JP4890551B2 - Character conversion device and method for controlling character conversion device - Google Patents

Description

  The present invention is an apparatus for a user to input a character or a character string using a pointing device such as a keyboard, a touch panel, a tablet, or a mouse, receives a conversion instruction, and according to the input character or character string The present invention relates to a character conversion device that displays a converted character or character string so that the user can select the displayed character or character string, and allows the user to input a desired character or character string.

  In order for the user to input and display kanji and words such as Japanese and Chinese on the information device using the keyboard, the user can input readings (phonetic characters) for kanji and words and perform conversion instruction operations. It has been realized. In order to convert readings into kanji, dictionary data that associates readings with kanji and words is necessary, and the amount of vocabulary in the dictionary data is one aspect of the performance of the information device (input system).

  However, as the number of vocabularies increases, the information device increases the time for searching the dictionary data for kanji that matches the input reading. For this reason, the response from when the user inputs a reading and gives a conversion instruction until the conversion result is displayed deteriorates, resulting in difficulty in using the information device.

Therefore, ingenuity has been made to reduce the search time for kanji and words. For example, for words in the dictionary, the index information has the maximum number of readings of words starting with the reading of the first character for each first character of the reading, thereby limiting the search target for the input character string and There is a conversion device that reduces the amount of processing (for example, see Patent Document 1).
Japanese Patent Publication No. 6-149787 (published May 31, 1994)

  However, in the above-described conventional apparatus, if the number of input characters is within the maximum number of words that start reading the first character of the input character string recorded in the index information, It is necessary to check all the data to see if it matches all dictionary data. Therefore, when the number of vocabulary recorded in the dictionary data is increased in order to increase the number of vocabularies that can be converted, the number of kanji and words for each first reading character increases, and the processing time under the above conditions increases. Eventually, the response until display output is reduced, and the information device becomes difficult to use.

  The present invention has been made in view of the above-described problems, and its object is to provide a character conversion device, a character input method, a character input program, and a computer readable program that can shorten a word search time for an input phonetic character. Is to realize a simple recording medium.

In order to solve the above problems, a character conversion device according to the present invention converts an input character string made up of one or more input phonetic characters into a converted character, and can determine a word made up of the converted character. In
(A-1) conversion character identification information for identifying a conversion character and conversion character correspondence information in which conversion character phonetic information indicating the phonetic of the conversion character is associated, and (a-2) input And the converted character storage location information indicating the storage location in the converted character correspondence information of the converted character identification information having the phonetic character as the head of the converted character phonetic information. A converted character dictionary having a converted character index;
(B-1) word correspondence information in which word identification information for identifying a word and word phonetic information indicating the phonetic expression of the word are associated with each other, and (b-2) head conversion that becomes the head of the word First conversion character identification information for identifying a character is associated with word storage destination information indicating the storage destination of the word having the first conversion character indicated by the first conversion character identification information in the word correspondence information. A word dictionary having a word index,
(C) Conversion with reference to the converted character dictionary based on the first phonogram of the input character string that is input, and having the first phonogram of the input character string as the head of the converted character phonogram information First conversion character acquisition means for acquiring character identification information;
(D) word acquisition means for referring to the word dictionary based on the acquired head conversion character identification information and acquiring a word having the conversion character indicated by the head conversion character identification information at the head; It is characterized by that.

  According to the above configuration, the first conversion character acquisition unit first searches for the conversion character from the first phonogram of the input character string with reference to the conversion character dictionary, and then the word acquisition unit references the word dictionary. In addition, it is possible to acquire a word having the conversion character acquired by the head conversion character acquisition means at the head. Therefore, the time related to the word search process can be shortened. That is, since the number of comparisons between the input character string and the phonetic (reading) information can be limited, a character conversion device with a high response speed can be provided.

  Here, the conversion character indicates a character after conversion. If the converted character is kanji, the first character of the phonetic character (reading) is as follows. For example, in Chinese,

(Hereinafter referred to as “external character 1”) is “mainian”, and “m” is the first character of the reading. Also, in Japanese, the reading of “every year” is “Mai”, so the first character is “ma”.

  For example, in the case of “external character 1”, the leading conversion character is “

(Hereinafter referred to as “external character 2”). The converted character phonetic information is, for example, “mainian” in “external character 1”.

  The converted character identification information may be any information as long as it is information that can identify the converted character that is the character after conversion, or may be the converted character itself. For example, when the conversion character is a kanji, it may be a kanji itself or a kanji code. The same applies to the word identification information and the head conversion character identification information.

  Other objects, features, and advantages of the present invention will be fully understood from the following description. The benefits of the present invention will become apparent from the following description with reference to the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1, showing an embodiment of the present invention, is a block diagram showing a main configuration of a character conversion device. It is a figure which shows the flow of a process about the conversion of the character which the said character conversion apparatus performs. It is a figure explaining the Chinese character index which concerns on the Example of this invention. It is a figure explaining the kanji corresponding | compatible information which concerns on the Example of this invention. It is a figure explaining the primary index which concerns on the Example of this invention. It is a figure explaining the secondary index which concerns on the Example of this invention. It is a figure explaining the word corresponding | compatible information which concerns on the Example of this invention. It is a figure explaining the pinyin dictionary index based on the Example of this invention. It is a figure explaining the Pinyin dictionary which concerns on the Example of this invention. It is a figure showing the example of a pinyin division | segmentation process based on the Example of this invention. In the pinyin division process according to the embodiment of the present invention, FIG. It is a block diagram which shows the principal part structure of the character conversion apparatus which concerns on embodiment of this invention. 1 is a block diagram showing a schematic configuration of a multifunction printer to which a character conversion device of the present invention is applied. It is a figure which shows an example of the operation screen at the time of the character input process displayed on the operation panel of a character conversion apparatus. It is a figure which shows the example of the Chinese character dictionary recorded in the Chinese character dictionary recording part referred when the candidate extraction part of a character conversion apparatus extracts a candidate Chinese character. It is a block diagram which shows the principal part structure of the character conversion apparatus which extracts the candidate idiom which concerns on other embodiment of this invention. It is a figure which shows the example of the candidate idiom displayed on the display part of the operation panel of a character conversion apparatus. It is a figure which shows the example of the idiom dictionary which a idiom dictionary recording part records. It is a flowchart which shows the flow of a process of the character conversion apparatus which concerns on other embodiment of this invention. It is a figure which shows the example of a display of the candidate character display area and fixed character display area on an operation panel. It is a block diagram which shows the principal part structure of the character conversion apparatus provided with the learning function which concerns on other embodiment of this invention. 4 is a block diagram showing a more detailed configuration of a candidate order determination unit 32. FIG. It is a figure which shows the example of the connection degree table which showed the connection degree with the character string just before (front) for every kanji (jukugo) recorded in a connection degree recording part. It is a figure which shows the classification procedure of a candidate when the candidate order determination part of a character conversion apparatus determines a display order. It is a figure which shows an example of the data structure which memorize | stores the evaluation value of the display order calculated for every candidate. It is a figure which shows an example of the priority information recorded on a candidate order determination element recording part. It is a figure which shows the example of the kanji written in the character input frame of the operation panel, and the example of the coordinate information of this kanji. It is a figure which shows the example of a display of an operation panel, (a) of FIG. 28 is a figure which shows an example of the Chinese character written in the character input frame of a display-integrated tablet, (b) of FIG. 28 is a candidate character It is a figure which shows the example of the recognition character candidate displayed on the display area, (c) of FIG. 28 is a figure which shows the example of the candidate idiom displayed on the candidate character display area. It is a figure which shows the example of the recognition result of handwritten data "external character 2". It is a block diagram which shows the principal part structure of the character conversion apparatus which inputs the traditional character based on other embodiment of this invention. It is a figure which shows an example of the operation screen at the time of the phonetic character input process for the traditional character input in other embodiment of this invention. It is a figure which shows the example of the Chinese character dictionary recorded on the Chinese character dictionary recording part referred when the candidate extraction part of the character conversion apparatus in another embodiment of this invention extracts a candidate Chinese character. It is a figure which shows the example of the idiom dictionary which the idiom dictionary recording part of the character converter in other embodiment of this invention records. The figure which shows the example of the connection degree table | surface which showed the connection degree with the character string immediately before for every kanji (jukugo) recorded in the connection degree recording part of the character converter in other embodiment of this invention. is there. It is a figure which shows an example of the operation screen displayed on the display part of an operation panel. It is a figure which shows an example of the operation screen displayed on the display part of an operation panel. It is a figure which shows an example of the operation screen displayed on the display part of an operation panel. It is a figure which shows an example of the operation screen displayed on the display part of an operation panel.

  One embodiment (Embodiments 1 to 4) of the present invention is described below with reference to FIGS. In this embodiment, as an example, a case will be described in which the character conversion device of the present invention is applied to a multifunction printer (copier) having a touch panel in which an input device and an output device are integrated. However, the copying machine described in this embodiment is merely an example, and the character conversion device of the present invention can be applied to a mobile phone, a mobile terminal, a personal computer, a remote controller, and the like.

[Background technology and problems]
If an attempt is made to obtain an ideogram in a conventional character conversion device, the operation becomes more complicated than a phonetic character. There are an enormous amount of ideographic character types, and it is not realistic to assign keys one by one. Furthermore, it is common for ideograms to have many homonyms, and even when a plurality of phonograms are combined, the number of ideogram candidates corresponding to them is enormous. As a result, the task of selecting an appropriate ideographic character from a huge amount of candidates requires a lot of time and effort for the user.

  This problem does not occur only in a Japanese character input system, but similarly occurs in a similar ideographic character input system. For example, Chinese characters are input using Pinyin (four voices that are characteristic of Chinese characters are omitted). To enter kanji, follow the steps below. A pinyin notation (alphabet character string) of a kanji to be input is input from the keyboard, and a conversion to a kanji corresponding to the pinyin is performed to obtain a desired kanji. In order to solve the above problem, a character conversion device capable of simplifying an ideographic character input operation for obtaining a desired ideographic character will be described.

  In this embodiment, a character conversion device capable of inputting Chinese (simplified characters) from Pinyin notation (alphabet string) will be described. However, the character conversion device of the present invention is not limited to Chinese, and various phonetic characters are used. From which the desired ideogram can be obtained. Note that although the Pinyin notation originally has four voices representing four different pronunciations in Chinese, the concept of the four voices will be omitted for the sake of simplicity.

Embodiment 1
(Configuration of multi-function printer)
FIG. 13 is a block diagram showing a schematic configuration of a multifunction printer to which the character conversion apparatus 1 of the present invention is applied.

  As shown in FIG. 13, the multi-function printer includes a control unit 10, an operation panel (tablet) 2, a temporary storage unit 3, a recording unit 40, an image reading unit 4, an image transfer unit 5, and a paper feed unit 6. Note that the configuration of the multifunction printer described here is shown as an example for carrying out the present invention, and does not limit the scope of the present invention.

  The control unit 10 performs overall control of the multifunction printer. The control unit 10 reads out various programs recorded in the recording unit 40, controls the image reading unit 4, the image transfer unit 5, the paper feeding unit 6, and the like that function as a copying machine, and performs copying, etc. Perform the process. Each unit for performing the character input processing according to the present invention is also included in the control unit 10.

  The recording unit 40 records a control program for the above-described various units for operating as the operation panel 2 and a copying machine, and a control program for the character conversion device 1, and is a non-volatile memory such as a ROM (read only memory). Realized by a memory element.

  The temporary storage unit 3 is used for temporary storage of a work area for processing a program and a read image. In this embodiment, the temporary storage unit 3 is used for temporary storage of learning characters (strings) during character input processing. It is realized by a volatile memory such as a RAM (random access memory).

  The operation panel 2 is used by the user to instruct the multifunction printer to execute various processes. For example, it is realized by a touch panel in which a display device and an input device are integrally formed, or a display-integrated tablet. Furthermore, you may provide the handwritten character recognition function which recognizes the character written using the display integrated tablet. In addition, it has a function of detecting information such as the designated character, character string, and other function keys based on the position designated by the user on the display-integrated tablet.

  The user can input a desired alphabet into the multifunction printer by lightly touching each key displayed on the operation panel 2 corresponding to the alphabet (phonetic character) with a pen or a finger.

  Alternatively, the multifunction printer may include a keyboard (first input unit / third input unit) 7 (for example, a general-purpose keyboard used in a PC) in addition to the operation panel 2. In this case, the alphabet can be input using any of the keys displayed on the operation panel 2 and the keys on the keyboard 7. An input unit (first input unit) having a key for inputting a phonetic character can be realized simultaneously with both the operation panel 2 and the keyboard 7, or can be realized with only the keyboard 7. Or it is good also as a structure which can switch whether the input of a phonetic character is performed with the operation panel 2 or with the keyboard 7. FIG.

  Hereinafter, the character conversion device 1 of the present invention that performs the character input process will be described in more detail.

(Configuration of character conversion device)
FIG. 12 is a block diagram illustrating a main configuration of the character conversion device 1 according to the present embodiment. As shown in FIG. 12, the character conversion device 1 includes an operation panel 2, a control unit 10, a recording unit 40, and a temporary storage unit 3.

  As described above, the operation panel 2 is realized by a touch panel in which a display device and an input device are integrally formed.

  The display unit 12 provides various information to the user. FIG. 14 is a diagram illustrating an example of an operation screen during character input processing displayed on the operation panel 2. As shown in FIG. 14, the display unit 12 has a character input area 121 for displaying a keyboard for the user to input characters, and candidate characters (or character strings) for displaying characters (or character strings) listed as input candidates. Column) display area 122 and a confirmed character (column) display area 123 for displaying input confirmed characters that have already been entered and confirmed. The input unit 11 is for the user to input Chinese phonetic notation Pinyin (phonetic character string / phonetic information), and when the alphabet key displayed in the character input area 121 is pressed by the user. The alphabet signal corresponding to the pressed position on the panel is input to the input / output control unit 20 of the character conversion device 1. The input unit 11 is also used by the user to select a character for which input is to be confirmed. When a Chinese character (ideographic character / converted character) displayed in the candidate character display area 122 is pressed, a signal of the Chinese character displayed at the pressed position is input to the input / output control unit 20, and the input of the Chinese character is performed. Is confirmed.

  The control unit 10 controls the input / output control unit 20 and the character input control unit 30. The input / output control unit 20 controls input / output of signals exchanged between the operation panel 2 and the control unit 10, and includes an input receiving unit 21, a display control unit (display control means) 22, and candidates. A selection receiving unit 23 is included. The character input control unit 30 executes character input processing, and includes a candidate extraction unit (character candidate extraction unit / character string candidate extraction unit) 31, a candidate order determination unit (candidate order determination unit) 32, and a character confirmation. Part (character input confirmation means) 33 is included. Each unit in the control unit 10 is realized by a central processing unit (CPU) of a computer and software executed by the CPU. This software is also stored in the recording unit 40 described above, read into the temporary storage unit 3, and executed by the CPU.

  The input receiving unit 21 of the input / output control unit 20 receives a signal of a character (such as an alphabet) input from the input unit 11 and supplies it to the candidate extracting unit 31. The display control unit 22 controls the display unit 12 to display information to be displayed in each of the character input area 121, the candidate character display area 122, and the confirmed character display area 123 of the display unit 12. The candidate selection receiving unit 23 receives the signal of the selected candidate kanji in the candidate character display area 122 and the instruction signal for confirming the input of the selected candidate kanji input from the input unit 11, This is supplied to the determination unit 33.

  The candidate extraction unit 31 of the character input control unit 30 acquires candidate kanji (or) words from each dictionary unit stored in the recording unit 40 based on the input characters. In the present embodiment, the candidate extraction unit 31 extracts candidate Chinese characters from the Chinese character dictionary unit (character recording unit) 41 of the recording unit 40 based on the input alphabet (column).

  More specifically, first, pinyin corresponding to the input alphabet (column) is specified, and kanji corresponding to each specified pinyin is extracted as candidate kanji. The candidate extraction unit 31 stores the extracted candidate kanji in an internal buffer (not shown), and instructs the candidate order determination unit 32 and the display control unit 22 to display the candidate kanji.

  In the Chinese character dictionary unit 41 referred to by the candidate extracting unit 31, alphabet (column) -pinyin-kanji characters are recorded in association with each other. Details of the Chinese character dictionary unit 41 will be described later.

  The candidate order determination unit 32 determines the display order of the candidate kanji extracted by the candidate extraction unit 31 based on a predetermined condition. The display order is the order in which candidate Chinese characters are displayed in the candidate character display area 122, as shown in FIG. When all the extracted candidate kanji characters can be displayed in the candidate character display area 122 at a time, the upper candidate kanji characters are displayed at the top of the candidate character display area 122 so that the user can easily select them. do it. In addition, when the number of candidate kanji characters exceeds the number that can be displayed at one time, the pages may be switched so as to display candidate kanji characters in the order following the lower order. Even in this case, the upper candidate kanji characters are displayed on the first page so that the higher candidate kanji characters are more easily selected by the user. Therefore, the candidate order determination unit 32 determines the order in which the extracted candidate Kanji characters are displayed in order from those that are easy for the user to select based on a predetermined condition. The condition for determining the order of candidate kanji based on what conditions will be described in detail later.

  The character confirmation unit 33 receives the candidate kanji signal selected by the user input from the input unit 11 and the instruction signal for confirming the input of the selected candidate kanji from the candidate selection reception unit 23, and is selected by the user. The input of the candidate kanji is performed. First, the selected candidate kanji is displayed in the confirmed character display area 123 together with an underline indicating the selected state before confirmation. Further, the candidate Kanji characters determined by the character confirmation unit 33 according to the input confirmation instruction signal are displayed in the confirmed character display area 123 via the display control unit 22 without being underlined as input confirmation characters.

(Kanji dictionary)
Next, the kanji dictionary recorded in the kanji dictionary unit 41 will be described. FIG. 15 is a diagram illustrating an example of a Chinese character dictionary recorded in the Chinese character dictionary unit 41 that is referred to when the candidate extraction unit 31 extracts candidate Chinese characters.

  As shown in FIG. 15, the Chinese character dictionary 140 in the present embodiment is an alphabet-pinyin correspondence table (A / P correspondence table 141) in which a correspondence between alphabets (columns) received by the input receiving unit 21 and pinyin is recorded. And a Pinyin-Kanji correspondence table (P / C correspondence table 142) in which the correspondence between Pinyin and Kanji is recorded. This makes it possible to directly extract candidate Chinese characters corresponding to the input alphabet (column).

