation target character string.

When the process of step S240 ends, the process proceeds to step S206.
On the other hand, if it is determined in step S238 that the corresponding key morpheme is not found (NO), the process proceeds to step S242 to indicate that neither a key morpheme nor a similar key morpheme could be searched for “?”. Are stored in the array, the process proceeds to step S244, "1" is added to the value of the variable ind, and the process proceeds to step S206.

  On the other hand, if it is determined in step S206 that the value of the variable ind is greater than or equal to the length of the character string to be translated (NO), the process proceeds to step S246, and the original process (step S104) that called the morpheme translation process ( The array result is returned to (function), a series of processing is terminated, and the original processing is restored.

Next, the similar key morpheme search process in step S236 will be described.
FIG. 7 is a flowchart showing the similar key morpheme search process.

  When the similar key morpheme search process is executed in step S236, as shown in FIG. 7, first, the process proceeds to step S300.

  In step S300, the minimum value of the variable m is calculated as in step S208. Specifically, (1) the number of characters of the longest key morpheme in the translation information table 400, (2) the number of characters from the position indicated by the value of the variable ind in the translation target character string to the end of the translation target character string, (3 ) The number of characters from the position indicated by the value of the variable ind in the character string to be translated to the point immediately before the next punctuation mark is calculated, and the minimum value of the calculated number of characters is set as the variable m.

  Next, the process proceeds to step S302, where a character string corresponding to the length of the variable m is obtained from the position indicated by the value of the variable ind in the translation target character string as a sub-character string from the translation target character string, and the process proceeds to step S304.

  In step S304, a key morpheme including a sub-character string is searched from the translation information table 400, the process proceeds to step S306, it is determined whether or not the corresponding key morpheme has been searched, and the corresponding key morpheme has been searched. When it determines (YES), it transfers to step S308.

  In step S308, for each searched key morpheme, the appearance frequency of the key morpheme is acquired from the translation information table 400, all the searched key morphemes and the acquired appearance frequencies are stored in an array, and the process proceeds to step S310. Then, “1” is subtracted from the value of the variable m, and the process proceeds to step S312 to determine whether the length of the sub character string is “0”, and the length of the sub character string is not “0”. (NO), it transfers to step S302.

  On the other hand, if it is determined in step S312 that the length of the sub-character string is “0” (YES), the process proceeds to step S314, and the original process (function) that called the similar key morpheme search process in step S236. The array result is returned to, the series of processing is terminated, and the original processing is restored.

  On the other hand, if it is determined in step S306 that the corresponding key morpheme is not found (NO), the process proceeds to step S310.

Next, the operation of the present embodiment will be described.
An example will be described in which the translation target character string “Foreigners like Japanese food” is translated into English for each morpheme.

FIG. 8 is a sequence chart showing the operation of translating the character string to be translated.
In FIG. 8, processes 1 to 43 are shown. Hereinafter, processes 1 to 43 will be sequentially described.

<Process 1>
In step S200, the character string to be translated "foreigner likes Japanese cuisine" is acquired.

<Process 2>
The value of the variable ind is set to “0”. Since the variable n has a minimum number of 14 characters from the position indicated by the value of the variable ind in the translation target character string to immediately before the next punctuation mark, the value of the variable n is set to “14”. Then, in step S210, the substring “foreigner likes Japanese cuisine” is acquired. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 3>
As a result of the subtraction, since the value of the variable n becomes “13”, the sub-character string “Foreigners like Japanese cuisine” is acquired in step S210. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 4>
Sub-strings "Foreigners like Japanese food""Foreigners like Japanese food""Foreigners like Japanese food""Foreigners Japanese food" People are searched in order for “Japanese fee”, “Foreigners for Japan”, “Foreigners for Japan” and “Foreigners for Japan”, but there is a key morpheme in the translation information table 400 that exactly matches these substrings. Therefore, the value of the variable n is subtracted.

<Process 5>
As a result of the subtraction in the process 4, the value of the variable n becomes “4”, so that the sub character string “foreigner is” is acquired in step S210. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 6>
As a result of the subtraction, since the value of the variable n becomes “3”, the sub character string “foreigner” is acquired in step S210. Since a key morpheme that completely matches this sub-character string exists in the translation information table 400, the key morpheme “foreigner” is searched.