  Specifically, for example, it is assumed that the user first presses the “m” key using the input unit 11 in order to input the Chinese character “external character 2” (pinyin is “mei”). When the “m” signal is supplied from the input receiving unit 21, the candidate extracting unit 31 first searches the A / P correspondence table 141 of the Chinese character dictionary 140 from the pinyin (in the example of FIG. 15) that matches forward with “m”. , “Ma” and “mei”), and the address of the P / C correspondence table 142 where the Chinese character corresponding to the pinyin is recorded is detected (“cp_ml” and “cp_mi”).

  Subsequently, based on “cp_ml” and “cp_mi” in the P / C correspondence table 142, the candidate extraction unit 31 selects a Chinese character (

、 Hemp 、 December ... "Gaiji 2"

...) is extracted. The extracted Chinese characters are supplied to the candidate order determination unit 32 as candidate Chinese characters as described above. Here, when “mei” is further input by the user, the candidate extraction unit 31 extracts candidate kanji from the kanji dictionary unit 41 by narrowing down to kanji characters that coincide with “mei”.

  According to the above configuration, the candidate extraction unit 31 receives an alphabet input by the user and specifies Pinyin including the alphabet. The candidate extraction unit 31 can extract, from the kanji dictionary unit 41, candidates for kanji that the user desires to input based on the specified pinyin.

  The candidate order determination unit 32 rearranges the kanji candidates extracted by the candidate extraction unit 31 based on a predetermined condition, and determines a candidate display order. The display control unit 22 controls the display unit 12 so that kanji candidates are displayed in the display order determined by the candidate order determination unit 32.

  Finally, when a desired Chinese character is designated from the displayed candidate Chinese characters to be input, the character confirmation unit 33 confirms the input of the designated Chinese character.

  Thereby, it becomes possible to simplify input operation for obtaining a desired kanji.

  In addition, the kanji characters that are candidates for extraction by the candidate extraction unit 31 may be stored in an external storage device (not shown) and read from the external storage device by a user's alphabet input operation. The read Chinese characters may be used when displaying candidates.

(Modification 1—Output of candidate word)
Furthermore, the character conversion device 1 of the present invention is configured to select a Chinese character string (hereinafter referred to as a word (ideographic character string)) having one or more Chinese characters having one meaning after the Chinese character is selected as an input confirmed character. Is displayed as a candidate word in the candidate character display area 122 (FIG. 14). Accordingly, the user can obtain kanji characters that can form words together with the kanji characters in the candidate character display area 122, so that it is possible to obtain a desired kanji character without separately inputting the subsequent kanji pinyin. Become. Therefore, it is possible to simplify the input operation for obtaining a desired kanji.

  FIG. 16 is a block diagram illustrating a main configuration of the character conversion apparatus 1 that outputs candidate words. In addition, the code | symbol attached | subjected to each component of FIG. 16 respond | corresponds to the code | symbol attached | subjected to each component of FIG. 12, and the same code | symbol has shown the same component. Therefore, the description about the component already demonstrated by each above-mentioned embodiment is not repeated. In the character conversion device 1 shown in FIG. 16, the difference from FIG. 12 is that a control signal is transmitted from the character determination unit 33 to the candidate extraction unit 31 and a word dictionary unit (character string recording unit) 42 in the recording unit 40. Is included.

  When the Chinese character (or input confirmed character) selected by the user is input via the candidate selection receiving unit 23, the character determination unit 33 continues to the candidate extraction unit 31 and starts with the selected word Send instructions to extract.

  The candidate extraction unit 31 extracts a word starting from the selected kanji character from the word dictionary unit 42 as a candidate word based on the above instruction. FIG. 17 is a diagram illustrating an example of candidate words displayed on the display unit 12 of the operation panel 2. In the example shown in FIG. 17, when the Chinese character “external character 2” is selected, a word starting from “external character 2” is displayed as a candidate word in the candidate character display area 122 of the display unit 12.

  In the word dictionary unit 42 referred to by the candidate extraction unit 31, kanji-words are recorded in association with each other. FIG. 18 is a diagram illustrating an example of a word dictionary recorded by the word dictionary unit 42. As shown in FIG. 18, the word dictionary in the present embodiment is a kanji-word correspondence table (C / I correspondence table) in which associations between kanji recorded in the kanji dictionary unit 41 and words starting from the kanji are recorded. 143). This makes it possible to further extract candidate words corresponding to the selected kanji.

  Specifically, for example, when the user selects the Chinese character “external character 2”, the candidate extraction unit 31 detects the address of “external character 2” in the P / C correspondence table 142 (FIG. 15) (“wp_mi + j”). "). Subsequently, referring to the word dictionary unit 42, based on the address “wp_mi + j”, a word starting from “external character 2” (“external character 2” month, “external character 2” heaven...) Is extracted. The extracted words are supplied as candidate words to the candidate order determining unit 32 and displayed in the candidate character display area 122 as in the above-described configuration (FIG. 17).

  When a certain Chinese character is input, the character that follows the Chinese character (the Chinese character desired by the user) is more likely to be a Chinese character that constitutes a word starting from the Chinese character than an unrelated kanji. Therefore, if a word starting from a kanji that has been confirmed to be input is displayed as a candidate, the user can select the desired word from the word candidates and input Chinese without inputting pinyin separately. It becomes possible. Thus, the user can obtain a desired kanji input with a smaller number of operations related to the pinyin input.

(flowchart)
Next, the flow of processing of the character conversion device 1 (FIG. 16) of the present invention is shown in the flowchart shown in FIG. 19 and display examples of the candidate character display area 122 and the confirmed character display area 123 on the operation panel 2 shown in FIG. Based on In the following, a case where the user tries to input ““ External character 2 ”walk to the suburbs” will be described.

  When the user inputs an alphabet (the first alphabet “m” of Pinyin of the Chinese character “external character 2” (mei)) from the input unit 11 of the operation panel 2 (YES in S201), the input reception unit 21 receives the input alphabet “m”. ”Is supplied to the candidate extraction unit 31. Next, the candidate extraction unit 31 specifies Pinyin corresponding to the input alphabet, and extracts a Chinese character corresponding to the specified Pinyin from the Chinese character dictionary unit 41 as a candidate Chinese character (S202).

  The candidate order determination unit 32 determines the display order of the extracted candidate Chinese characters based on a predetermined condition (details will be described later) (S203). Subsequently, the display control unit 22 determines a display position so that n candidate kanji characters are displayed in the display order determined by the candidate order determining unit 32, outputs the candidate kanji characters to the display unit 12, and presents them to the user ( S204). FIG. 20A shows a display example in the case where n = 5 and five Chinese character candidates are displayed.

  Here, the user does not have an instruction to display the next candidate kanji other than the kanji displayed in the candidate character display area 122 for the first time (since there is no desired kanji “external character 2” in FIG. When a “next candidate display” button is pressed (NO in S205, YES in S206), the process returns to S204, and the display control unit 22 displays a list of next candidate kanji characters (( b)). Furthermore, when neither the next candidate kanji display instruction nor the candidate kanji selection instruction is input (NO in S205 and S206), if a new alphabet is input, the process returns to S202, and the candidate extraction unit 31 has so far performed. The pinyin corresponding to the alphabet string input to is further narrowed down and specified, and the kanji corresponding to the specified pinyin is extracted as a candidate kanji.

  On the other hand, when the user selects a desired Chinese character “external character 2” (“3011” in FIG. 20B) displayed in the candidate character display area 122 via the input unit 11 (YES in S205), the candidate is selected. The selection receiving unit 23 receives a signal of the selected candidate Chinese character “external character 2” and supplies it to the character determination unit 33.

  The character determination unit 33 instructs the display control unit 22 to display the selected “external character 2” and also instructs the candidate extraction unit 31 to extract a word starting from “external character 2” next. Subsequently, the candidate extraction unit 31 extracts a word starting from “external character 2” as a candidate word from the word dictionary unit 42 (S207) ((c) of FIG. 20).

  The candidate order determination unit 32 determines the display order of the extracted candidate words based on a predetermined condition (details will be described later) (S208). Subsequently, the display control unit 22 outputs the candidate words to the display unit 12 in the display order determined by the candidate order determination unit 32 and presents them to the user (S209).

  Here, when an instruction to display the next candidate word is input by the user (NO in S210, YES in S211), the process returns to S209, and the display control unit 22 displays a list of next candidate words. Further, when neither a next candidate word display instruction nor a candidate word selection instruction is input (NO in S210 and S211) and a new alphabet is input, the process returns to S202, and the candidate extraction unit 31 has so far Pinyin corresponding to the alphabet input in is identified, and kanji corresponding to the identified pinyin is extracted as candidate kanji.

  On the other hand, when the user selects the desired word ““ external character 2 ”heaven” (“020” in FIG. 20C) displayed in the candidate character display area 122 via the input unit 11 (YES in S210). ), The candidate selection receiving unit 23 receives a signal of the selected candidate word ““ external character 2 ”heaven” and supplies it to the character determination unit 33.

  The character determination unit 33 instructs the display control unit 22 to display the selected candidate word, and the display control unit 22 displays the selected candidate word on the display unit 12 (S212) ((d in FIG. 20) )). At the same time, the candidate extraction unit 31 is instructed to extract the next word starting from the candidate word. If an instruction to end character input is not input (NO in S213), the candidate extraction unit 31 returns to S207, and extracts a word starting from the selected candidate word from the word dictionary unit 42 as a candidate word.

  When the user's next desired Chinese character “To” is not displayed as a candidate Chinese character, the user can input the first Pinyin “d” of “To” (dao) (NO in S211, YES in S201) (FIG. 20 (e)). Thereafter, it is possible to input a desired kanji by repeating the same processing.

  After the selected candidate word is selected (after input is confirmed until “walk”), if an instruction to end character input is input (YES in S213), the process ends.

(Modification 2-Learning function)
Furthermore, it is preferable that the character conversion device 1 of the present invention has a configuration for recording kanji (words) whose input is determined by the user's selection. Thereby, the character conversion apparatus 1 can display a candidate Chinese character based on the recording of the Chinese character input in the past, ie, the Chinese character which the user has used. There is a high possibility that Chinese characters used in the past will be used again by the user. Therefore, by using a list of candidate kanji that reflects such usage history, the user can obtain a desired kanji with a small number of operations.

  FIG. 21 is a block diagram showing a main configuration of the character conversion device 1 having a learning function. Note that the reference numerals given to the respective constituent elements in FIG. 21 correspond to the reference numerals given to the respective constituent elements in FIG. 12, and the same reference numerals indicate the same constituent elements. Therefore, the description about the component already demonstrated by each above-mentioned embodiment is not repeated. In the character conversion device 1 shown in FIG. 21, the difference from FIG. 12 is that the character input control unit 30 further has a learning control unit (learning control means) 34 and that the recording unit 40 records the learning dictionary. Part (learning word recording part) 43 and pair learning dictionary recording part (learning word recording part) 44 are included.

  The learning control unit 34 controls recording (learning) of Chinese characters input (used) to the character conversion device 1 in accordance with a user instruction. Specifically, the kanji used by the user is stored as a learning kanji in the learning dictionary recording unit 43 (hereinafter referred to as independent learning (function)), or the kanji used by the user is input immediately before. It is stored in the pair learning dictionary recording unit 44 in association with the kanji that has been set (hereinafter referred to as pair learning (function)). Kanji characters that have been input by the user in the past are highly likely to be used again by the user, so by using such learning kanji as candidate kanji, the user can obtain a desired kanji with fewer operations. Increases the probability of getting.

  The learning dictionary recording unit 43 records kanji (input confirmed characters) used in the past by the user as independent learning kanji. The learning dictionary recorded in the learning dictionary recording unit 43 includes information on the date and time when the single learning kanji was recorded, information on the usage frequency (how many times it was used in the past, how many times it was used within a predetermined period, etc. ) May be recorded in association with each Chinese character.

  When the learning control unit 34 receives a candidate kanji signal selected by the user from the character determination unit 33, the learning control unit 34 writes the selected candidate kanji into the temporary storage unit 3. The learning control unit 34 receives an instruction for a single learning process for the input confirmed character from the character confirmation unit 33 when the character confirmation unit 33 confirms the input according to the confirmation instruction signal. The learning control unit 34 reads the candidate kanji from the temporary storage unit 3 into the internal buffer, and deletes the candidate kanji read from the temporary storage unit 3 and updates the content. Then, the candidate kanji stored in the internal buffer is stored in the learning dictionary recording unit 43 as a single learning kanji.

  The pair learning dictionary recording unit 44 records kanji used in the past by the user as a pair learning kanji in association with information (preceding kanji) indicating which kanji is determined following the kanji. . The immediately preceding kanji has already been recorded as a single learning kanji in the learning dictionary recording unit 43. In the learning dictionary recorded in the pair learning dictionary recording unit 44, the preceding kanji is recorded in association with each pair learning kanji. For each pair of Kanji characters, recorded date and time information and usage frequency information may be recorded in association with each other.

  The learning control unit 34 receives an instruction of the input confirmed character pair learning process from the character confirming unit 33. The learning control unit 34 reads the input confirmed character and the previous kanji character from the temporary storage unit 3 into the internal buffer, and deletes the character string read from the temporary storage unit 3 and updates the content. Then, the input confirmed character stored in the internal buffer is associated with the preceding kanji and stored in the pair learning dictionary recording unit 44 as a pair learning kanji.

  In the above description, the learning dictionary recording unit 43 and the pair learning dictionary recording unit 44 have described the configuration in which each learning dictionary is recorded for each Chinese character. However, the present invention is not limited to this. In addition to the above configuration, the word may be recorded in a dictionary. That is, the used word may be recorded in the learning dictionary recording unit 43 by the single learning function of the learning control unit 34. For example, when the word “water supply” is used, the word “water supply” is recorded in the learning dictionary recording unit 43 as a single learning kanji (word) in addition to the kanji “supply”.

  Further, according to the pair learning function of the learning control unit 34, when the words “university” and “professor” are continuously input, the kanji “teaching” is changed to “kanji” and / or “university” immediately before the kanji (word ) And can be recorded as pair learning kanji. Furthermore, the word “professor” can be recorded as a pair learning word in association with the previous kanji (word) of “study” and / or “university”.

(Candidate order determining element)
Next, a predetermined condition (hereinafter referred to as a candidate order determination element) for the candidate order determination unit 32 (FIGS. 12, 16, and 21) of the character conversion device 1 to determine the display order of candidates (candidate kanji / candidate words). ).

  As described above, in the display unit 12 provided on the operation panel 2, there is an advantage that the possibility that the user can obtain a desired Chinese character earlier is higher when more candidates can be displayed at once. In addition, when the operation panel 2 is realized by a display / input integrated input / output device such as a touch panel, the number of operations for the user to obtain a desired kanji can be significantly reduced. However, on the other hand, even if many candidates are displayed as a list on a large display screen, there arises a problem that the burden of searching for a desired candidate from among them increases. Therefore, when displaying a limited number of candidates in a limited display area, it is desirable that the kanji (words) that are most likely to be desired (selected) by the user are displayed in order.

  In the present invention, the order of candidates to be displayed is determined so that the candidates are displayed in a more accurate display order that matches the user's preference. Thereby, the character conversion apparatus which can simplify input operation for obtaining a desired Chinese character is realizable.

  In the following, a method of determining the display order of candidate words will be described. However, the display order of any candidate kanji / candidate word is determined by the same method.

  In this embodiment, the candidate order determination unit 32 determines a candidate display order based on the following seven candidate order determination elements.

(Element A) Appearance Frequency Appearance frequency is information that serves as an index of how often a character (string) in a language appears. The higher the appearance frequency, the more frequently it is used in the language usage environment. Therefore, the higher the appearance frequency, the higher the display order. The appearance frequency may be fixed information using known statistical data.

(Element B) Pinyin length match / mismatch The display order is determined based on whether the number of characters in the input alphabet (column) matches the number of characters in the corresponding candidate Pinyin. The display order of candidates whose pinyin length matches the number of input pinyin characters is made higher.

(Element C) Continuity with the immediately preceding character string Possibility of continuing immediately after the input confirmed character and / or the currently selected character (string) at the current time (continuation with the immediately preceding character string) The higher the candidate), the higher the display order. Alternatively, from the past input history, the presence / absence of continuous input to the immediately preceding character string is determined, and the display order of candidates in which the history input continuously remains remains higher.

(Element D) Relevance to the forward character string There is a possibility that an input is confirmed after the character (string) in relation to the input confirmed character (string) that has already been confirmed by that time (the forward character string and The display order is set higher in the candidate with the greater relevance).

(Element E) Single learning kanji / word The display order is determined depending on whether or not it is a single learning kanji (word) recorded in the learning dictionary recording unit 43. If a Chinese character (word) that has been input (used) in the past is likely to be used again. Therefore, the display order of candidates recorded as independent learning kanji (words) is set higher.

(Element F) Pair learning kanji / word The display order is determined depending on whether or not the pair learning kanji (word) is recorded in the pair learning dictionary recording unit 44. If a combination of Chinese characters (words) that have been input (used) in the past is likely to be used again in that combination. Therefore, the display order of candidates recorded as pair learning kanji (words) is set higher.

(Element G) Learning Time The display order is determined based on the time when the learning kanji (word) matching the candidate kanji (word) is recorded in the learning dictionary recording unit 43 and / or the pair learning dictionary recording unit 44. The newer the recorded time, the greater the possibility of selection by the user. Therefore, the more recently recorded Chinese character (word) candidates, the higher the display order.

  By setting each of the above elements as a condition, a candidate that is more likely to be selected by the user can be preferentially displayed in the candidate character display area 122 (FIG. 14). Thus, it becomes possible to obtain a desired kanji.