<Process 7>
In step S 216, the part of speech corresponding to the key morpheme “foreigner” is acquired from the translation information table 400.

<Process 8>
In step S216, based on the acquired part of speech, the key morpheme “foreigner” uses all parts of speech “not foreign”, “not foreign”, “not foreign”, “not foreign” “Foreigner” was generated, but the longest part-of-speech utilization that matches the character string after the sub-character string in the character string to be translated does not exist in the character string to be translated.

<Process 9>
In step S230, a value “3” corresponding to the length of the key morpheme “foreigner” is added to the variable ind.

<Process 10>
In steps S228 and S230, the key morpheme “foreigner”, the part of speech “noun”, the optimal translation “foreigner” and the next candidate translation “1) foreigner, 2) foreign citizen, 3) alien” are stored in the array.

<Process 11>
As a result of the addition, the value of the variable ind becomes “3”. The variable n has a minimum value of 11 characters from the position indicated by the value of the variable ind in the character string to be translated to immediately before the next punctuation mark, so the value of the variable n is set to “11”. Then, in step S210, the sub-character string “I like Japanese food” is acquired. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 12>
As a result of the subtraction, since the value of the variable n becomes “10”, the sub-character string “I like Japanese food” is acquired in step S210. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 13>
Similar to Processes 11-12, the substrings “I like Japanese food” “I like Japanese food” “Has Japanese food” “Has Japanese food” “Ha “Search is performed in order for“ is Japan ”, but since there is no key morpheme in the translation information table 400 that completely matches these sub-character strings, the value of the variable n is subtracted.

<Process 14>
As a result of the subtraction in the process 13, the value of the variable n becomes “2”, so that the sub-character string “has day” is acquired in step S210. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 15>
As a result of the subtraction, since the value of the variable n becomes “1”, the sub-character string “ha” is acquired in step S210. Since a key morpheme that completely matches the sub-character string exists in the translation information table 400, the key morpheme “ha” is searched.

<Process 16>
In step S 216, the part of speech corresponding to the key morpheme “ha” is acquired from the translation information table 400.

<Process 17>
In step S230, a value “1” corresponding to the length of the key morpheme “ha” is added to the variable ind. In steps S228 and S230, the key morpheme “ha”, the part of speech “particle”, the optimum translation “(object)”, and the next candidate translation “-” are stored in the array.

<Process 18>
As a result of the addition, the value of the variable ind becomes “4”. Since the variable n has a minimum number of 10 characters from the position indicated by the value of the variable ind in the character string to be translated to immediately before the next punctuation mark, the value of the variable n is set to “10”. Then, in step S210, the substring “I like Japanese cuisine” is acquired. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 19>
As a result of the subtraction, since the value of the variable n becomes “9”, the sub-character string “I like Japanese cuisine” is acquired in step S210. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 20>
As in the processes 18 to 19, the sub-strings “I like Japanese cuisine”, “Japanese cuisine likes”, “Japanese cuisine”, “Japanese cuisine”, “Japanese fee” are sequentially searched. Since the key morpheme that completely matches the sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 21>
As a result of the subtraction of the process 20, the value of the variable n becomes “3”, so that the sub-character string “Japan” is acquired in step S210. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 22>
As a result of the subtraction, since the value of the variable n becomes “2”, the sub-character string “Japan” is acquired in step S210. Since a key morpheme that completely matches this sub-character string exists in the translation information table 400, the key morpheme “Japan” is searched.

<Process 23>
In step S 216, the part of speech corresponding to the key morpheme “Japan” is acquired from the translation information table 400.

<Process 24>
In step S216, based on the acquired part of speech, all the parts of speech of the key morpheme “Japan” are “Not Japan”, “Not Japan”, “Not Japan”, “Not Japan”, “It was Japan” Although it is generated, the longest part of speech utilization that matches the character string after the sub-character string in the translation target character string does not exist in the translation target character string.

<Process 25>
In step S230, the value “2” corresponding to the length of the key morpheme “Japan” is added to the variable ind.

<Process 26>
In steps S228 and S230, the key morpheme “Japan”, the part of speech “noun”, the optimal translation “Japan”, and the next candidate translation “Japan” are stored in the array.