  As described above, the candidate order determination element is information necessary for the candidate order determination unit 32 to determine the display order of candidate kanji / candidate words. More specifically, this is information used by the candidate order determination unit 32 to calculate an evaluation value indicating the likelihood that a candidate kanji / candidate word will be selected by the user. Information on candidate order determination elements is associated with each candidate (candidate kanji / candidate word) and stored in each (recording) unit of the recording unit 40. The candidate order determination unit 32 refers to each unit of the recording unit 40. Thus, it becomes possible to specify a candidate order determination element for each candidate. Furthermore, the recording unit 40 may store an evaluation value calculation rule that defines how the evaluation value is calculated from the candidate order determination element specified by the candidate order determination unit 32. Thereby, the candidate order determination unit 32 can obtain the evaluation value for each candidate according to the identified candidate order determination element and the evaluation value calculation rule, and determine the display order of the candidates. The evaluation value calculation method of the candidate order determination unit 32 will be described later.

(Configuration of candidate order determination unit)
FIG. 22 is a block diagram showing a more detailed configuration of the candidate order determination unit 32. The candidate order determination unit 32 includes a learning kanji determination unit 131, a pinyin length determination unit 132, a connectivity determination unit 133, an appearance frequency determination unit 134, and a learning time determination unit 135. The recording unit 40 (FIGS. 12, 16, and 21) includes an appearance frequency recording unit 45 and a connection degree recording unit (relevance degree recording unit) 46 in addition to the storage units described above. The temporary storage unit 3 includes an extraction candidate temporary storage unit 51, an evaluation value temporary storage unit 52, and a display order temporary storage unit 53 therein.

  The extraction candidate temporary storage unit 51 in the temporary storage unit 3 temporarily holds kanji and / or words as candidates extracted by the candidate extraction unit 31 (hereinafter referred to as extraction candidates). The candidate order determination unit 32 reads out extraction candidates from the extraction candidate temporary storage unit 51 and performs a process of determining the display order of each extraction candidate.

  The evaluation value temporary storage unit 52 temporarily stores the evaluation value given to each extraction candidate calculated based on the candidate order determination element and the evaluation value calculation rule. The display order may be determined based on the finally calculated evaluation value.

  The display order temporary storage unit 53 temporarily holds information on the display order determined for each extraction candidate. The candidate order determination unit 32 stores the determined display order in the display order temporary storage unit 53. The display order information in the display order temporary storage unit 53 is referred to by the display control unit 22 when the extraction candidates are displayed on the display unit 12.

  The learning kanji determination unit 131 of the candidate order determination unit 32 determines whether or not the extraction candidate is a learning kanji (word) and whether it is single learning or pair learning. Specifically, if a kanji (word) that matches the extracted candidate is recorded in the learning dictionary recording unit 43, it is determined that it is a single learning kanji (word), and the pair learning dictionary recording unit 44 Is recorded as a paired kanji (word). The learning kanji determination unit 131 reflects the determination result in the evaluation value in the evaluation value temporary storage unit 52 or the display order in the display order temporary storage unit 53. Thereby, the display order based on the element E and / or the element F can be determined.

  The pinyin length determination unit 132 determines whether or not the number of characters in the alphabet (column) input to the input receiving unit 21 (FIGS. 12, 16, and 21) matches the number of characters in the corresponding candidate pinyin. . For example, suppose that “me” has been input to obtain the Chinese character “External character 2” (pinyin: mei). At this time, since the number of input characters is 2 and the pinyin length of the corresponding candidate Chinese character ““ external character 2 ”(mei)” is 3, the pinyin length determination unit 132 determines that the pinyin length does not match. On the other hand, when the user inputs “mei”, it is determined that the Pinyin lengths match, and the evaluation value or the display order in the evaluation value temporary storage unit 52 is temporarily increased so as to increase the display order of the candidate kanji ““ External character 2 ””. The display order of the storage unit 53 is changed. Thereby, the display order based on the element B can be determined.

  The connection degree determination unit 133 determines whether the connection with the input confirmed character and / or the currently selected character (kanji / word) has been confirmed until that time, that is, according to the degree of connection. Predicts a desired character and makes a determination as a candidate. The connectivity determination unit 133 refers to the connectivity table 150 recorded in the connectivity recording unit 46 to determine the connectivity with the confirmed or selected forward character string and / or immediately preceding character string. The higher the degree of connectivity with the preceding character string and / or the immediately preceding character string, the higher the probability that the character is desired by the user, and the higher the display order. Thereby, the display order based on the element C and / or the element D can be determined. Details of the connectivity table 150 and the connectivity determination unit 133 of the connectivity recording unit 46 will be described later.

  The appearance frequency determination unit 134 determines the frequency at which candidate kanji (words) appear (use) based on known statistical data recorded in the appearance frequency recording unit 45. The higher the appearance frequency, the more frequently the character is used, so the possibility of being selected by the user is high. Therefore, the higher the appearance frequency, the higher the display order. Thereby, the display order based on the element A can be determined.

  The learning time determination unit 135 determines the time when the learning kanji (word) matching the candidate kanji (word) was recorded in the learning dictionary recording unit 43 and / or the pair learning dictionary recording unit 44. The newer the time when the learning kanji (word) is recorded, the more likely it is to be selected by the user. Therefore, the display order of candidates is determined so that the more recent kanji (word) is at the higher learning time. Thereby, the display order based on the element G can be determined.

  According to the above configuration, the candidate order determination unit 32 can determine the display order of candidates based on the various conditions described above.

  This makes it possible to display candidates in a display order that more accurately reflects the user's wishes. As a result, the number of operations related to the input of phonograms can be input with a smaller number of operations. Can be obtained.

(Determining connectivity)
Next, the operation of the connectivity determination unit 133 will be described in more detail.

  FIG. 23 is a diagram showing an example of the connectivity table 150 showing the connectivity with the immediately preceding (front) character string for each Chinese character (word) recorded in the connectivity recording unit 46. As shown in FIG. 23, the connectivity table 150 detects the connectivity of the continuity information with the immediately preceding character string and the address of the relevance information with the preceding character string stored for each Chinese character (word). A correspondence table 151, a continuity table 152 that stores continuity with the immediately preceding character string for each Chinese character (word), and a forward relationship table 153 that stores the relationship with the forward character string for each Chinese character (word). Contains.

  For example, when the word “traffic” is selected by the user, the connectivity determination unit 133 searches the connectivity table 150 for a kanji (word) having a high probability of being continued after “traffic” as the immediately preceding character string. . The connectivity determination unit 133 refers to the connectivity correspondence table 151 and detects that continuity information with the immediately preceding character string “traffic” is stored at the address “npi + l” of the continuity table 152.

  Next, the connectivity determination unit 133 refers to the address “npi + 1” in the continuity table 152, and in the order of “rule (continuity 21)”, “university (16)”, “tool (15)”,. , Detect high continuity. Finally, it is determined that the display order of the above candidate words is in the order of “rule”, “university”, “tool”,.

  Further, when “traffic” is selected after the input of the word “Shanghai” is confirmed, the connectivity determination unit 133 has a probability that it will be input in association with “Shanghai” as the forward character string. A high kanji (word) is searched from the connectivity table 150. The connectivity determination unit 133 refers to the connectivity correspondence table 151 and detects that the relevance information with the forward character string “Shanghai” is stored at the address “ppi + s” of the forward relevance table 153.

  Next, the connectivity determination unit 133 refers to the address “ppi + s” in the forward relevance table 153 and selects “university (relevance 10)”, “rule (5)”, “police (3)”,. In order, it detects that the relevance is high. Finally, it is determined that the display order of the above candidate words is in the order of “university” “rule” “police”.

  As a result, the display order of candidate kanji (words) can be determined as a sequence of words frequently input by the user, so that the candidate kanji / words can be displayed in a display order that reflects user preferences. As a result, the number of key operations by the user can be reduced, and the user's input operation for obtaining a desired kanji can be simplified.

(Determining connectivity-modification)
When the display order is determined using both the element C (continuity with the immediately preceding character string) and the element D (relevance with the preceding character string), the determination result based on the element D may be prioritized. . In this case, in the above example, when “Transport” is selected, the candidate word “Rule” is not always displayed with priority, but “Shanghai” is input ahead. The candidate word “university” can be displayed with priority.

  The operation of the character conversion device 1 when the user inputs “Shanghai Jiaotong University” in the character conversion device 1 where the user frequently inputs “Shanghai Jiaotong University” will be described. When the user inputs the first character “s” of Pinyin to input “up” (shang), “up” is displayed as a candidate kanji in the above-described procedure. When “upper” is selected by the user, words starting with “upper” are displayed as candidate words.

  When “Shanghai” is selected by the user, the word “Shanghai” is displayed on the fixed character display area 123 (FIG. 14). At the same time, the candidate extraction unit 31 of the character conversion device 1 extracts candidate words that are input following the input confirmed character “Shanghai”. The candidate order determining unit 32 determines the display order of the candidate words with reference to the connectivity recording unit 46 of FIG. 22 in order to display the word with a high probability following “Shanghai” at the top.

  First, the connectivity determination unit 133 refers to the address “npi + s” of the continuity table 152 based on the connectivity correspondence table 151 (FIG. 23), and the probability that the word candidate is directly connected to the word “Shanghai” (continuous). Arrange in descending order. That is, in the example shown in FIG. 23, “University (15)” “Transport (13)” “Police (12)” “Rule (8)”. . . Candidates are found in the order of. Subsequently, the connectivity determination unit 133 rearranges the display order of the candidate words following “Shanghai” based on the relevance to the word positioned in front of the input confirmed character “Shanghai”. However, here, since there is no kanji / word located in front of “Shanghai”, the display order is not changed due to the relationship with the forward character string.

  When “traffic” is selected by the user from the candidate words displayed in the above procedure, the character conversion device 1 subsequently displays the candidate words directly continuous to the input confirmed character “traffic” at the top. Determine the display order.

  The connectivity determination unit 133 refers to the address “npi + 1” in the continuity table 152 and determines “rule (21)”, “university (16)”, “tool (15)”, “police (14) based on the continuity. ) ". . . Search for candidates in the following order. Next, referring to the address “ppi + s” in the forward relationship table 153, based on the relationship (relevance) with the kanji / word (here “Shanghai”) located in front of “Transport”, “ Rearrange the display order of the candidate words that follow "Transport". More specifically, in the example shown in FIG. 23, according to the forward relevance table 153, the degree of association with the candidate word “Shanghai” is “University (10)”, “Rule (5)”, "Police (3)", "Tool (2)". . . It has become.

  Therefore, the connectivity determination unit 133 prioritizes the relevance of the candidate word and changes the display order in the order of relevance. In this case, the display order of candidate words following “Shanghai Jiao” is “rule”, “university”, “tool”, “police”. . . From the order of "University", "Rules", "Police", "Tools". . . The order is changed. Alternatively, the display order may be changed based on the sum of the degree of connection and the degree of association. Further, the display order may be changed based on the average value and the maximum value of the degree of connection and the degree of association. For example, an evaluation value calculation rule that prescribes that the total of the connectivity and the relevance is calculated as an evaluation value is stored in the recording unit 40, and the candidate order determination unit 32 follows the evaluation value calculation rule described above. The sum of the degree of connection and the degree of association can be calculated as an evaluation value, and the display order can be determined based on the evaluation value.

  From the above, candidate kanji / words can be displayed in a display order with higher accuracy according to the user's preference, so that the number of key operations by the user can be reduced. As a result, the user's input operation for obtaining a desired kanji can be simplified.

(Display order determination procedure)
In this embodiment, the display order of the extracted candidates is determined by combining the above-described candidate order determination elements. One example will be described below, but the present invention is not limited to the combination described below, and the display order can be determined by appropriately combining the above-described elements.

  FIG. 24 is a diagram illustrating a candidate classification procedure when the candidate order determination unit 32 determines the display order. First, the candidate order determination unit 32 classifies the candidates into several groups for each display order according to predetermined conditions (the elements A to G and combinations thereof), and sets the final display order within the groups. decide.

  The candidate word G extracted by the candidate extraction unit 31 is first stored in the extraction candidate temporary storage unit 51 (FIG. 22) of the temporary storage unit 3. For the stored candidate word G, the learning kanji determination unit 131 (FIG. 22) determines whether or not it is a learning word (S301). Further, the pinyin length determination unit 132 compares the pinyin length (number of letters of the alphabet) input at the time when the candidate word G is extracted with each pinyin length of the candidate word G (S302).

  Here, among the candidate words G, candidate words that are determined to be learning words and have the same pinyin length are classified into the group A having the highest display order. Candidate words that are learning words but are determined not to have the same Pinyin length are classified into group B, which is higher than group A. Candidate words that are not learning words but are determined to have the same Pinyin length are classified into group C, which is lower than group B. Candidate words that are not learning words and are determined not to have the same Pinyin length are classified into the lowest group D.

  For the candidate words that are learning words classified into groups A and B, the learning kanji determination unit 131 further determines whether the learning word is a single learning word or a pair learning word (S303). Further, the connectivity determination unit 133 determines whether or not there is continuity with the immediately preceding character string (S304).

  Here, among the candidate words of group A, the candidate words that are pair learning words and are determined to have continuity with the immediately preceding character string are classified into the highest group A1. Although it is a pair learning word, the candidate word determined to have no continuity with the immediately preceding character string is classified into the upper group A2. Although it is not a pair learning word, the candidate word determined to have continuity with the immediately preceding character string is classified into the lower group A3. Candidate words that are not pair-learned words and are determined not to have continuity with the immediately preceding character string are classified into the lowest group A4. The group B candidate words are also classified in the same manner.

  On the other hand, for candidate words classified into groups C and D, the connectivity determination unit 133 determines the presence or absence of continuity with the immediately preceding character string (S305), and the upper group C (D) 1 and lower And group C (D) 2.

  Finally, all the candidate words (G) are group A1, group A2, group A3, group A4, group B1, group B2, group B3, group B4, group C1, group C2, and group D1 in order from the upper group. , Group D2, and a small group. The order of each small group of groups A and B is further determined by the learning time (element G) (the more recently learned kanji is higher), and each small group of groups C and D is further determined by the appearance frequency (element A) The order is determined (the more commonly used words are higher).

  In the above-described embodiment, it has been described that learned candidate kanji / words are rearranged under a predetermined condition (that is, in the newly learned order). However, learning is performed in the past after candidate kanji extracted as candidates. You may arrange as a candidate word made. Or you may arrange so that the word learned in the past may be displayed with priority over a candidate Chinese character. Moreover, when there are a plurality of words used in the past including candidate Chinese characters, they may be arranged in ascending order of the number of characters. Alternatively, as time course, they may be arranged in the learned order or vice versa.

(Modification 3-Input of four voices)
In the present embodiment, the configuration in which the alphabet (column) and Pinyin are associated with each other while ignoring the distinction between the four voices has been described, but the character conversion device 1 of the present invention is not limited to the above configuration. A configuration is also possible in which alphabets (rows) and Pinyin are made to correspond by distinguishing four voices. For example, if the A / P correspondence table 141 shown in FIG. 15 is created, including a four-voice code (for example, one to four voices are represented by 1 to 4 and a light voice is represented by 5 in PC input). Good. More specifically, in the A / P correspondence table 141 and the P / C correspondence table 142 shown in FIG. 15, it can be seen that Pinyin “shang” is associated with the Chinese character “above”. Here, since the pinyin with four voices “upper” is “three voices shang”, “shang” corresponding to “cp_si” in the A / P correspondence table 141 is described as “shang3”. Pinyin “shang3” with voice can be associated with “above”.

[Embodiment 2]
The character conversion device 1 of the present invention further assigns priority information to the above-mentioned candidate order determination elements, and the conditions (evaluation value calculation rules) when the user determines the display order of candidate kanji (words) are determined by the user. You may comprise so that it can set arbitrarily. Since the user can arbitrarily set the conditions for determining the display order, the multi-function printer provided with the character conversion device 1 can display candidates in the display order that matches the user's preference. As a result, an input operation for obtaining a desired kanji can be simplified.

(Configuration of character conversion device)
The character conversion device 1 according to the present embodiment includes a candidate order determination element setting unit (candidate order determination element setting means) 35 inside the character input control unit 30 in addition to the configuration of the first embodiment described above. A candidate order determination element recording unit 47 is further included inside the recording unit 40 (FIG. 12).

  In this embodiment, the evaluation value calculated for each candidate is temporarily stored in the evaluation value temporary storage unit 52 shown in FIG. 22, and the display order is determined based on the evaluation value. That is, a candidate order determination element having a higher priority set by the candidate order determination element setting unit 35 has a greater influence on the calculation of the evaluation value. The candidate order determination elements set by the candidate order determination element setting unit 35 and their priorities are recorded in the candidate order determination element recording unit 47.

  First, the data structure of the evaluation value for each candidate stored in the evaluation value temporary storage unit 52 will be described.

  FIG. 25 is a diagram illustrating an example of a data structure that stores display order evaluation values calculated for each candidate. In this embodiment, the evaluation value for each candidate is expressed by 16 bits. That is, the value of “0” or “1” indicating the characteristics (elements A to G) of a certain candidate Chinese character (word) is stored in the 16 areas shown in FIG. In the present embodiment, the bit position is related to the level of the evaluation value, and represents the upper digit of the evaluation value as it becomes higher bits. Therefore, the determination result regarding the candidate order determination element having a high priority is reflected in the upper bits. That is, the evaluation value calculation rule defines how to obtain the value of “0” or “1” from the candidate order determination element element of the candidate kanji (word), and the value thus obtained is used as the evaluation value. This specifies the bit position to be reflected.

  According to such a data structure, when the character conversion device 1 calculates an evaluation value for each candidate, it is possible to adjust the importance of an element given to the evaluation value by the bit position. In other words, since important elements are assigned to higher positions, the evaluation value of the candidate display order can be easily quantified easily only by performing the bit ON “1” / OFF “0” operation. . In addition, if a variable is assigned to each element individually, the data size of evaluation values to be secured for determining the display order increases. Therefore, the data size can be reduced by assigning the determination result of each element to 1 bit.

  The candidate order determination element recording unit 47 is for recording priority information of candidate order determination elements for determining a candidate display order. FIGS. 26A and 26B are diagrams illustrating an example of priority information recorded in the candidate order determination element recording unit 47. FIG.

  The priority information includes, for each candidate order determination element, bit position information indicating in which bit position of the evaluation value data structure the determination result related to the candidate order determination element is reflected. In the example shown in FIG. 26A, the determination result of the element B is set to be stored in the area of the bit position 12 shown in FIG. That is, when the Pinyin lengths match, “1” is stored as the evaluation value in the area of the bit position 12 and when they do not match, “0” is stored as the evaluation value. The candidate order determination element to which a higher bit position is assigned has a higher priority and has a larger influence on determining the evaluation value of the candidate.

  In this embodiment, the evaluation value is expressed by 16 bits, and the evaluation value is calculated using the region of bit positions 0 to 12, but the present invention is not limited to this example. When the number of candidate order determination elements is reduced, or when a new element is added or a range width is changed as an evaluation value of each element, it is possible to make the bit length sufficiently large. For example, in the example shown in FIG. 26 (a) and FIG. 26 (b), the appearance frequency is expressed by 8 bits, but 16 bits are secured or the appearance appears to clarify the difference in the appearance frequency. Since the difference in frequency is not important, it can be reduced to 4 bits. Also, empty bit positions 13 to 15 may be used.

  The candidate order determination element setting unit 35 sets priorities of candidate order determination elements designated by the user via the input unit 11 (FIG. 12), and stores priority information in the candidate order determination element recording unit 47. is there.

  More specifically, for example, when the user instructs the priority conversion menu of the candidate order determination element to the character conversion device 1 using the operation panel 2, the display unit 12 of the operation panel 2 gives priority to the candidate order determination element. Displays an operation screen that prompts you to specify the order. For example, a list of candidate order determination elements for which priority can be set may be displayed together with a message such as “Please specify each element (A to G) in order of priority”. When the user indicates each element displayed in the order of priority, a message asking whether to decide is displayed. For example, a message such as “If you decide to change, press the enter key” may be displayed.

  When the user presses the determination key on the operation panel 2, the candidate order determination element setting unit 35 determines the candidate order determination element region having a low priority designated by the user from a low bit position according to the input instruction signal. The bit position information for storing the determination result of each candidate order determination element is changed in order.

  For example, from the state of the example shown in FIG. 26A, the user sets the element learning pair Kanji / word of element F as the candidate order determination element with the highest priority in order of element E, element F, element B, If element C, element D, and element A / G are designated, the priority information is changed as shown in FIG. In the example shown in (b) of FIG. 26, the determination result of the element F is stored in the area of the bit position 12.

  According to the above configuration, each part (FIG. 22) of the candidate order determination unit 32 is designated with the evaluation value related to the determination result based on the priority information of the candidate order determination element specified by the candidate order determination element recording unit 47. The evaluation value temporary storage unit 52 stores the bit position.

  As a result, the evaluation value for determining the display order can be calculated based on the candidate order determining element that the user wants to prioritize, so that the candidates are displayed in a display order that more matches the user's preference. It becomes possible.

  The priority information of the candidate order determination element may be configured to be set separately for candidate kanji and candidate words, or one priority information that is similarly applied to the display order of both candidate kanji and candidate words. It is good also as composition which can be set up. In a configuration in which candidate kanji and candidate words can be set separately, the priority change operation can be performed by changing the priority of the candidate order determination element in the order of candidate kanji and candidate words by one priority setting menu call operation. It may be possible to perform a priority setting menu call operation and a priority change operation for candidate Chinese characters and candidate words.

  Further, a setting end key may be provided, and desired priority information may be changed with respect to contents up to the point when the setting end key is pressed. For example, when the setting end key is pressed in the middle of changing the priority of the candidate order determination element, it is possible not to reflect the change contents of the priority and reflect only the change of the candidate order determination element. is there.

  Next, a method for calculating the evaluation value shown in FIG. 25 based on the determination result of each part of the candidate order determination unit 32 shown in FIG. 22 will be described in more detail. In the example described below, it is assumed that the priority is set as shown in FIG.

  First, the candidates extracted by the candidate extraction unit 31 are temporarily stored in the extraction candidate temporary storage unit 51. In the evaluation value temporary storage unit 52, a 16-bit region (FIG. 25) representing the evaluation value is secured for each stored candidate.

  Next, when the candidate is a learning kanji (word), the learning time determination unit 135 calculates the learning time or the appearance frequency, respectively, otherwise, the evaluation frequency calculation rule. Based on the above, the calculation result is stored in bit positions 0 to 7 (FIG. 25).

  Appearance frequency is the usage frequency obtained statistically for each character and word for various document character data such as a large number of newspapers and magazines in advance, and the higher the frequency, the more commonly used characters and words. . In this embodiment, since the appearance frequency area is 8 bits, the candidate evaluation value can be set between 0 and 255 according to the evaluation value calculation rule based on the appearance frequency.

Similarly, the learning time can be set between 0 and 255 so that the more recently learned, the higher the evaluation value. The learning time represents when the candidate kanji (word) is learned. The evaluation value calculation rule will be specifically described as follows. For example, if the learning time is X, the maximum number that can be learned is N, the learning time of a kanji / word that can be learned is n, and the current learning time is m, the learning time X of any kanji / word can be expressed by the following equation: It is obtained by.
When m−n ≧ 0, X = N + mn−1 (Formula 1-1)
When m−n <0, X = n−m−1 (Formula 1-2)
Here, in the learning period X, it is assumed that learning is performed in the past as the value of X is small.

  When the value range of X in the above formula exceeds the bit length range (8 bits in this embodiment) reserved as the evaluation value for the learning time, the value of X is converted into a numerical value range of bit length, The bit length may be changed.

  Bit positions 8 to 12 store determination results for elements D, C, F, E, and B, respectively. The learning kanji determination unit 131, the pinyin length determination unit 132, and the connectivity determination unit 133 set “1” when it is determined that the characteristics of the candidate satisfy the condition, and “0” otherwise. Set to the specified bit position. For example, the connectivity determination unit 133 sets “1” in the bit position 8 when determining that the candidate is related to the forward character string already input.

  When the display order evaluation values are calculated for all candidates, the candidate order determination unit 32 rearranges the candidates in descending order of evaluation values stored in the evaluation value temporary storage unit 52, and temporarily stores the final display order in the display order temporarily. Stored in the unit 53. Finally, the display control unit 22 (FIG. 12) displays candidates on the display unit 12 based on the display order stored in the display order temporary storage unit 53.

[Embodiment 3]
In each of the above-described embodiments, Chinese input by inputting alphabet keys constituting Pinyin has been described. However, the character conversion device 1 of the present invention is not limited to this. The input unit 11 (FIG. 12) of the operation panel 2 has an input function of handwritten characters, and the input / output control unit 20 can recognize the handwritten characters to configure the character conversion device 1 that inputs Chinese. It is.

  Moreover, the input part (written character input part) 11 for inputting a handwritten character is implement | achieved by displaying the frame for a handwritten character input in the display area of the operation panel 2 as a display integrated tablet, for example. Can do.

  The character conversion device 1 includes a character recognition unit (character recognition means) (not shown) for recognizing writing information (coordinate data) of characters written by the user in the character input frame inside the input / output control unit 20. Prepare. The recording unit 40 further includes a recognition dictionary (not shown) used for matching with the input character pattern recognized by the character recognition unit. And the display control part 22 is comprised so that the recognition result of an input character may be output.

  When the user writes a desired character in the character input frame of the operation panel 2, the operation panel 2 detects the coordinate information 60 (FIG. 27) of the written character and stores it in the internal buffer of the input / output control unit 20. The character recognition unit matches the coordinate information stored in the internal buffer with the dictionary pattern of the recognition dictionary (for example, DP matching), obtains the similarity (for example, distance) between the input pattern and the dictionary pattern, and corresponds to the dictionary pattern The character code to be associated with the similarity is stored in the internal buffer. When the character recognition unit performs matching between the input pattern and each pattern in the recognition dictionary, the character recognition unit rearranges the recognition results in the internal buffer in descending order of similarity. Then, the recognition result is supplied to the candidate extraction unit 31 as a candidate kanji (word).

  Hereinafter, the operation of the character conversion device 1 having a handwritten character recognition function will be described in detail.

  When the user writes “external character 2” in the character input frame of the display-integrated tablet (FIG. 28A), the operation panel 2 detects the coordinate data of “external character 2” and stores it in the internal buffer (FIG. 28). 27). The character recognition unit performs DP matching with the coordinate data of “external character 2” in the internal buffer and each dictionary pattern of the recognition dictionary, calculates distances, and rearranges them in ascending order of distance (FIG. 29). The input / output control unit 20 transfers the recognition result to the display control unit 22.

  The display control unit 22 obtains a display position of n characters (here, kanji characters) that are recognition character candidates based on the recognition result, and displays the display position information and n recognition character candidates n on the display unit of the operation panel 2. 12 (FIG. 28B). The operation panel 2 stores the displayed recognized character candidate and the display position in association with each other in the internal buffer.

  When the user confirms that “external character 2” is displayed, the user designates “external character 2” on the operation panel 2. The operation panel 2 detects the instructed position and transfers the position information to the candidate selection receiving unit 23 of the input / output control unit 20. Then, it is detected that “external character 2” is instructed by the candidate selection receiving unit 23, and a word starting from “external character 2” (“external character 2” heaven ”...) Is displayed ((c) in FIG. )). Since the operation at this time has already been described, it will not be repeated here.

  Thereby, even when the user does not grasp the pinyin corresponding to the desired kanji, it becomes possible to input the desired kanji, and the input operation for obtaining the desired kanji can be simplified.

  In addition, although the function of the said character recognition part of the character conversion apparatus 1 for recognizing a handwritten character was demonstrated taking the example of the multifunction printer, it is not limited to this. Any device provided with a display-integrated touch panel can apply the character recognition unit. Therefore, the character conversion device of the present invention that can input a desired kanji / word by inputting written characters can be realized in devices other than the multifunction printer.

[Embodiment 4]
(Traditional input)
In the above-described embodiment, the character conversion device capable of inputting Chinese (simplified characters) from Pinyin notation has been described. However, the character conversion device 1 of the present invention is not limited to the above configuration.

  For example, the character conversion device 1 of the present invention can also be realized as a character conversion device for inputting traditional characters used mainly in Taiwan.

  In the present embodiment, hereinafter, a character conversion device 1 that accepts input of a phonetic character (a phonetic character) as a phonetic character and presents a traditional character as an ideogram as a candidate kanji (word) to the user will be described. .

  FIG. 30 is a block diagram showing the main configuration of the character conversion device 1 according to this embodiment. 30 correspond to the reference numerals assigned to the constituent elements in FIGS. 12 and 22, and the same reference numerals indicate the same constituent elements. Therefore, the description about the component already demonstrated by each above-mentioned embodiment is not repeated.

  In the character conversion device 1 shown in FIG. 30, the difference from FIGS. 12 and 22 is that the input unit 11 includes a key associated with each sung character for inputting the sung character. The reception unit 21 includes a utterance character input reception unit 24 for receiving the input phonon character and supplying it to the candidate extraction unit 31. In the kanji dictionary unit 41, the phonation character-note sound and the sound injection -A correspondence table that records the correspondence of traditional characters is included.

  FIG. 31 is a diagram illustrating an example of an operation screen at the time of a sung character input process for inputting traditional characters according to the present embodiment. In the display-integrated operation panel 2 in the present embodiment, in addition to displaying the alphabet, a sung character may be displayed (character input area 121). In inputting traditional Chinese characters, it is common to input phonic characters instead of alphabets. As shown in FIG. 31, a sound-injecting character is assigned to a key on the keyboard, so that a sound-injecting sound corresponding to a traditional character can be input.

  FIG. 32 is a diagram showing an example of a Chinese character dictionary (traditional character) recorded in the Chinese character dictionary unit 41 (FIG. 30) in the present embodiment.

  FIG. 33 is a diagram showing an example of a word dictionary recorded in the word dictionary unit 42 in the present embodiment.

  FIG. 34 is a diagram illustrating an example of the connectivity table 150a recorded in the connectivity recording unit 46 in the present embodiment.

  According to the above configuration, as in the case of inputting Chinese (simplified characters), an operation for inputting a phonetic character (a syllabary character) to obtain a desired ideogram (a traditional character used mainly in Taiwan). It is possible to realize a character conversion device that can simplify the above.

[About the learning function]
(Method for determining the display order of learning kanji and words)
The learning kanji / word display order can also be determined by calculating the learned order (referred to as learning order Y), for example, by the following calculation. The learning order Y may be calculated by the learning time determination unit 135 shown in FIG.
Y = learning number−learning counter−1 (Expression 2)
The learning number represents a unique number that should be assigned to a newly learned kanji, and this number is recorded with the learning kanji as a learning counter so that it can be known when the kanji has been learned. It is. If it is possible to record 500 learning kanji (words) in the learning dictionary recording unit 43 (or pair learning dictionary recording unit 44) in FIG. 21, learning numbers 0 to 499 are prepared. do it.

  If Y is negative by the calculation of Equation 2, the upper limit value of learning numbers (here, about 500) is added.

Furthermore, the learning time determination unit 135 may perform a process of rounding the learning rank Y obtained for each candidate in order to save the time and effort of rearranging the candidates. For example, the learning order can be rounded according to the following procedure.
(B) When 0 ≦ Y ≦ 9, Y ← Y (as it is)
(B) When 10 ≦ Y ≦ 29, Y ← (Y−10) / 2 + 10
(C) When 30 ≦ Y ≦ 109, Y ← (Y−30) / 8 + 20
(D) When Y ≧ 110, Y ← (Y−110) / 32 + 30, and when Y> 39, Y ← 39
As described above, by using the above formula 2, it is possible to efficiently rearrange the learning kanji (words) in the order in which they were learned, or in the order in which the learning time was new, so that they are displayed as candidates according to the learning time. The order can be determined.

(Learning counter assignment method)
Next, the learning number (initial value = 0) and the learning counter (initial value = 0) controlled by the learning control unit 34 shown in FIG. 21 will be described more specifically. It is assumed that 500 learning kanji characters can be recorded in the learning dictionary recording unit 43, and the initial value of the recording position (address) of the learning kanji in the learning dictionary recording unit 43 is ID = 0.

  When learning the kanji “gaku” when the learning number = 0, the learning control unit 34 searches the learning kanji learned in the learning dictionary recording unit 43 for learning kanji having a learning counter equal to the current learning number. To do. The kanji with the address ID of the searched learning kanji is deleted, and a new learning kanji is set. If there is no learning kanji having a learning counter equal to the current learning number, both the learning kanji and the learning counter are recorded at the ID position. Then, 1 is added to the learning number. Here, since it is the first learning, the learning kanji character “gaku” is recorded at the position of the address “ID = 0” together with the learning counter = 0, and 1 is added to the learning number, so that the learning number = 1.

  When this operation is repeated and “sentence” is learned when the learning number = 499, the learning control unit 34 sets the learning kanji “sentence” to the learning counter 499 at the address “k” in the learning dictionary recording unit 43. And the learning number is updated to 0 (if the learning number becomes 500 or more as a result of adding 1 to the learning number 499, 500 is subtracted and the learning number is returned to 0).

  Further, when a new kanji character (for example, “Large”) is learned, the learning control unit 34 records the new learned Kanji character “Large” overwriting the “Language” recorded together with the learning counter 0. At this time, “Large” is recorded with the learning counter 0 associated therewith. Subsequently, the learning number is updated to 2 as described above.

(Data structure of learning dictionary)
Next, the data structure of learning words recorded in the learning dictionary recording unit 43 (and the pair learning dictionary recording unit 44) will be described.

  The learning word recorded in the learning dictionary recording unit 43 is assigned a group identifier (GId value) generated from the character codes of kanji characters constituting the word, and the learning word is recorded in association with the GId value. . By using this GId value, the processing efficiency for searching the target learning word from the learning dictionary recording unit 43 is improved.

  A method for generating the GId value will be described with a specific example. In the following description, both the kanji code for uniquely identifying the kanji and the GId value are represented by 16 bits. In this embodiment, as an example, the GId value is composed of the upper 4 bits indicating the number of characters of the word and the lower 12 bits indicating the numerical value obtained by the operation of the kanji code of each kanji constituting the word. However, the data structure of the GId value used in the character conversion device 1 of the present invention is not limited to this.

  For example, assume that the input of the word “Shanghai” is confirmed. At this time, the learning control unit 34 (FIG. 21) first generates a GId value of the word “Shanghai”. In order to generate the GId value, the lower 8 bits of the kanji code of each kanji constituting the word “Shanghai” are expressed by numerical values (“upper” = “009F (hexadecimal number)”, “sea” = “00A3”. (Hexadecimal number) "). For example, if the lower 12 bits of the GId value are obtained by adding the lower 8 bits of the 16 Kanji code, the learning control unit 34 calculates 9F + A3 = “141 (hexadecimal number)” as the lower 12 bits. .

  Next, since “Shanghai” has two characters, “2 (hexadecimal number)” indicating that this word is composed of two characters is calculated as the upper 4 bits. As described above, the GId value of the word “Shanghai” is calculated as “2141 (hexadecimal number)”.

  The GId value generated as described above is a word consisting of any number of characters (however, when the number of characters is expressed by 4 bits as in the above example, up to 8 characters are learned as words) And a fixed length (for example, 16 bits). Therefore, the search processing can be executed after narrowing down the target word by the GId value configured with a fixed length, so that the search efficiency of the learning word can be improved.

[About the operation panel]
The operation screen displayed during the character input process displayed on the operation panel 2 is not limited to the examples shown in FIGS. Another example of the operation screen displayed on the operation panel 2 will be described.

  The alphabet keys for inputting pinyin displayed on the operation panel 2 may be arranged as a key imitating a PC keyboard as shown in FIG. 35, or the input characters are arranged in alphabetical order as shown in FIG. It may be applied to an m × n key arrangement.

  Also, a candidate character display area 122 for displaying candidate kanji characters may be displayed superimposed on the confirmed character display area 123 ((a) to (c) of FIG. 37).

  Further, after selecting a candidate kanji, a second candidate character display area 122 ′ for displaying a candidate word starting from the selected kanji may be displayed superimposed on the candidate character display area 122 and the confirmed character display area 123. Alternatively, it may be displayed side by side with the candidate character display area 122 (FIG. 38B).

[Embodiment 5]
An embodiment of the present invention will be described below with reference to FIGS. Here, although it is set as the apparatus which inputs the reading (phonetic character) of a Chinese character (converted character, the character after conversion), and determines the word containing a Chinese character, it is not limited to this. That is, the embodiment described below is an example for carrying out the present invention, and is not limited in realizing the present invention.