<Process 27>
As a result of the addition, the value of the variable ind becomes “6”. Since the variable n has a minimum number of 8 characters from the position indicated by the value of the variable ind in the character string to be translated to immediately before the next punctuation mark, the value of the variable n is set to “8”. Then, in step S210, the sub-character string “I like cooking” is acquired. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 28>
As in the process 27, the substrings “I like cooking”, “I like cooking”, “I like cooking”, “Cooking” and “Cooking” are sequentially searched, but these substrings are completely matched. Since there is no key morpheme in the translation information table 400, the value of the variable n is subtracted.

<Process 29>
As a result of the subtraction in the process 28, the value of the variable n becomes “2”. Therefore, the sub character string “no charge” is acquired in step S210. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 30>
As a result of the subtraction, since the value of the variable n becomes “1”, the sub character string “NO” is acquired in step S210. Since a key morpheme that completely matches the sub-character string exists in the translation information table 400, the key morpheme “NO” is searched.

<Process 31>
In step S 216, the part of speech corresponding to the key morpheme “NO” is acquired from the translation information table 400.

<Process 32>
In step S230, a value “1” corresponding to the length of the key morpheme “no” is added to the variable ind. In steps S228 and S230, the key morpheme “NO”, the part of speech “particle”, the optimum translation ““ 's ””, and the next candidate translation “-” are stored in the array.

<Process 33>
As a result of the addition, the value of the variable ind becomes “7”. Since the variable n has a minimum number of 7 characters from the position indicated by the value of the variable ind in the translation target character string to immediately before the next punctuation mark, the value of the variable n is set to “7”. Then, in step S210, the sub-character string “I like cooking” is acquired. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 34>
Similarly to the process 33, the sub-character strings “I like cooking”, “I like cooking”, “Cooking food”, and “Cooking” are sequentially searched. The key morphemes that completely match these sub-character strings are translated information. Since it does not exist in the table 400, the value of the variable n is subtracted.

<Process 35>
As a result of the subtraction of the process 34, the value of the variable n becomes “2”, and thus the sub character string “cooking” is acquired in step S210. Since a key morpheme that completely matches the sub-character string exists in the translation information table 400, the key morpheme “dish” is searched.

<Process 36>
In step S 216, the part of speech corresponding to the key morpheme “dish” is acquired from the translation information table 400. And in step S216, based on the acquired part of speech, all the parts of speech of the key morpheme “cooking” were used. “It was not cooking” “It was not cooking” “It was not cooking” “It was not cooking” “It was cooking ”Is generated, but the longest part-of-speech utilization that matches the character string after the sub-character string in the translation target character string does not exist in the translation target character string.

<Process 37>
In step S230, a value “2” corresponding to the length of the key morpheme “dish” is added to the variable ind.

<Process 38>
In steps S228 and S230, the key morpheme “cooking”, the part of speech “noun and suru verb”, the optimal translation “cooking” and the next candidate translation “1) cooking, 2) cuisine, 3) dealing with something, 4) handling, 5) administration, 6) management ”is stored in the array.

<Process 39>
As a result of the addition, the value of the variable ind becomes “9”. Since the variable n has a minimum number of 5 characters from the position indicated by the value of the variable ind in the character string to be translated to immediately before the next punctuation mark, the value of the variable n is set to “5”. Then, in step S210, the substring “I like you” is acquired. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 40>
As in the process 39, the sub-character strings “I like”, “I like”, and “I like” are searched in order, but there is no key morpheme that completely matches these sub-character strings in the translation information table 400. The value of variable n is subtracted.

<Process 41>
As a result of the subtraction in the process 40, the value of the variable n becomes “1”, so that the sub character string “ga” is acquired in step S210. Since a key morpheme that completely matches the sub-character string exists in the translation information table 400, the key morpheme “ga” is searched.

<Process 42>
In step S 216, the part of speech corresponding to the key morpheme “GA” is acquired from the translation information table 400.

<Process 43>
In step S230, a value “1” corresponding to the length of the key morpheme “ga” is added to the variable ind. In steps S228 and S230, the key morpheme “ga”, the part of speech “particle”, the optimum translation “(subject)”, and the next candidate translation “-” are stored in the array.