  FIG. 1 is a block diagram showing a configuration of a character conversion device 1 according to the present embodiment. The character conversion device 1 includes an input unit 11, a dictionary search unit 310, a display control unit 22, a kanji dictionary unit (converted character dictionary) 41, a word dictionary unit (word dictionary) 42, a temporary storage unit 3, a control unit 10, and a display unit. 12 is provided. 1 correspond to the reference numerals given to the constituent elements in FIGS. 12, 16, 21, 21, 22 and 30, and the same reference numerals denote the same constituent elements. Indicates an element. 1 may be provided with each part of FIG. 12, FIG. 16, FIG. 21, FIG. 22, and FIG. 30 that are not shown in FIG.

  The input unit 11 includes a keyboard or the like, and is configured so that the user can input a desired input character string (phonetic character), a selected kanji character, a selected word, and the like described later. . The input unit 11 is configured such that the user can input various instructions to the character conversion device 1. Note that the input unit 11 and the display unit 12 may also be used as the operation panel 2 (touch panel).

  Based on the readings and readings input from the input unit 11, the display control unit 22 displays the kanji and words searched by the dictionary search unit 310 on the display unit 12 so that the user can select them. The display unit 12 displays various data under the control of the display control unit 22.

  In this embodiment, the candidate extraction unit 31 includes a dictionary search unit 310. The dictionary search unit 310 searches each dictionary unit (the kanji dictionary unit 41 and the word dictionary unit 42) stored in the recording unit 40, and detects candidate kanji (or words). In the present embodiment, the dictionary search unit 310 further includes a kanji acquisition unit (first conversion character acquisition unit) 311 and a word acquisition unit (word acquisition unit) 312.

  The kanji acquisition unit 311 receives the character string of the reading received by the input receiving unit (accepting means) 21 and searches the kanji dictionary unit 41 to match the desired condition (or kanji information described below). Is stored in the temporary storage unit 3 as a search result. Further, the word acquisition unit 312 uses the search result of the kanji acquisition unit 311 and the character string of the reading received by the input reception unit 21 as input, and searches the word dictionary unit 42 to search for a word (or below) that matches a desired condition. Word information to be described) is stored in the temporary storage unit 3 as a search result. Details of kanji acquisition and word acquisition will be described later.

  The kanji dictionary unit 41 holds kanji correspondence information 144 that associates readings with kanji. Furthermore, a kanji index (kanji index table) 145 is stored in which the first character of the reading is associated with the address information of the storage destination of the kanji corresponding to the first character in the kanji correspondence information 144. Here, it is assumed that the kanji index 145 is arranged in the order of reading kanji.

  The word dictionary unit 42 holds word correspondence information 162 in which a word and a reading of the word are associated with each other. Furthermore, a word index (word index table) 161 in which the first character that can be the head of the word and the address information of the storage destination in the word correspondence information 162 of the word having the first character at the head is stored. Yes. The word correspondence information 162 and the word index 161 are grouped for each leading character of the word.

  The kanji correspondence information 144 stores auxiliary information in association with kanji. The word correspondence information 162 stores auxiliary information in association with words. Examples of the auxiliary information include the number of reading characters (for example, the Pinyin length (element B) in each of the above-described embodiments) (search auxiliary information), the appearance frequency (the above-described element A) (display auxiliary information), and the like. . The auxiliary information is a numerical value obtained by converting the number of characters read and the appearance frequency into a predetermined range, for example, a range of 1 to 100. In the following, kanji information refers to kanji and information (reading, auxiliary information) associated with the kanji. The word information refers to a word and information (reading and auxiliary information) associated with the word. However, the auxiliary information is of course not limited to the above, and may be other information. For example, the candidate order determination elements (element C to element G) described in the above embodiments may be stored as auxiliary information in association with candidates (kanji / words).

  The control unit 10 comprehensively controls operations of various components in the character conversion device 1. For example, the number of characters in the character string input from the input unit 11 is counted. Further, the dictionary search unit 310 and the display control unit 22 are controlled. Moreover, the input reception part 21 is provided and the character string which the user input from the input part 11 is received.

  Next, the processing operation of the character conversion apparatus 1 will be described using the flowchart shown in FIG.

  When the user inputs one or more reading character strings including the first character of the reading as the reading of the first character of the character string that the user wants to input from the input unit 11, the input receiving unit 21 inputs the input reading character string. Transfer to the control unit 10.

  The control unit 10 counts the number of characters in the reading character string transferred from the input unit 11 and stores the reading character string and the number of characters in the temporary storage unit 3. Then, the control unit 10 instructs the dictionary search unit 310 to search for Chinese characters (S101).

  Next, the kanji acquisition unit 311 of the dictionary search unit 310 reads the read character string and the number of characters of the read character string from the temporary storage unit 3 into an internal buffer (not shown) of the dictionary search unit 310. The kanji acquisition unit 311 refers to the kanji index 145 and obtains the address (iNum) of the end position of the search target range to be searched in the kanji correspondence information 144 based on the first character of the reading character string in the internal buffer. Hold (S102). Further, the address (iStart) of the start position of the search target range to be searched in the kanji correspondence information 144 is obtained and held in the internal buffer (S103).

  The kanji acquisition unit 311 reads kanji information by referring to the kanji correspondence information 144 based on the address held in the internal buffer of the dictionary search unit 310, and reads the kanji correspondence information 144 and the input character string. Are compared (S104). When the reading of the input character string and the reading in the kanji correspondence information 144 match (YES in S104), the kanji acquisition unit 311 stores the kanji information that satisfies the matching condition in the temporary storage unit 3. (S105).

  Then, iStart is updated (S106). Thereafter, the kanji acquisition unit 311 compares the iNum held in the internal buffer with the iStart updated in S106 (S107). As a result of the comparison, if iStart is equal to or greater than iNum for the kanji to be searched (NO in S107), the control unit 10 is informed that the kanji search has ended.

  A specific description of Kanji search will be described later using an embodiment.

  Upon receiving a notification from the dictionary search unit 310 that the kanji search has been completed, the control unit 10 rearranges the kanji search results stored in the temporary storage unit 3 in a predetermined order (S108) and adds display position information. And stored in the temporary storage unit 3. In addition, the display control unit 22 displays the kanji search result on the display unit 12 (S109). In addition, when rearranging in the predetermined order, it may be rearranged in the order of appearance frequency of auxiliary information, and as described in each of the above embodiments, candidates are determined based on the candidate order determining element (and the evaluation value calculation rule). The order determination unit 32 may calculate evaluation values and rearrange them based on the evaluation values.

  Next, it is confirmed whether there is a user instruction input (S110). The user visually checks the kanji search result displayed on the display unit 12, selects a desired kanji, and uses the input unit 11 to instruct and input the position where the kanji is displayed. When there is such a user instruction input (YES in S110), the input unit 11 transmits the position information of the indicated position to the input receiving unit 21 of the control unit 10.

  The control unit 10 compares the position information received by the input receiving unit 21 with the information on the display position of the kanji search result stored in the temporary storage unit 3, and detects the input kanji (S111). Further, the control unit 10 transfers the kanji specified by the user to the dictionary search unit 310 and instructs to perform a word search. At this time, the candidate selection receiving unit 23 detects the kanji selected by the user, and the learning control means stores it in the learning dictionary recording unit 43 or the pair learning dictionary recording unit 44 as the learning kanji. Good.

  Upon receiving a word search instruction, the word acquisition unit 312 of the dictionary search unit 310 refers to the word index 161 and should search in the word correspondence information 162 based on the acquired kanji identification information (character code). The address (iNum) of the end position of the search target range is obtained and held in the internal buffer (S112). Further, the address (iStart) of the start position of the search target range to be searched in the word correspondence information 162 is obtained and held in the internal buffer (S113).

  Then, the word acquisition unit 312 reads the word information with reference to the word correspondence information 162 based on the address held in the internal buffer of the dictionary search unit 310, and stores the word information matching the predetermined condition in the temporary storage unit 3. Store (S114).

  Thereafter, the word acquisition unit 312 updates the iStart for the search target word (S114a), and compares the iNum held in the internal buffer with the iStart updated in S114a (S115). As a result of the comparison, if iStart is equal to or greater than iNum for the search target word (NO in S115), the control unit 10 is informed that the word search has ended.

  A specific description of word search will be described later using an embodiment.

  Then, upon receiving notification from the dictionary search unit 310 that the word search has been completed, the control unit 10 rearranges the word search results stored in the temporary storage unit 3 in a predetermined order (S116), and displays the display position information. In addition, it is stored in the temporary storage unit 3. In addition, the display control unit 22 causes the display unit 12 to display the word search result on the display unit 12 (S117). In addition, when rearranging in the predetermined order, it may be rearranged in the order of appearance frequency of auxiliary information, and as described in each of the above embodiments, candidates are determined based on the candidate order determining element (and the evaluation value calculation rule). The order determination unit 32 may calculate evaluation values and rearrange them based on the evaluation values.

  Further, the user visually looks at a desired word from the word search result displayed on the display unit 12, selects the desired word, and inputs the position where the word is displayed. When there is an instruction input, the input unit 11 transmits the position information of the input position to the input receiving unit 21 of the control unit 10. At this time, the candidate selection receiving unit 23 detects the kanji selected by the user, and the learning control means stores it in the learning dictionary recording unit 43 or the pair learning dictionary recording unit 44 as the learning kanji. Good.

  The control unit 10 compares the position information received by the input receiving unit 21 with the information on the display position of the word search result stored in the temporary storage unit 3 and searches for the word specified by the user. Then, the display control unit 22 displays the word searched by the display unit 12.

  That is, in the character conversion device according to the present invention, the kanji dictionary unit 41 and the word dictionary unit 42 have the index, the kanji correspondence information 144, and the word correspondence information 162, group words of the same first kanji, By adding auxiliary information to the word correspondence information and limiting kanji and words to be searched based on the auxiliary information, the search time can be shortened. Furthermore, the display order of the kanji (words) that are searched candidates is higher the possibility that the candidate order determining unit 32 will select the kanji (words) by the user based on the candidate auxiliary information (candidate order determining element). It is determined to be displayed with priority. Therefore, candidate Chinese characters (words) rearranged in the order desired by the user can be displayed in a shorter time. As a result, user convenience can be improved.

  In Examples 5-1 and 5-2 below, the character conversion device 1 described above will be described as a Chinese conversion device that inputs Pinyin (phonetic characters) and converts them into kanji (converted characters). In this case, as described in detail below, it is assumed that the kanji index 145, the kanji correspondence information 144, the word index 161, and the word correspondence information 162 are as follows.

  The Chinese character index 145 is as shown in FIG. 3, and the Chinese character of the Chinese character having the first character of Pinyin (first phonetic character) and the first character of Pinyin as the first character of Pinyin (phonetic information) of the Chinese character. The correspondence information 144 is associated with the kanji address (storage destination information). In FIG. 3, the description (cidx_m1, cidx_m2, etc.) on the left of the kanji index 145 is an address in the kanji index 145. Specifically, cidx_m1 is an address in the kanji index 145 of “the reading starts with m and one voice”.

  In the kanji correspondence information 144, as shown in FIG. 4, a kanji, its pinyin, and auxiliary information (in FIG. 4, search conditions and frequency information) are associated with each other. The description (cp_m1, cp_m2, etc.) on the left side of the kanji correspondence information 144 in FIG. 4 is an address in the kanji correspondence information 144.

  Further, the word index 161 has a primary index (word primary index table) 161a shown in FIG. 5 and a secondary index (word secondary index table) 161b shown in FIG. As shown in FIG. 5, in the primary index 161a, the kanji, the code thereof, and the address in the secondary index 161b of the word having the kanji as the head are associated with each other. In the secondary index 161b, the leading Pinyin of the Chinese character and the second Chinese character is associated with the address in the word correspondence information 162. The description (widx_di, widx_gi, etc.) on the left side of the secondary index 161b in FIG. 6 is an address in the secondary index 161b.

  Furthermore, as shown in FIG. 7, the word correspondence information 162 associates words with their pinyin. The description (wp_di, wp_gi, etc.) on the left of the word correspondence information 162 in FIG. 7 is an address in the word correspondence information 162. In the kanji correspondence information 144, when the kanji code corresponds to Pinyin, the primary index 161a is not necessarily required.

(Example 5-1)
In the present embodiment, the above-described character conversion device 1 is described as a Chinese conversion device that inputs Pinyin and converts it into Kanji. However, the present invention is not limited to this. For example, Hiragana (or Katakana) is input to input Kanji. An apparatus for determining a word to be included may be used.

  Specifically, a case where the user inputs “Mei” of Pinyin, which is a Chinese reading of “External character 2”, in order to input “External character 1” will be described.

  First, the kanji search process will be described. Note that the condition that the input reading matches the reading of the kanji information is that the input reading matches the reading of the kanji information forward.

(Kanji search)
The kanji acquisition unit 311 of the dictionary search unit 310 reads the read character string “mei” from the temporary storage unit 3 into the internal buffer of the dictionary search unit 310. Then, the dictionary search unit 310 obtains addresses in the range of “m” of the kanji index 145 based on the first character “m” of the reading character, and stores the kanji address described in those addresses in the internal buffer. Remember.

The address is
(1) When there is a four-voice code in the input reading Start address = (Character code of input reading-Character code of base reading)
× Kanji index unit size × 5
+ (Input reading four-voice code-1) x Kanji index unit size end address = (input reading character code-base reading character code)
× Kanji index unit size × 5
+ Input reading four-voice code x Kanji index unit size (2) When there is no four-voice code in the input reading Start address = (Character code of input reading-Character code of base reading)
× Kanji index unit size × 5
End address = (Character code of input reading-Character code of base reading + 1)
× Kanji index unit size × 5
As determined.

  The reason why the address is determined by “5” is as follows. Chinese pronunciation is a combination of a part that represents reading, such as Pinyin and phoning, and a tone (four voices) that represents the ups and downs (tone) of the utterance. Since there are five types of tone (1 to 4 voices and light voices), it is set to 5.

  In Chinese input by Pinyin, Pinyin is expressed in alphabet, so the character code of the base reading is “a” (= 0x61 (ASCII code)). "

(= 0xa373) ".

  The unit size of the kanji index (kanji index table) 41 is configured so that the word index (word index table) 51 is in alphabetical order of the first characters of Pinyin, and the first characters of the same Pinyin are arranged in the order of 1 to 5 of the four voices. If so, it is the byte length of the Kanji address. For example, a device operating as the present embodiment has 4 bytes.

  Here, since there is no four-voice code, the address in the range of “m (= 0x6d)” is represented by the formula (2), and the head address is “(0x6d-0x61) × unit size of kanji index × 5” (that is, FIG. 3), and the end address is calculated by “(0x6d-0x61 + 1) × unit size of kanji index × 5” (that is, cidx_n1 in FIG. 3), and the kanji address (cp_m1) described in the kanji index indicated by each address , And cp_n1) are stored in the internal buffer. CIDX_M1 in FIG. 3 represents an address in the Kanji index 145 of the first reading of m and one voice. The corresponding cp_m1 represents the address of the kanji correspondence information 144 for the reading of one voice with m as the beginning of reading. In cidx_m5 and cp_m5, the beginning of reading is m and corresponds to a light voice.

  Note that (0x6d-0x61) = 12, and (0x6d-0x61 + 1) = 13.

  The dictionary search unit 310 sets the search target address cpi to cp_m1. Next, the dictionary search unit 310 reads the kanji information at the address cpi from the kanji correspondence information 144 and determines the search condition based on the number of input pinyins. As a result, if the search condition is satisfied, the input reading is compared with the reading of the kanji information. If the input reading matches the reading of the kanji information, the kanji information is stored in the temporary storage unit 3 as a match. And only cpi is updated.

In the example of the kanji correspondence information 144 shown in FIG. 4, the search conditions are as follows.
The search condition is
0: Unconditionally search target 1: If the number of input pinyins is 1 character, it is not searched 2: If the number of input pinyins is 2 characters or less, it is not searched 3: The number of input pinyins is 3 characters In the following cases, it is not considered as a search target.

  When the input pinyin is “mei”, since the number of input pinyins is 3, the Chinese characters corresponding to the search conditions 0, 1, and 2 are targeted for the search process.

  The above process continues until cpi is equal to cp_n1, and if cpi is equal to cp_n1, the kanji information stored in temporary storage unit 3 is rearranged at the appearance frequency as one of the auxiliary information, and kanji The search process ends.

  In the above description, the case where a four-voice code is not input has been described. However, when “mei3” is input with a four-voice code, the address of the third voice of “m” is the above (1). The start address is “(0x6d-0x61) × unit size of kanji index × 5 + (3-1) × unit size of kanji index” (that is, cidx_m3 in FIG. 3), and the end address is “(0x6d− 0x61) × unit size of kanji index × 5 + 3 × unit size of kanji index ”(that is, cidx_m4 in FIG. 3), and kanji addresses (cp_m3 and cp_m4) described in the kanji index indicated by each address are stored in the internal buffer. Remember.

  Note that the processing after obtaining the kanji address is the same as the search processing described above, and therefore the description thereof is omitted.

(Word search)
The word search is explained below. Specifically, the user inputs Pinyin's “mei”, which is a Chinese reading of “External character 1”, and the Chinese character search result is held in the temporary storage unit 3 and displayed on the display control unit 22 display device. The description will be made under the condition that “external character 2” is input.

  As shown in FIG. 5, the primary index 161a of the word dictionary unit 42 is stored in ascending order of kanji codes.

  The control unit 10 displays the user's designated position (X1, Y1) and the display position (xi_min, yi_min, xi_max, yi_max) of the kanji search result stored in the temporary storage unit 3 (i = 1 to n, n is the number of candidates) ) And xi_min ≦ X1 ≦ xi_max and yi_min ≦ Y1 ≦ yi_max, it is determined that the i-th candidate is selected, the corresponding kanji is transferred to the dictionary search unit 310, and the word search is performed. Instruct to do.