FIG. 9 is a sequence chart showing the operation of translating the character string to be translated.
In FIG. 9, processes 44-56 are shown. Hereinafter, the processes 44 to 56 will be sequentially described.

<Process 44>
As a result of the addition, the value of the variable ind becomes “10”. Since the variable n has a minimum number of 4 characters from the position indicated by the value of the variable ind in the translation target character string to immediately before the next punctuation mark, the value of the variable n is set to “4”. Then, the sub-character string “I like” is acquired in step S210. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

<Process 45>
Similarly to the process 44, the sub-character strings “like”, “like”, and “good” are searched in order, but there is no key morpheme that completely matches these sub-character strings in the translation information table 400. The value is subtracted.

<Process 46>
Since the length of the sub-character string is “0”, it is impossible to search for a completely matching key morpheme.

<Process 47>
Therefore, similar key morpheme search processing is executed in step S236.

<Process 48>
In steps S300 to S304, the value of the variable m changes in the range of “4” to “1”, and this substring is changed for each of the substrings “I like”, “I like”, “I like”, and “Good”. The key morpheme to be included is retrieved from the translation information table 400. Here, since the key morpheme including the sub-character string “like” exists in the translation information table 400, the key morpheme “love”, “things like”, and “people like” are searched.

<Process 49>
In step S308, the key morpheme “love” and the appearance frequency “1362”, the key morpheme “thing like” and the appearance frequency “197”, and the key morpheme “people like” and the appearance frequency “39” are stored in the array.

<Process 50>
Of the key morphemes “love”, “things like” and “people like”, the appearance frequency “1362” of “love” is the highest. Therefore, in step S240, the target character string “like” is the key morpheme “love”. Is replaced by

<Process 51>
Since the variable n has a minimum number of 5 characters from the position indicated by the value of the variable ind in the character string to be translated to immediately before the next punctuation mark, the value of the variable n is set to “5”. Then, in step S210, the sub-character string “I love you” is acquired. However, since a key morpheme that completely matches this sub-character string does not exist in the translation information table 400, the value of the variable n is subtracted.

  Similarly, the sub-character string “I love you” is searched in order, but since there is no key morpheme that completely matches this sub-character string in the translation information table 400, the value of the variable n is subtracted.

  As a result of the subtraction, the value of the variable n becomes “3”, so that the sub-character string “love” is acquired in step S210. Since a key morpheme that completely matches this sub-character string exists in the translation information table 400, the key morpheme “love” is searched.

<Process 52>
In step S 216, the part of speech corresponding to the key morpheme “love” is acquired from the translation information table 400.

<Process 53>
In step S216, based on the acquired part of speech, the key morpheme “I love” all the parts of speech “I did n’t love” “I don't love” “I loved” “I loved” “I didn't love” The longest part-of-speech utilization “I love you” that is generated and matches the character string after the sub-character string in the character string to be translated is searched.

<Process 54>
In step S230, a value “5” corresponding to the length of the key morpheme “I love you” is added to the variable ind.

<Process 55>
In steps S228 and S230, the key morphemes “I love” “I love you”, Part of speech “na-adjective”, Optimal translation “favourite (is)” and next candidate translation “1) loveable, 2) very likeable, 3) like very much / (P) / "is stored in the array.

<Process 56>
As a result of the addition, since the value of the variable ind is “15”, the array result is returned in step S246.

FIG. 10 is a screen for changing the optimal translation to the next candidate translation.
In step S106, as shown in FIG. 10, the optimal translation is displayed corresponding to each morpheme of the character string to be translated. Here, for example, when the user designates the optimal translation “foreigner” with a tap or the like, as shown in FIG. 5A, the next candidate translations are “(1) foreigner”, “(2) foreign citizen”, “(3) "alien" is displayed. The next candidate translation is displayed in, for example, a drop-down list 440. If there is another optimal translation in the next candidate translation, for example, when the user selects the next candidate translation “(3) alien”, the optimum translation “foreigner” is changed to the next candidate translation “alien”.

  For example, when the user designates the optimal translation “cooking” with a tap or the like, as shown in FIG. 5B, the next candidate translations are “(1) cooking”, “(2) cuisine”, and “(3) dealing with”. Something "" (4) handling "" (5) administration "" (6) management "is displayed. The next candidate translation is displayed in, for example, a drop-down list 442. If there is another optimal translation in the next candidate translation, for example, when the user selects the next candidate translation “(2) cuisine”, the optimal translation “cooking” is changed to the next candidate translation “cuisine”.

  In this way, in this embodiment, a morpheme is acquired as a sub-character string from a character string to be translated, and the key morpheme and the key morpheme are translated into the second language for a plurality of different key morphemes made of the first language. Corresponding to the most suitable translation as a candidate for translation among the multiple translations obtained in this case, the next candidate translation other than the optimal translation among multiple translations, and the appearance frequency of the key morpheme in a given Japanese document If the key morpheme that matches the sub-character string is searched from the translation information table 400 that is stored together, and the key morpheme is searched, the optimal translation corresponding to the searched key morpheme is acquired from the translation information table 400, If the key morpheme is not searched, the key morpheme including the substring is searched from the translation information table 400. If a plurality of key morphemes are searched, the key morpheme is searched. Get the best translation that frequency corresponding to the highest key morpheme among the number of keys morphemes from the translation information table 400, and outputs the acquired optimum translation as translation of the sub-string.

  As a result, the translation information table 400 that stores the key morpheme, the optimal translation, the next candidate translation, and the appearance frequency in association with each other is sufficient, and it is not necessary to use a large number of libraries. Can be reduced. Moreover, since it is a simple algorithm of translating a character string to be translated into a second language for each morpheme, the processing load can be reduced as compared with the conventional case. Furthermore, in addition to translation conversion by exact match search with key morphemes, translation conversion by partial match search with key morphemes considering appearance frequency is performed, so that a certain translation accuracy is ensured while adopting a simple algorithm. Can do.

  Furthermore, in the present embodiment, for the specified optimal translation among the acquired optimal translations, the next candidate translation corresponding to the optimal translation is displayed, and one of the displayed next candidate translations is selected and designated. The selected optimal translation is changed to the selected next candidate translation.

  Thereby, since an optimal translation can be changed into a next candidate translation, usability can be improved.

  Further, in the present embodiment, all the part-of-speech utilizations of the key morpheme are generated based on the part-of-speech corresponding to the searched key morpheme, and the sub-character strings after the sub-character string in the translation target character string are generated based on the generated part-of-speech utilization. The part-of-speech utilization that matches the character string is searched from the translation target character string, and the retrieved part-of-speech utilization is output.

  As a result, the morpheme part-of-speech utilization in the character string to be translated can be obtained, thereby improving the translation accuracy.

  In the present embodiment, translation information table 400 corresponds to the translation information storage means of inventions 1 to 5, step S110 corresponds to the next candidate translation output means of invention 2, and step S112 corresponds to the next candidate of invention 2. Corresponding to the translated word selecting means, step S114 corresponds to the translated word changing means of the second aspect. Step S210 corresponds to the morpheme acquisition unit or the morpheme acquisition step of inventions 1 to 3 or 5, and step S212 corresponds to the first key morpheme search unit of inventions 1 to 3 or 5 or the morpheme acquisition unit of inventions 4 or 5. Corresponding to the 1-key morpheme search step, step S214 corresponds to the first translation acquisition means of the invention 1, 2, or 5 or the first translation acquisition step of the invention 4 or 5.

  In the present embodiment, steps S304, S240, and S212 correspond to the second key morpheme search means of the invention 1, 2, or 5, or the second key morpheme search step of the invention 4 or 5, and step S214 includes This corresponds to the second translation acquisition means of the invention 1, 2, or 5, or the second translation acquisition step of the invention 4 or 5. Step S216 corresponds to the part-of-speech utilization generating unit of the invention 3, step S218 corresponds to the part-of-speech utilization searching unit of the invention 3, and step S228 is the translation output unit of the invention 1, 2, or 5 or the invention 4 Alternatively, it corresponds to 5 translated word output steps.