  Upon receiving a word search instruction, the dictionary search unit 310 searches the word index 161 that matches the transferred kanji code, and obtains a search target range in the word correspondence information 162. Specifically, the dictionary search unit 310 searches the primary index 161a in half for the transferred “external character 2”, and acquires a Chinese character code C3BF that matches “external character 2”. Then, using the address “widx_mi” in the secondary index 161b corresponding to the Chinese character code C3BF, the secondary index 161b is referred to, and “external character 2” is the address (word address) “in the word correspondence information 162 of the first word. “wp_mi” is stored in the internal buffer as the start address of the word correspondence information 162 having “gaiji2” as the head. Further, the dictionary search unit 310 uses the K3 code C3C0 (in FIG. 5) next to C3BF in the primary index 161a.

The secondary index 161b is referred to using the secondary index address “widx_mi + 1” (hereinafter referred to as “external character 6”), and “external character 6” is the address (word address) in the word correspondence information 162 of the first word. ) “Wp_mi + 1” is stored in the internal buffer as the end address in the word correspondence information 162 with “External character 2” first.

  Thereafter, the word acquisition unit 312 of the dictionary search unit 310 reads word information from “wp_mi” to “wp_mi + 1”, and stores the kanji string of the word information in the temporary storage unit 3 as a search result. And the candidate order determination part 32 rearranges in order with the appearance frequency which is auxiliary information (not shown) of the word information (candidate word) of the temporary storage part 3, and complete | finishes a word search process. Note that this rearrangement is merely an example, and the rearrangement may be performed by other methods. For example, the display order may be determined based on the candidate order determination elements (element A to element G). Finally, as described above, the searched and rearranged words are displayed on the display unit 12.

(Example 5-2)
In Example 5-2, an operation when the user inputs a word reading to input a word will be described. Here, in Example 5-1, a reading of the first character of a character string to be input is input, whereas in Example 5-2, a reading of a word to be input is input. However, the process in which the input receiving unit 21 of the control unit 10 receives the reading character string input from the input unit 11 is the same as that of the embodiment 5-1. Therefore, in this embodiment, a process after the user inputs “Meinian” of Pinyin, which is a Chinese reading of “External character 1”, in order to input “External character 1” will be described.

  It is assumed that the control unit 10 has a function of dividing the pinyin of the input reading character string. The pinyin division process of the input reading character string will be described later.

  When the control unit 10 receives the “mainian” reading character string from the input unit 11, the control unit 10 divides the reading character string into pinyin. As a result of the division, two divided into “pinyin” by “mei” and “nian” are obtained. The number of divisions indicating that there are two (or the number of division positions indicating that there are two division positions) 2 is stored in the temporary storage unit 3.

  The kanji acquisition unit 311 of the dictionary search unit 310 reads the read character string “mei” from the temporary storage unit 3 and stores it in the internal buffer of the dictionary search unit 310. The kanji acquisition unit 311 obtains addresses in the range of “m” by referring to the kanji index 145 using the above equation (2) based on the first character “m” of “mei”, and describes these addresses. Stores the kanji address being stored in the internal buffer.

  Here, since there is no four-voice code, the start address is “(0x6d-0x61) × unit size of kanji index × 5” (that is, cidx_m1 in FIG. 3), and the end address is “(0x6d-0x61 + 1) × unit of kanji index. The kanji address (cp_m1 and cp_n1) indicated by each address is stored in the internal buffer by referring to the kanji index 145 of the kanji dictionary unit 41, calculated by “size × 5” (that is, cidx_n1 in FIG. 3).

  The kanji acquisition unit 311 of the dictionary search unit 310 sets the search target address cpi to cp_m1. Next, the Chinese character acquisition unit 311 reads the Chinese character information at the address cpi, refers to the Chinese character correspondence information 144, and determines the search condition based on the number of input pinyins. As a result, if the search condition is satisfied, the input reading is compared with the reading of kanji information.

  If the input reading is completely coincident with the reading of the kanji information, the temporary storage unit 3 stores the fact that it is completely coincident with the kanji information (for example, 1), and updates cpi.

  If the input reading matches the reading of the kanji information, the temporary storage unit 3 stores that the kanji information and the front matching (for example, 0) in association with each other, and updates cpi.

  If it is not an exact match or a forward match, only cpi is updated.

  If cpi is equal to cp_n1, the candidate order determination unit 32 rearranges the completely matched kanji information among the kanji information stored in the temporary storage unit 3 according to the appearance frequency, and further, the kanji that matches the front. The information is rearranged at the appearance frequency so that the candidate rank is lower than the completely matched kanji information, and the kanji search process is terminated. The candidate order determination unit 32 may further determine the display order based on other auxiliary information (for example, candidate order determination elements (element A to element G)).

  In the above processing, for the reading character string “mei”, the kanji information of “external character 2” is stored in the temporary storage unit 3 as the search result as kanji information completely matching “mei”.

  Next, the word search process will be described.

  A description will be made assuming that the user has designated “external character 2” from the kanji search candidates displayed after the kanji search. Note that the candidate selection process for the instruction “external character 2” by the user is the same as the process described in Example 5-1, and thus the description thereof is omitted.

  In addition, the primary index 161a, the secondary index 161b, and the word correspondence information 162 are arranged in ascending order of kanji codes.

  When receiving a word search instruction, the word acquisition unit 312 of the dictionary search unit 310 refers to the primary index of the word index 161 and acquires first index information that matches the transferred kanji code. For example, in the primary index, the words are arranged in the order of the kanji code of the first kanji of the word, and the word acquisition unit 312 searches for the first index information that matches the kanji code transferred by the binary search and is transferred. The first index information that matches the kanji code and the first index information of the next kanji code arranged in the primary index are stored in the internal buffer.

  Here, since the user designates “external character 2”, in the primary index 161a, the secondary index address matching the kanji code is widx_mi, and the secondary index address of the next kanji code is widx_mi + 1. Stored in internal buffer.

  The word acquisition unit 312 of the dictionary search unit 310 reads the second and subsequent character strings of the read character string divided from the temporary storage unit 3 into the internal buffer. The dictionary search unit 310 narrows the search range of the word correspondence information 162 by using the kanji (first kanji, first kanji) searched by the first divided reading and the first pinyin character of the second divided reading. Therefore, based on the address of the secondary index 161b added to the primary index 161a, the secondary index 161b is searched for a match with the first character “n” of the second divided reading, The secondary index 161b searched for in the internal buffer and the next second index information arranged in the secondary index 161b are stored. Here, secondary index information wp_mi that matches the first Kanji character “gaiji 2” and the first character “n” of the second divided reading, and the next secondary index arranged in the secondary index Information wp_mi + k is stored in the internal buffer.

  The word acquisition unit 312 of the dictionary search unit 310 sets the search target address wpi as the address wp_mi of the word information added to the second index information. The word information from “wp_mi” to “wp_mi + k” is read out. Here, the reading of the word information in the word correspondence information 162 is associated with each kanji of the word, for example, the reading of “External character 1” is changed to “MeiNian”. Thus, it is assumed that the first part of the reading corresponding to each kanji of the word is capitalized.

  Next, the word acquisition unit 312 of the dictionary search unit 310 reads the word information at the address wpi, refers to the word correspondence information 162, and compares the number of Chinese characters in the word correspondence information 162 with the number of input reading divisions. If the number of kanji in the word information is less than or equal to the number of divisions in the input reading, the reading corresponding to each kanji in the word information and the divided readings are compared from the top, and the number of kanji with the exact match and the number of kanji with the reading forward match Count.

  If the “number of kanji characters in word information” and “number of kanji characters whose readings are completely matched + the number of kanji characters whose readings are forward matching” match, the number of kanji characters whose readings are completely matched with the word information in the temporary storage unit 3 and the readings are forwarded. Store the number of matching kanji and update wpi. If “the number of kanji characters in word information” and “the number of kanji characters whose reading is completely matched + the number of kanji characters whose reading matches the front” do not match, wpi is updated.

  If wpi <wp_mi + k, the process is continued. If wpi = wp_mi + k, the process ends.

The dictionary search unit 310 ranks the word information stored in the temporary storage unit 3 under the following conditions.
(I) Word information in which the number of reading divisions is equal to “the number of Kanji characters whose readings are completely matched + the number of Kanji characters whose readings are the same as the preceding characters” (Ii) Word information where the number of reading divisions is not equal to the number of Kanji characters whose readings are exact matches + the number of Kanji characters whose readings are front-matched. The higher the number of “Kanji with perfect matching”, the higher the candidate. ・ If the “Number of Kanji with perfect reading” is the same, rank them by frequency of appearance. However, if a search result corresponding to the above condition (i) is selected, a new reading character string is input, but a search corresponding to the condition (ii) is performed. If you select a result, it will be part of the entered reading string. To respond, it is possible to repeat the above processing read character string except the read character string corresponding to the selected character string as input to read in the control unit 10.

  In the above description, the word search result is obtained only when the number of kanji in the word information matches “the number of kanji whose reading is completely matched + the number of kanji whose reading matches the front” in the word search. However, the word search result may also be used when the number of kanji in the word information is larger than “the number of kanji whose readings are completely matched + the number of kanji whose readings are forward matching”. With such a configuration, it is possible to input by reading only a few characters from the beginning of a desired word.

  In the above description, the word search is started by the user selecting a desired kanji from the displayed kanji search result. However, after completing the kanji search, the word search may be started for the kanji ranked first in the candidate ranking without displaying the kanji search result, and the search result may be displayed when the word search is completed. Alternatively, after completing the kanji search, the word search may be started for some or all of the candidates without displaying the kanji search result, and the search result may be displayed when the word search is completed.

  Next, an input reading pinyin (n character) dividing process will be described. Here, the unit of division is called a syllable.

The division process is performed as follows.
(A) The division position i is set to 1. (b) The i-th character from the first character is set as the first syllable, and the longest pinyin that is established as the final, final, and initial + final from the i + 1th character is obtained, and the second syllable is obtained. (C) The first syllable length + the second syllable length is held as the evaluation value (hereinafter referred to as the position evaluation value) of the i-th character division position. (D) 1 is added to the division position i (i = i + 1). ), If i> n (e), otherwise move to (b) (e) determine the final division position based on the number of division positions processed as having the maximum position evaluation value (E-1) When the number of division positions processed as having the maximum position evaluation value is 1 The division position processed as having the maximum position evaluation value is set as the final division position (e-2 ) Multiple split positions processed as having the maximum position evaluation value (1) It consists of only one syllable (2) Two or more characters in both the first and second syllables (3) The first syllable has more characters than the second syllable As a specific example, the division process of “mainian” is described as the final division position. Whether or not part or all of the input pinyin is established as a pinyin is compared with the pinyin registered in the pinyin dictionary, and if there is a matching pinyin, there is a possibility that it is established as a pinyin. In the pinyin dictionary, as shown in FIG. 9, pinyins having a length of 1 to 7 that may be established as pinyin are registered.

Here, the input pinyin is a character string input as pinyin,
(1) Character or character string registered as Pinyin itself (2) Character string or character string registered as Pinyin (3) Character string other than (1) (2) If)
There are three types.

  In the fifth embodiment, the position evaluation value calculated at the time of the division process is obtained by quantifying the likelihood that the syllable is divided at that position when the separation position for dividing the input Pinyin into two syllables is determined. It is. This position evaluation value is calculated for each syllable break position. Therefore, the evaluation value (evaluation value indicating the possibility that the candidate kanji / candidate word is selected by the user) calculated for each candidate kanji / candidate word used in the first to fourth embodiments described above. Differentiated.

  The division position i of the input pinyin “mainian” is set to 1. With reference to the Pinyin dictionary index (pinyin dictionary index table) shown in FIG. 8 using the first character “m”, the first character “m” and the one character Pinyin dictionary address pr_m1 and the two character Pinyin dictionary address pr_m2 are read. . And it compares with the pinyin memorize | stored from pr_m1 to pr_m2 of a pinyin dictionary. In this case, since “m” exists, it is established as pinyin.

  Next, “einian” excluding “m” is set as the second syllable, the first character is “e”, the six-character Pinyin dictionary address pr_e6 and the seven-character Pinyin dictionary address pr_e7 are read and stored from pr_e6 to pr_e7. Compare with Pinyin. If there is a matching pinyin, the position evaluation value of the first character is 7. However, if there is no matching pinyin, one character is reduced from the last character of the second syllable to form a new second syllable, and similarly the Pinyin dictionary and If there is a matching Pinyin, the length of the second syllable + the length of the first syllable (here, 1) is set as the position evaluation value of the division position 1. If there is no matching pinyin, a process of further reducing one character from the last character of the second syllable and comparing it with the pinyin dictionary is repeated until a position evaluation value is obtained.

  When the above process is completed, the division position i of the input pinyin “mainian” is set to i = i + 1 (= 2). Therefore, based on the two characters “me” from the first character, “m” from the Pinyin dictionary index table is compared with the Pinyin stored from the Pinyin dictionary addresses pr_m2 to pr_m3 of the first two characters. If “me” exists, the first syllable is set to “me”. Next, “inian” excluding “me” is set as the second syllable, and similarly, it is searched whether there is a pinyin that matches the pinyin dictionary. Processing from the beginning to the cutout position 7 is performed, and the division position of the maximum position evaluation value is evaluated. In FIG. 10, since the maximum position evaluation value is 7, and the number of division positions at which the maximum position evaluation value is 7 is 1, the final division position is determined to be 3. In FIG. 10, “division position number 2” for a division position of 3 means that “mainian” has a first division position between the third character i and the fourth character n, Since it is determined here that there is a second division position between the seventh character n and the following (no character), it indicates that there are two division positions.

  Next, the same processing as described above is repeated for “nian” excluding pinyin up to the dividing position, and the dividing position of the input pinyin is obtained. In “nian”, as shown in FIG. 11, the division position is obtained, and there are two division positions of the position evaluation value 4. Therefore, “nian” composed of one syllable is selected, and the division position 4 is obtained.

  If the pinyin up to the division position is excluded, there is no remaining pinyin, so the input pinyin is corrected to the position from the head position (division positions 3 and 7), and the process ends.

  In addition, although the division | segmentation process with respect to Pinyin was demonstrated above, it can divide | segment by the said process by comprising a sound-injection table similarly about sound-injection. The index table of the Pinyin dictionary has a maximum Pinyin length of 7, so the first character of each Pinyin is composed of 7 units of length 1-7. On the other hand, since the maximum length of the sound injection is 4, the first character of each sound injection may be configured in units of 4 from length 1 to 4.

  As described above, in the present invention, words having the same leading kanji are grouped, and the kanji dictionary and the word dictionary have information for limiting the search target according to the input pinyin. Specifically, the Chinese character dictionary has information on whether or not it is a search target corresponding to the length of Pinyin. The word dictionary has the first character of the reading of the second character in the index information, divides the Pinyin sequence (n characters of kanji), and the word from the first character and the second character of the first character searched Search target can be limited. Because of this structure, search for items that match the search conditions, so even if the number of vocabulary is increased, the number of comparisons between the input character string and reading information is limited, and character string input with a fast response speed An apparatus can be provided. When the reading of the second and subsequent characters of the kanji to be input is input, the word search target can be further narrowed down, so that a character conversion device with a fast search process can be provided.

  Furthermore, candidates for kanji (words) obtained by the above-described search process are more likely to be selected by the user based on auxiliary information (or candidate order determination elements) associated with each candidate. Since the images are rearranged in the order in which they are displayed with priority, the response time until the candidates are displayed can be reduced, and the candidates can be displayed in the order desired by the user. Can be further improved.

  Finally, each block of the character conversion device 1, particularly the candidate order determination unit 32, the dictionary search unit 310, the display control unit 22, and the control unit 10 may be configured by hardware logic, or as described below. It may be realized by software using Specifically, the recording unit 40, the kanji dictionary unit 41, the word dictionary unit 42, and the temporary storage unit 3 are realized by a storage device such as a memory or a hard disk, and the candidate order determination unit 32, dictionary search unit 310, display control unit 22. The control unit 10 is realized by a central processing unit (CPU) of a computer and software executed on the CPU. This software is also stored in the storage device, read into the memory, and executed by the CPU.

  That is, the character conversion device 1 includes a CPU (central processing unit) that executes instructions of a control program that realizes each function, a ROM (read only memory) that stores the program, and a RAM (random access memory) that expands the program. And a storage device (recording medium) such as a memory for storing the program and various data. An object of the present invention is a recording medium on which a program code (execution format program, intermediate code program, source program) of a control program of the character conversion device 1 which is software for realizing the functions described above is recorded so as to be readable by a computer. This can also be achieved by supplying the character conversion apparatus 1 and reading and executing the program code recorded on the recording medium by the computer (or CPU or MPU).

  Examples of the recording medium include a tape system such as a magnetic tape and a cassette tape, a magnetic disk such as a floppy (registered trademark) disk / hard disk, and an optical disk such as a CD-ROM / MO / MD / DVD / CD-R. Card system such as IC card, IC card (including memory card) / optical card, or semiconductor memory system such as mask ROM / EPROM / EEPROM / flash ROM.

  Further, the character conversion device 1 may be configured to be connectable to a communication network, and the program code may be supplied via the communication network. The communication network is not particularly limited. For example, the Internet, intranet, extranet, LAN, ISDN, VAN, CATV communication network, virtual private network, telephone line network, mobile communication network, satellite communication. A net or the like is available. Further, the transmission medium constituting the communication network is not particularly limited. For example, even in the case of wired such as IEEE 1394, USB, power line carrier, cable TV line, telephone line, ADSL line, etc., infrared rays such as IrDA and remote control, Bluetooth ( (Registered trademark), 802.11 wireless, HDR, mobile phone network, satellite line, terrestrial digital network, and the like can also be used. The present invention can also be realized in the form of a computer data signal embedded in a carrier wave in which the program code is embodied by electronic transmission.

As described above, the character conversion device according to the present invention is a character conversion device that converts an input character string composed of one or more input phonograms into a converted character and can determine a word composed of the converted character.
(A-1) conversion character identification information for identifying a conversion character and conversion character correspondence information in which conversion character phonetic information indicating the phonetic of the conversion character is associated, and (a-2) input And the converted character storage location information indicating the storage location in the converted character correspondence information of the converted character identification information having the phonetic character as the head of the converted character phonetic information. A converted character dictionary having a converted character index;
(B-1) word correspondence information in which word identification information for identifying a word and word phonetic information indicating the phonetic expression of the word are associated with each other, and (b-2) head conversion that becomes the head of the word First conversion character identification information for identifying a character is associated with word storage destination information indicating the storage destination of the word having the first conversion character indicated by the first conversion character identification information in the word correspondence information. A word dictionary having a word index,
(C) Conversion with reference to the converted character dictionary based on the first phonogram of the input character string that is input, and having the first phonogram of the input character string as the head of the converted character phonogram information First conversion character acquisition means for acquiring character identification information;
(D) a word acquisition unit that refers to the word dictionary based on the acquired head conversion character identification information and acquires a word having a conversion character at the head indicated by the head conversion character identification information;
It has.