  In the present embodiment, step S228 corresponds to the part-of-speech utilization output means of invention 3, Japanese corresponds to the first language of invention 1, 2, 4 or 5, and English corresponds to inventions 1, 2 and 5. 4 or 5 second languages are supported.

  In the embodiment described above, the appearance frequency of key morphemes in a predetermined Japanese document is registered in the translation information table 400. However, the present invention is not limited to this, and the appearance frequency of the optimum translation word in a predetermined English document may be registered. it can.

  In the above embodiment and its modification, the key morpheme reading is not output. However, the present invention is not limited to this. In steps S226 and S228, the key morpheme corresponding to the key morpheme retrieved in step S212 is read. It is also possible to acquire the optimal translation and the next candidate translation from the translation information table 400 and store the retrieved key morpheme and the read key morpheme, the part of speech, the optimal translation and the next candidate translation in the array.

  Moreover, in the said embodiment and its modification, although implement | achieved as the mobile apparatus 100 which is a single apparatus, it can implement | achieve as not only this but a network system. A translation server and a plurality of user terminals are connected to a network such as the Internet. The translation server has a function equivalent to that of the mobile device 100 in the above embodiment. The difference is that information is received from the user terminal instead of input by the input device 40, and information is transmitted to the user terminal instead of display by the display device 44.

  In the above embodiment and its modifications, the present invention is applied to the case where a translation target character string made of Japanese is translated into English for each morpheme. However, the present invention is not limited thereto, and does not depart from the gist of the present invention. The scope is applicable to other cases. For example, the source language and the target language can be arbitrarily set.

DESCRIPTION OF SYMBOLS 100 ... Mobile device, 10 ... Cabinet, 11 ... Microphone, 12 ... Speaker, 21 ... Touch panel, 22 ... Communication part, 23 ... Location information acquisition part, 24 ... Shooting part, 25 ... Memory | storage part, 26 ... Output part, 27 ... Control unit 28 ... Display unit 28a ... Display surface 29 ... Detection unit 400 ... Translation information table 420 ... Kanji correction information table 402-414, 422, 424 ... Field, 440, 442 ... Drop-down list

Claims (5)