  In addition to the above configuration, the character conversion device according to the present invention includes a display control unit that displays the first conversion character indicated by the first conversion character identification information acquired by the first conversion character acquisition unit on the display unit in a selectable manner by the user. And receiving means for receiving the first conversion character selected by the user, wherein the word acquisition means performs processing for acquiring a word based on the first conversion character identification information of the selected first conversion character. May be.

  According to the above configuration, a word can be acquired based on the first conversion character selected by the user, that is, by narrowing down the first conversion character.

  In addition to the above configuration, the character conversion device according to the present invention further includes display auxiliary information associated with the converted character indicated by the converted character identification information in the converted character correspondence information. The display control means may rearrange the head conversion characters on the basis of the display auxiliary information and display them on the display unit so that the user can select them.

  According to the said structure, based on display auxiliary information, a head conversion character can be rearranged and it can display on a display part so that a user can select. The display auxiliary information may be information on the appearance frequency of the Chinese character, for example. The display auxiliary information may be other information, and may be any information as long as the display control means can rearrange the head conversion characters.

  Further, in addition to the above configuration, the character conversion device according to the present invention includes a word index of the word dictionary in which the first conversion character identification information and the first character of the second character The storage location in the word correspondence information may be allocated to the combination with the second first phonogram that is the phonogram.

  According to the above configuration, the word acquisition unit can search for a word from the combination of the first converted character identification information and the second first phonogram which is the first phonogram of the second character of the word. Therefore, the search range can be narrowed down more quickly.

  Further, in the character conversion device according to the present invention, in addition to the above configuration, in the conversion character correspondence information, acquisition auxiliary information about the conversion character indicated by the conversion character identification information is further associated with the conversion character identification information. The head conversion character acquisition means may acquire the conversion character identification information based on the acquisition auxiliary information.

  According to the above configuration, the head conversion character acquisition unit can acquire the conversion character identification information based on the acquisition auxiliary information about the conversion character indicated by the conversion character identification information. The acquisition auxiliary information may be information that assists the search using the conversion character dictionary of the head conversion character acquisition means, and is, for example, the length of the phonetic character of the head conversion character.

In the character conversion device according to the present invention, in addition to the above configuration, the acquisition auxiliary information may be classified into the following four pieces of information.
(1) Information that the converted character indicated by the converted character identification information is unconditionally acquired with respect to the input character string.
(2) Information that the conversion character indicated by the conversion character identification information is excluded from acquisition if the input character string is one character.
(3) Information that the conversion character indicated by the conversion character identification information is excluded from acquisition if the input character string is two characters or less.
(4) Information that the conversion character indicated by the conversion character identification information is not subject to acquisition if the input character string is three characters or less.

  According to the above configuration, the leading conversion character acquisition means can acquire the leading conversion character quickly and effectively.

  In addition to the above configuration, the character conversion device according to the present invention may be arranged as a group of words having the same conversion character at the beginning of the word in the word correspondence information.

  According to the above configuration, the word acquisition unit can search for words grouped as a group, and thus can acquire words quickly and effectively.

  Further, in each of the above groups, the words may be arranged in a predetermined arrangement order of the first phonetic characters of the words.

  According to the above configuration, the word acquisition unit can search for a word in the group arranged in the predetermined arrangement order of the phonograms at the beginning of the word, so that the word can be acquired more quickly. Here, the predetermined arrangement order of the phonograms at the beginning of the word is, for example, alphabetical order when the word reading is Pinyin, or alphabetical order when hiragana is used.

  In addition to the above configuration, the character conversion device according to the present invention is such that the conversion character is a kanji character, and the conversion character correspondence information includes the same phonetic code and the same phonetic code in the kanji character. Kanji identification information that is converted character identification information may be grouped together.

  According to the above configuration, the head conversion character acquisition unit only needs to search for the head conversion character to be obtained in the group, so that the response speed is increased.

  Further, in the converted character correspondence information of the character conversion device used as a Chinese conversion device, the kanji identification information may be arranged in a predetermined arrangement order of the first phonetic characters of the kanji in each group. Good.

  Since the first phonetic characters of the Chinese characters are arranged in a predetermined arrangement order, the first converted character acquisition means can acquire the first converted character more quickly.

  Further, in the conversion character index of any one of the character conversion devices used as a Chinese conversion device, the first phonetic characters of the Chinese characters are in a predetermined arrangement order, and the conversion character corresponding to each group is supported. Information indicating the storage location in the information may be associated with the first phonetic character of the Chinese character as the converted character storage information.

  According to the above configuration, in the conversion character index, the first phonetic characters of the kanji are in a predetermined arrangement order, and information indicating the storage destination in the conversion character correspondence information of each group is used as the conversion character storage information. It is associated with the first phonetic character of the kanji. Therefore, the head conversion character acquisition means can search by narrowing down the group from the storage destination in the conversion character correspondence information of the group and increasing the response speed.

  Here, in any one of the character conversion devices used as the Chinese conversion device, the kanji may be simplified and the phonetic may be expressed in pinyin.

According to the above configuration, when a user inputs a pinyin string of one or more characters, the first kanji candidates are narrowed down from the first character in the pinyin string, and the search target words can be limited based on the first kanji candidates. This time can be shortened. Therefore, a Chinese input device with a high response speed can be provided. In addition, it is possible to provide a character conversion device that can input a desired character string even if only the first character Pinyin of the character string to be input is input or if the second and subsequent characters are input continuously. On the other hand, the pinyin of the first character of the character string to be input is incomplete, and a Chinese input device can be provided that can input a desired character string even if the pinyin after the second character is input. In any one of the character conversion devices used as the word conversion device, the kanji may be a traditional character, and the phonetic may be expressed by a note.

  According to the above configuration, when the user inputs one or more sung character strings, the first kanji candidates are narrowed down from the first character of the sung character string, and the search target words can be limited based on the first kanji candidates. Processing time can be shortened. Therefore, a Chinese input device with a high response speed can be provided. Also, a Chinese input device that can input a desired character string even when only the first sung character of the character string to be input is input or when the second and subsequent sung characters are continuously input. Can be provided. On the other hand, it is possible to provide a Chinese input device capable of inputting a desired character string even when a second or subsequent character is input because the first character of the character string to be input is incomplete.

Moreover, the control method of the character conversion apparatus according to the present invention converts the input character string composed of one or more input phonograms into a converted character, and can determine a word composed of the converted character. (A-1) conversion character correspondence information in which conversion character identification information for identifying conversion characters and conversion character phonetic information indicating the phonetics of the conversion characters are associated with each other, and (a -2) Conversion character storage location information indicating a storage location in the conversion character correspondence information of the input phonetic character and the conversion character identification information having the phonetic character as the head of the conversion character phonetic information; A conversion character dictionary having a conversion character index associated with the word, (b-1) word identification information for identifying the word, and word correspondence information indicating the phonetic expression of the word Correspondence information and (b-2) First conversion character identification information for identifying a conversion character at the beginning of a word, and word storage destination information indicating a storage destination in the word correspondence information of a word having the conversion character indicated by the first conversion character identification information at the head And a word dictionary having a word index associated with
Conversion character identification information that refers to the conversion character dictionary based on the input phonetic character of the input character string and has the head phonetic character of the input character string as the head of the conversion character phonetic information A first conversion character acquisition step for acquiring
Based on the acquired head conversion character identification information, referring to the word dictionary, a word acquisition step of acquiring a word having a conversion character at the head indicated by the head conversion character identification information;
Processing including

  According to the above method, as in the character conversion device, it is possible to reduce the time required for word search processing.

  The character conversion device according to the present invention may be realized by a computer. In this case, the character conversion device is operated by the computer by causing the computer to operate as each of the above-described means in any of the character conversion devices. A character conversion program to be realized and a computer-readable recording medium on which the character conversion program is recorded also fall within the scope of the present invention.

  According to these structures, the same effect as the said character conversion apparatus is realizable by making a computer read and execute a character conversion program.

[Additional notes]
The character conversion device according to the present invention is a character conversion device that converts an input phonetic character or phonetic character string into a corresponding ideogram and inputs the ideogram or ideogram string selected by the user. A character recording unit that records a phonetic character and an ideogram corresponding to the phonetic character string including the phonetic character in association with each other, and accepts a phonetic character input by the user and associates the phonetic character with the phonetic character. Character candidate extraction means for extracting ideographic characters from the character recording unit as ideographic character candidates, candidate order determination means for determining the display order of candidates extracted by the candidate extraction means based on predetermined conditions, and the candidates Display control means for outputting data so as to display candidates based on the display order determined by the order determination means is provided.

  According to the said structure, a character candidate extraction means receives the phonetic character which the user input, and specifies the phonetic character string containing the phonetic character. The phonetic character string corresponds to the ideogram. The correspondence relationship between the phonetic character and the phonetic character string and the correspondence relationship between the phonetic character string and the ideographic character are recorded in the character recording unit. Therefore, the character candidate extraction unit can extract from the character recording unit candidates of ideographic characters that the user desires to input based on the phonetic characters received from the user.

  Subsequently, the candidate order determination means rearranges the ideographic character candidates extracted by the character candidate extraction means based on a predetermined condition, and determines the candidate display order. The display control means outputs candidate data so that the ideographic character candidates are displayed in the display order determined by the candidate order determination means.

  Thus, the user can input the ideogram only by designating the desired ideogram from the displayed candidates.

  From the above, the user can obtain the input of a desired ideogram with a smaller number of operations related to the input of phonogram characters.

  More specifically, for example, it is assumed that the phonogram is an alphabet and the ideogram is a Chinese kanji. That is, the alphabet character string is Pinyin, and one Pinyin corresponds to one or a plurality of Chinese characters. In the character recording portion, alphabet-pinyin and pinyin-kanji correspondences are recorded.

  Therefore, when one alphabetic character is input, the character candidate extraction unit refers to the character recording unit, identifies pinyin starting from the alphabet, and extracts one or more kanji characters corresponding to the identified pinyin as candidates. .

  The kanji extracted as the candidate is presented to the user. The user only has to specify a desired kanji from the presented candidates. That is, after inputting the alphabet, it is possible to designate and obtain a desired kanji without requiring an operation of selecting the corresponding pinyin.

  From the above, it is possible to provide a character conversion device that can simplify the input operation of phonetic characters (pinyin) to obtain a desired ideogram (kanji).

  Since there are four voices in Chinese pronunciation, there are four types of pronunciation for one Pinyin notation. The information for distinguishing these four voices may be added to record the corresponding kanji. For example, if the field for storing the distinction of the four voices is prepared in the character recording unit, and the corresponding kanji is further narrowed down based on the distinction of the four voices input by the user, the extraction of the character candidate extraction means Processing can be performed efficiently.

  Alternatively, for example, the phonetic character may be a letter of the syllabary and the ideogram may be a traditional character. In other words, the phonetic character string is a note, and one or more traditional characters correspond to one note. In the character recording unit, it is only necessary to record the correspondence between the note-moji-to-note and the note-to-traditional character.

  The character conversion device further accepts an ideogram selected by the user by associating the ideogram with a character string recording unit that records the ideogram and an ideogram including the ideogram, and is associated with the ideogram. There may be provided character string candidate extracting means for extracting the ideographic character string from the character string recording unit as an ideographic character string candidate.

  Alternatively, the character conversion device of the present invention converts the input phonetic character or phonetic character string into a corresponding ideographic character and inputs the ideographic character selected by the user. A character string recording unit that records the ideographic character string including the ideographic character in association with each other, accepts the ideographic character selected by the user, and converts the ideographic character string associated with the ideographic character into candidate ideographic character strings. The character string candidate extracting means for extracting from the character string recording unit, the candidate order determining means for determining the display order of the candidates extracted by the character string candidate extracting means based on a predetermined condition, and the candidate order determining means Display control means for outputting data so as to display candidates based on the determined display order is provided.

  According to the above configuration, the correspondence between the ideographic characters (for example, kanji) and the ideographic character strings including the ideographic characters (for example, idioms starting with the kanji) is recorded in the character string recording unit. Yes. Therefore, the character string candidate extraction unit extracts an ideographic character string associated with the ideographic character from the character string recording unit as a candidate in response to selection of the ideographic character by the user.

  For example, explanation will be given using an example of Chinese. When a certain Chinese character is input, the character following the Chinese character (the Chinese character desired by the user) is more likely to be a Chinese character that constitutes an idiom starting from the Chinese character than a Chinese character without any context. Therefore, if a idiom that starts with a kanji that has been confirmed to be input is displayed as a candidate, the user can select the desired idiom from the idiom candidates and input Chinese without inputting pinyin separately. It becomes possible.

  As a result, the user can obtain a desired ideogram input with a smaller number of operations related to the input of phonogram characters.

  The character conversion device further includes a written character input unit for inputting an ideogram written by a user, recognizes the ideographic character input to the written character input unit, and converts the information of the ideographic character into the character string. It is preferable that character recognition means to be supplied to the candidate extraction means is provided.

  Thereby, even when the phonetic character (sequence) corresponding to the desired ideographic character is not clear, it is possible to obtain the input of the desired ideographic character.

  Furthermore, it is preferable that the character candidate extraction unit of the character conversion device extracts, as candidates, ideographic characters corresponding to the phonetic character strings received continuously when the phonetic characters input by the user are received.

  According to the above configuration, each time a phonetic character is input by the user, the number of candidates for the phonetic character string to be specified is reduced. Therefore, the character candidate extraction means can significantly narrow down and extract the number of corresponding ideographic character candidates, and the display control means can output the narrowed ideographic character candidates.

  From the above, the character conversion device can be configured so that the probability that the displayed ideographic character candidate includes the ideographic character desired by the user increases each time the phonic character is input by the user. . As a result, it is possible to provide a character conversion device capable of simplifying a phonetic character input operation for obtaining a desired ideogram.

  Further, the candidate order determination means of the character conversion device displays the ideographic character string candidates extracted by the character string candidate extraction means in preference to the ideographic character candidates extracted by the character candidate extraction means. It is preferable to determine the display order.

  According to the above configuration, the ideographic character string candidate extracted by the character string candidate extraction unit from the character string recording unit is displayed first than the ideographic character candidate extracted by the character candidate extraction unit from the character recording unit. The candidate order determining means determines the candidate display order.

  For example, in Japanese or Chinese kanji input, a kanji that forms an idiom (ideographic character string) follows an already input kanji, rather than a kanji without any context (desired by the user). Probability is high. Therefore, for example, if an idiom starting from a kanji that has been confirmed to be input is displayed as a candidate, the user can select a desired kanji faster and with a smaller number of operations than when displayed from a candidate for a kanji (ideographic character). It becomes possible to input.

  The character conversion device further includes learning control means for recording an ideographic character selected by the user as a learning character in a learning word recording unit, and the candidate order determination means includes the candidate extracted by the character candidate extraction means as the candidate When the learning characters are recorded in the learning word recording unit, the display order may be determined so that the candidates are displayed with priority.

  Alternatively, it comprises learning control means for recording an ideographic character string selected by the user as a learned character string in a learned word recording unit, and the candidate order determining means is configured such that the candidate extracted by the character string candidate extracting means is recorded as the learned word record. In the case of the learned character string recorded in the part, the display order may be determined so that the candidate is displayed with priority.

  According to the above configuration, the learning control unit causes the learning word recording unit to record the ideographic character or the ideographic character string whose input is confirmed (desired by the user) in the character conversion device as the learning character (string).

  And after the said candidate extraction means extracts a candidate from a character (sequence) recording part, a candidate order determination means determines what was recorded on the said learning word recording part among the extracted candidates as a learning character. can do. That the candidate is a learning character means that the user is more likely to be desired again than a candidate that is an ideographic character (sequence) that the user has desired in the past and is not a learning character.

  From the above, it is possible to realize a character conversion device that allows a user to input a desired kanji faster and with a smaller number of operations by raising the display order of candidates that are learning characters.

  Alternatively, the character conversion device further records the ideographic character selected by the user in the learning word recording unit as a pair learning character in association with the immediately preceding character or the immediately preceding character string input immediately before the ideographic character. A learning control unit, wherein the candidate order determination unit is a pair learning character recorded in the learning word recording unit in association with the immediately preceding character or the immediately preceding character string. The display order may be determined so that the candidates are displayed with priority.

  Alternatively, a learning control unit is provided that records the ideographic character string selected by the user in the learning word recording unit as a pair learning character string in association with the immediately preceding character or the immediately preceding character string input immediately before the ideographic character string. The candidate order determining means, when the candidate extracted by the character string candidate extracting means is a pair learning character string recorded in the learning word recording unit in association with the immediately preceding character or the immediately preceding character string, the candidate The display order may be determined so that is displayed with priority.

  According to the above configuration, when the learning control unit records the ideographic character or the ideographic character string whose input is confirmed (desired by the user) in the character conversion device as the learning character (string) in the learning word recording unit, It is recorded as a pair learning character (string) in association with a character (string) whose input has already been confirmed immediately before the character (string) (hereinafter referred to as the immediately preceding character (string)).

  The pair learning character (sequence) has a history desired by the user following the associated immediately preceding character (sequence). Therefore, when the immediately preceding character (string) is input, the pair learning character (string) is highly likely to be desired by the user.

  When the candidate extracted by the candidate extraction unit is a pair learning character (string), the candidate order determination unit changes the display order of the candidates to the top.

  Thereby, it is possible to realize a character conversion device that allows a user to input a desired kanji character earlier and with a smaller number of operations.

  Furthermore, the learning control means of the character conversion device generates a group identifier for classifying the learning character or learning character string or the pair learning character or pair learning character string as a learning word, An identifier is associated with each learning word and recorded in the learning word recording unit, and the group identifier is generated based on a character code having a predetermined length assigned in advance to ideographic characters constituting the learning word. Is preferred.