A translation system that translates a translation target character string in a first language into a second language for each morpheme,
Morpheme acquisition means for acquiring a morpheme from the character string to be translated;
For a plurality of different key morphemes in the first language, the key morpheme, a translation obtained when the key morpheme is translated into the second language, and the appearance frequency of the key morpheme or the translation in a predetermined document A first key morpheme search unit for searching for the key morpheme that completely matches the morpheme acquired by the morpheme acquisition unit from the translation information storage unit stored in association;
When the key morpheme is searched by the first key morpheme search unit, a first translation acquisition unit that acquires the translation corresponding to the searched key morpheme from the translation information storage unit;
If the key morpheme is not searched by the first key morpheme search unit, a second key morpheme search unit that searches the translation information storage unit for the key morpheme including the morpheme acquired by the morpheme acquisition unit;
When a plurality of the key morphemes are searched for by the second key morpheme search means, the translation information storage means for the translation corresponding to the key morpheme having the highest appearance frequency among the searched key morphemes A second translated word acquisition means acquired from
A translation system comprising: a translation output unit that outputs the translation acquired by the first translation acquisition unit or the second translation acquisition unit as a translation of the morpheme acquired by the morpheme acquisition unit. A translation system that translates a translation target character string in a first language into a second language for each morpheme,
Morpheme acquisition means for acquiring a morpheme from the character string to be translated;
For the plurality of different key morphemes of the first language, the optimal translation that should be most recommended as a candidate for the translation among the plurality of translations obtained when the key morpheme and the key morpheme are translated into the second language, the plurality Morphemes acquired by the morpheme acquisition unit from the translation information storage unit that stores the next candidate translations other than the optimal translation and the key morpheme or the appearance frequency of the optimal translation in a predetermined document. First key morpheme search means for searching for the key morpheme that completely matches;
When the key morpheme is searched by the first key morpheme search unit, a first translation acquisition unit that acquires the optimal translation corresponding to the searched key morpheme from the translation information storage unit;
If the key morpheme is not searched by the first key morpheme search unit, a second key morpheme search unit that searches the translation information storage unit for the key morpheme including the morpheme acquired by the morpheme acquisition unit;
When a plurality of the key morphemes are retrieved by the second key morpheme search means, the optimal translation corresponding to the key morpheme having the highest appearance frequency among the plurality of retrieved key morphemes is stored in the translation information storage A second translation acquisition means for acquiring from the means;
A translation output unit that outputs the optimal translation acquired by the first translation acquisition unit or the second translation acquisition unit as a translation of the morpheme acquired by the morpheme acquisition unit;
Next candidate translation output means for outputting the next candidate translation corresponding to the optimum translation for the optimum translation specified among the optimum translation acquired by the first translation acquisition means or the second translation acquisition means;
A next candidate translation selecting unit that selects any one of the next candidate translations output by the next candidate translation output unit;
A translation system comprising: a translation change unit that changes the designated optimal translation to a next candidate translation selected by the next candidate translation selection unit. In claim 2,
The translation information storage means stores a plurality of different key morphemes in association with the key morpheme, the part of speech of the key morpheme, the optimal translation, the next candidate translation, and the appearance frequency,
A part-of-speech utilization generating unit that generates all part-of-speech utilizations of the key morpheme based on the part of speech corresponding to the key morpheme retrieved by the first key morpheme search unit;
Part-of-speech utilization search that retrieves part-of-speech utilization that matches the character string after the morpheme acquired by the morpheme acquisition unit in the translation-target character string based on the part-of-speech utilization generated by the part-of-speech utilization generation unit Means,
A translation system comprising: a part-of-speech utilization output means for outputting the part-of-speech utilization retrieved by the part-of-speech utilization retrieval means. A translation program for executing a process of translating a character string to be translated in a first language into a second language for each morpheme,
A morpheme acquisition step of acquiring a morpheme from the translation target character string;
For a plurality of different key morphemes in the first language, the key morpheme, a translation obtained when the key morpheme is translated into the second language, and the appearance frequency of the key morpheme or the translation in a predetermined document A first key morpheme search step for searching for the key morpheme that completely matches the morpheme acquired in the morpheme acquisition step, from the translation information storage means that stores the associated information;
If the key morpheme is searched in the first key morpheme search step, a first translation acquisition step of acquiring the translation corresponding to the searched key morpheme from the translation information storage unit;
If the key morpheme is not searched in the first key morpheme search step, a second key morpheme search step of searching the translation information storage unit for the key morpheme including the morpheme acquired in the morpheme acquisition step;
When a plurality of the key morphemes are searched in the second key morpheme search step, the translation information storage means stores the translation corresponding to the key morpheme having the highest appearance frequency among the searched key morphemes A second translation acquisition step acquired from
A program for causing a computer to execute processing including a translation output step of outputting the translation acquired in the first translation acquisition step or the second translation acquisition step as a translation of the morpheme acquired in the morpheme acquisition step A translation program characterized by A translation method for translating a translation target character string in a first language into a second language for each morpheme,
A morpheme acquisition unit acquires a morpheme from the translation target character string; and
The first key morpheme search means, for a plurality of different key morphemes made of the first language, the key morpheme, a translated word obtained when the key morpheme is translated into the second language, and the key in a predetermined document A first key morpheme search step for searching for the key morpheme that completely matches the morpheme acquired in the morpheme acquisition step, from translation information storage means for storing the appearance frequency of the morpheme or the corresponding translation word;
A first translated word acquisition unit that acquires the translated word corresponding to the retrieved key morpheme from the translation information storage unit when the first translated word acquiring unit searches for the key morpheme in the first key morpheme search step; ,
If the second key morpheme search means does not search for the key morpheme in the first key morpheme search step, the second key morpheme search means searches for the key morpheme including the morpheme acquired in the morpheme acquisition step from the translation information storage means. A key morpheme search step;
When the second translation acquisition unit searches for the plurality of key morphemes in the second key morpheme search step, the second translation acquisition unit corresponds to the key morpheme having the highest appearance frequency among the plurality of searched key morphemes. A second translation acquisition step of acquiring a translation from the translation information storage means;
A translation method characterized in that the translation output means includes a translation output step of outputting the translation acquired in the first translation acquisition step or the second translation acquisition step as a translation of the morpheme acquired in the morpheme acquisition step. .