  According to the above configuration, the learning control means divides the learning word (learning character or learning character string, or pair learning character or pair learning character string) recorded in the learning word recording unit into a plurality of groups. Generate and assign group identifiers for classification. The learning control means generates the group identifier using a numerical value that is assigned in advance to each ideographic character constituting the learning word and that is obtained based on the character code.

  The group identifier generated in this way has a fixed length regardless of what learning word is assigned. That is, to give a specific example, a group identifier of the same size is assigned to a learning word consisting of a two-character idiom or a learning word consisting of a six-character idiom.

  As the number of learning words recorded in the learning word recording unit becomes enormous, or when it is a idiom, the data size increases as the number of characters increases. That is, it takes a long time to search for a matching learning word from the learning word recording unit in order to determine whether or not the candidate is a learning word, resulting in a problem that the processing efficiency of character input decreases. .

  However, as described above, if a fixed-length group identifier is assigned to each learning character, the time for determining whether the input character (string) matches the learning word or the character (string) in the same group in the search is as follows. Regardless of the number of characters in the learning word, it is uniform. As described above, it is possible to suppress a proportional increase in the time required for the search with respect to the increase in learning words, and to reduce the influence on the decrease in processing efficiency.

  The character conversion device further records the degree of association between the ideogram or ideogram string selected by the user and the forward character or the forward character string whose input is confirmed before the ideogram or ideogram string. The candidate order determination unit may include a relevance degree recording unit, and the candidate order determination unit may determine the display order such that the candidates are displayed with higher priority as the degree of relevance with the front character or the front character string is higher.

  According to the above configuration, the candidate order determining means determines the candidate display order based on the degree of association with the forward character (string). The degree of relevance is the relationship between the ideogram or ideographic character string whose input has been confirmed by the character input confirmation means and the front character or the front character string whose input has been confirmed before the ideogram or ideographic character string. That is.

  More specifically, for example, a case where the phrase “Shanghai Jiaotong University” is frequently input will be described. Even if there is a high probability that another phrase such as “rule” will be input in succession after “traffic”, if “Shanghai” precedes it, the candidate order determining means Considering the relationship between “Shanghai” and “University”, “University” is changed to a higher rank candidate.

  Thereby, the candidate order determining means considers not only the continuity with the immediately preceding character (sequence) but also the relationship with the preceding character (sequence), and the candidate ideographic characters (sequence). Since the display order is determined, the candidates can be presented in an order more reflecting the user's wishes.

  As a result, the user can obtain the input of a desired ideographic character with a smaller number of operations with respect to the number of operations related to the input of phonetic characters.

  Furthermore, the character conversion device includes a candidate order determination element setting unit that sets a predetermined condition that is specified by the user as a predetermined condition to be referred to when the candidate order determination unit determines the display order of candidates. Preferably, the candidate order determining means determines a candidate display order based on the conditions set by the candidate order determining element setting means.

  According to the above configuration, when there are a plurality of predetermined conditions to be referred to when the candidate order determination unit determines the display order of candidates, the user's instruction is used to determine the display order based on what conditions The candidate order determination element setting unit sets the response according to the above.

Several examples of the predetermined condition are conceivable. For example, as mentioned above,
(1) whether the candidate is a character (kanji), a character string (jukugo), (2) whether the candidate is a learning character, (3) whether the candidate is a pair learning character, 4) When was it learned, (5) How often it is generally used, (6) Is the phonetic character length currently input as reading consistent with the candidate phonetic character length? No (for example, when “technical” is desired to be input, “giji (2 characters)” and “gijutsu (4 characters)” are inconsistent in phonetic character length), (7) continuous with previous character string There are various conditions such as (8) whether there is a relationship with the forward character string, (9) how often it is used by the user.

  Here, if the user can arbitrarily set the display order of the candidates based on what conditions (or what conditions are emphasized), the candidates determined by the candidate order determination means The display order reflects the user's wish more accurately.

  As a result, the user can obtain the input of a desired ideographic character with a smaller number of operations with respect to the number of operations related to the input of phonetic characters.

  A copying machine according to the present invention includes the above-described character conversion device and a first input unit that includes a key corresponding to each phonetic character and supplies the phonetic character corresponding to the key pressed by the user to the character conversion device. And the second input unit for inputting the candidate selected by the user to the character conversion device and the candidates extracted by the character candidate extraction unit or the character string candidate extraction unit in the display order determined by the candidate order determination unit. Accordingly, the display unit includes a display unit, and the first input unit, the second input unit, and the display unit are touch panels or tablets formed integrally with each other.

  Thus, the first and second input units for inputting the phonetic characters and ideograms selected by the user to the character conversion device, and the display unit for presenting the ideographic characters (sequences) as candidates to the user In a copier having a tablet integrated with the above, the user can obtain a desired ideogram input with a smaller number of operations related to the input of phonogram characters.

  The copying machine further includes a written character input unit for inputting an ideographic character written by a user, recognizes the ideographic character input to the written character input unit, and supplies information of the ideographic character to the character conversion device. It is preferable to include a character recognition means.

  As a result, it is possible to realize a copier that can obtain the input of a desired ideogram even when the phonogram (column) corresponding to the desired ideogram is not clear.

  A character input method according to the present invention is a character input method in a character conversion device that converts an input phonetic character or phonetic character string into a corresponding ideographic character and inputs an ideographic character or ideographic character string selected by the user. The phonetic character input by the user is received, and the ideogram associated with the phonetic character is the phonogram and the ideogram corresponding to the phonetic character string including the phonetic character. A first step of extracting as an ideographic character candidate from a character recording unit to be recorded in association; a second step of determining a display order of candidates extracted in the first step based on a predetermined condition; And a third step of outputting data so as to display candidates based on the display order determined in the steps.

  According to the above method, in the first step, candidates for ideographic characters corresponding to the phonetic character string including the phonetic character input by the user are extracted, and in the second step, the display order of the extracted candidates is determined. Is done. At this time, the display order may be determined based on the predetermined condition as described above, for example. In a third step, ideographic character candidates are output based on the determined display order.

  Thus, when a desired ideogram is designated from the candidates by the user, the input of the designated ideogram is confirmed.

  From the above, it is possible to provide a character conversion device that can simplify the input operation of phonetic characters (pinyin) to obtain a desired ideogram (kanji).

  The character conversion device and the copying machine may be realized by a computer. In this case, a control program for causing the character conversion device or the copying machine to be realized by the computer by causing the computer to operate as each of the above means, A computer-readable recording medium on which it is recorded also falls within the scope of the present invention.

[Supplement 2]
The character conversion device according to the present invention is a character conversion device that converts an input phonetic character or phonetic character string into a corresponding ideogram and inputs the ideogram or ideogram string selected by the user. A character recording unit that records a phonetic character and an ideogram corresponding to the phonetic character string including the phonetic character in association with each other, and accepts a phonetic character input by the user and associates the phonetic character with the phonetic character. Character candidate extraction means for extracting ideographic characters from the character recording unit as ideographic character candidates, candidate order determination means for determining the display order of candidates extracted by the candidate extraction means, and the candidate order determination means Display control means for outputting data so as to display candidates based on the display order, and an ideographic character or ideographic character string input in the past; The relevance degree is a degree of relevance indicating a high possibility that the ideogram or the ideographic character string is input behind the front character or the front character string in association with the front character or the front character string inputted in front. Stored in the recording unit, the candidate order determining means detects a degree of association between each candidate and a forward character or a forward character string input in front of the candidate from the association degree recording unit, The display order is determined so that candidates with higher relevance are displayed with higher priority.

  Alternatively, the character conversion device of the present invention converts the input phonetic character or phonetic character string into a corresponding ideogram, and inputs the ideogram selected by the user. A character string recording unit that records the ideographic character string including the ideographic character in association with each other, accepts the ideographic character selected by the user, and uses the ideographic character string associated with the ideographic character as a candidate for the ideographic character string. Based on the character string candidate extracting means extracted from the character string recording unit, the candidate order determining means for determining the display order of the candidates extracted by the character string candidate extracting means, and the display order determined by the candidate order determining means Display control means for outputting data so as to display candidates, and an ideographic character or ideographic character string input in the past, and further forward from immediately before the ideographic character or ideographic character string The relevance degree recording unit associates the inputted front character or the front character string with each other and indicates the degree of possibility that the ideographic character or the ideographic character string is input behind the front character or the front character string. And the candidate order determination means detects the degree of association between each candidate and the forward character or the forward character string input in front of the candidate from the association degree recording unit. The display order is determined so that the higher the candidate is, the higher the priority is displayed.

  As a result, the user can obtain a desired ideogram input with a smaller number of operations related to the input of phonogram characters.

  Further, an ideographic character or ideographic character string input in the past is associated with the immediately preceding character or immediately preceding character string input immediately before the ideographic character or ideographic character string, and immediately after the immediately preceding character or immediately preceding character string. The continuity indicating the high possibility that the ideogram or the ideogram is input is stored in the continuity recording unit, and the candidate order determining means includes the relevance with the forward character or the forward character string, In addition, it is preferable to determine the display order so that candidates with higher continuity with the immediately preceding character or the immediately preceding character string are displayed with higher priority.

  Further, the character candidate extraction unit of the character conversion device connects the phonetic character to the end of the received phonetic character string every time it receives a phonetic character input by the user, and is obtained by the connection. It is preferable to extract ideographic characters corresponding to the phonetic character string as candidates.

  Further, the learning control means of the character conversion device classifies the learning character or learning character string, or the pair learning character or pair learning character string as a learning word to be recorded in the learning word recording unit. A group identifier is generated, and the group identifier is recorded in the learning word recording unit in association with each learning word. The group identifier is a predetermined length assigned in advance to ideographic characters constituting the learning word. Preferably, it is generated based on the character code.

  Furthermore, the character conversion device calculates an evaluation value indicating a degree of possibility that the candidate is selected by the user, which is a predetermined condition to be referred to when the candidate order determination unit determines the display order of the candidates. A candidate order determination element setting means for setting an element designated by the user from a plurality of types of elements that can be used for the element, and an element different from the relevance level is stored in the element recording unit in association with each candidate The candidate order determining means preferably determines the display order of candidates based on the relevance and the elements stored in the element recording unit set by the candidate order determining element setting means.

  A character input method according to the present invention is a character input method in a character conversion device that converts an input phonetic character or phonetic character string into a corresponding ideographic character and inputs an ideographic character or ideographic character string selected by the user. The character conversion device includes character candidate extraction means, candidate order determination means, and display control means. The character candidate extraction means accepts a phonetic character input by a user, The ideogram associated with the phonetic character is extracted as an ideographic character candidate from a character recording unit that records the phonogram and the ideogram corresponding to the phonogram string including the phonogram in association with each other. A first step, a second step in which the candidate order determining means determines the display order of the candidates extracted in the first step, and a table determined by the display control means in the second step. A third step of outputting data so as to display candidates based on the order, wherein the second step further includes an ideographic character or ideographic character string input in the past, and immediately before the ideographic character or ideographic character string. Associating with the forward character or the forward character string inputted in front, the degree of relevance indicating the possibility that the ideographic character or the ideographic character string is inputted behind the forward character or the forward character string is stored. From the relevance level recording unit, the degree of relevance between the front character or the front character string input in front of the candidate and each candidate is detected, and the display order is set so that the candidate having the higher relevance degree is displayed with priority. It is characterized by including the step to determine.

  From the above, it is possible to provide a character conversion device that can simplify the input operation of phonetic characters (pinyin) to obtain a desired ideogram (kanji).

  The character input method according to the present invention is a character input method in a character conversion device for converting an input phonetic character or a phonetic character string into a corresponding ideogram and inputting an ideogram selected by a user. The conversion device includes character candidate extraction means, candidate order determination means, and display control means. The character candidate extraction means accepts an ideographic character selected by a user and associates it with the ideographic character. A first step of extracting the ideographic character string as a candidate for the ideographic character string from a character string recording unit for recording the ideographic character string in association with the ideographic character string including the ideographic character, and the candidate order determining means, A second step of determining the display order of the candidates extracted in the first step, and the display control means displays the candidates based on the display order determined in the second step A third step of outputting the data, wherein the second step includes an ideographic character or ideographic character string input in the past, and a forward character or a forward character input in front of the ideographic character or ideographic character string. From the relevance level recording unit that stores a relevance level that indicates a high possibility that the ideographic character or the ideographic character string is input behind the front character or the front character string in association with the front character string. Detecting a degree of association between the forward character or the forward character string input in front of each candidate and each candidate, and determining a display order so that a candidate having a higher degree of association is displayed with priority. It is said.

  The present invention is not limited to the above-described embodiments and examples, and various modifications can be made within the scope of the claims. That is, embodiments obtained by combining technical means appropriately modified within the scope of the claims are also included in the technical scope of the present invention.

  The present invention can be applied to various devices that convert readings (phonetic characters), and in particular, convert readings to kanji.

Claims (22)

In a character conversion device capable of converting an input character string made up of one or more input phonograms into converted characters and determining a word made up of the converted characters,
Conversion character correspondence information in which conversion character identification information for identifying a conversion character and conversion character phonetic information indicating a phonetic of the conversion character are associated with each other, an input phonetic character, and the phonetic A conversion character dictionary having a conversion character index in which the conversion character identification information having a character as the head of the conversion character phonetic information is associated with conversion character storage destination information indicating a storage destination in the conversion character correspondence information. When,
Word identification information for identifying a word, word correspondence information in which word phonetic information indicating the phonetic expression of the word is associated, and head conversion character identification for identifying a conversion character that is the head of the word A word dictionary having information and a word index in which word storage destination information indicating a storage destination in the word correspondence information of a word having a conversion character indicated by the head conversion character identification information at the head is associated with each other;
Conversion character identification information that refers to the conversion character dictionary based on the input phonetic character of the input character string and has the head phonetic character of the input character string as the head of the conversion character phonetic information First conversion character acquisition means for acquiring
Based on the acquired head conversion character identification information, referring to the word dictionary, a word acquisition means for acquiring a word having a conversion character at the head indicated by the head conversion character identification information;
A character conversion device comprising: Display control means for displaying on the display unit the head conversion character indicated by the head conversion character identification information acquired by the head conversion character acquisition means so that the user can select it;
Receiving means for receiving the first conversion character selected by the user,
The character conversion apparatus according to claim 1, wherein the word acquisition unit performs a process of acquiring a word based on the first conversion character identification information of the selected first conversion character. In the converted character correspondence information, display auxiliary information about the converted character indicated by the converted character identification information is further associated with the converted character identification information,
3. The character conversion apparatus according to claim 2, wherein the display control means rearranges the head conversion characters based on the display auxiliary information, and displays them on the display unit so that the user can select them. The word index of the above word dictionary is
For each head conversion character that is the head of a word, the above word correspondence information for the combination of the head conversion character identification information and the second head phonetic character that is the first phonetic character of the second character of the word The character conversion device according to any one of claims 1 to 3, wherein a storage location is assigned. In the conversion character correspondence information, acquisition auxiliary information about the conversion character indicated by the conversion character identification information is further associated with the conversion character identification information,
The character conversion device according to claim 1, wherein the head conversion character acquisition unit acquires the conversion character identification information based on the acquisition auxiliary information. 6. The character conversion apparatus according to claim 5, wherein the acquisition auxiliary information is classified into the following four pieces of information.
(1) Information that the conversion character indicated by the conversion character identification information is unconditionally acquired with respect to the input character string. (2) The conversion character indicated by the conversion character identification information is one character of the input character string. (3) Information that the converted character indicated by the converted character identification information is not acquired if the input character string is 2 characters or less (4) The converted character indicated by the converted character identification information However, if the input string is 3 characters or less   7. The character conversion apparatus according to claim 1, wherein the word correspondence information includes a group of characters having the same conversion character at the beginning of the word.   Furthermore, in each said group, the word is arranged in the predetermined arrangement | sequence order of the phonogram of the head of a word, The character conversion apparatus of Claim 7 characterized by the above-mentioned. The conversion character is Kanji,
In the conversion character correspondence information, the kanji identification information which is the conversion character identification information having the same phonogram at the beginning of the kanji and the same tone code is grouped together.
It is used as a Chinese conversion apparatus, The character conversion apparatus of any one of Claims 1-8 characterized by the above-mentioned.   10. The character conversion apparatus according to claim 9, wherein in the converted character correspondence information, the kanji identification information is arranged in a predetermined arrangement order of the first phonetic characters of the kanji in each of the groups.   In the conversion character index, the first phonetic characters of the kanji are in a predetermined arrangement order, and information indicating the storage destination in the conversion character correspondence information of each group is used as the conversion character storage information, and the kanji The character conversion device according to claim 9, wherein the character conversion device is associated with the first phonogram of the character.   The character conversion device according to any one of claims 9 to 11, wherein the character conversion device is used as a Chinese conversion device in which the kanji is simplified and the phonetic is expressed in pinyin.   The character conversion device according to any one of claims 9 to 11, wherein the character conversion device is used as a Chinese conversion device in which the Chinese character is a traditional character and the phonetic sound is expressed by a note. In a control method of a character conversion device capable of converting an input character string composed of one or more input phonograms into converted characters and determining a word composed of the converted characters,
Conversion character correspondence information in which conversion character identification information for identifying a conversion character and conversion character phonetic information indicating a phonetic of the conversion character are associated with each other, an input phonetic character, and the phonetic A conversion character dictionary having a conversion character index in which the conversion character identification information having a character as the head of the conversion character phonetic information is associated with conversion character storage destination information indicating a storage destination in the conversion character correspondence information. When,
Word identification information for identifying a word, word correspondence information in which word phonetic information indicating the phonetic expression of the word is associated, and head conversion character identification for identifying a conversion character that is the head of the word A word dictionary having information and a word index in which word storage destination information indicating a storage destination in the word correspondence information of a word having a conversion character indicated by the head conversion character identification information at the head is associated with each other;
Using,
Conversion character identification information that refers to the conversion character dictionary based on the input phonetic character of the input character string and has the head phonetic character of the input character string as the head of the conversion character phonetic information A first conversion character acquisition step for acquiring
A word acquisition step of referring to the word dictionary based on the acquired head conversion character identification information and acquiring a word having a conversion character at the head indicated by the head conversion character identification information;
A method for controlling a character conversion device, comprising performing processing including:   The character conversion program for functioning a computer as each means of the character conversion apparatus of any one of Claims 1-13 and 17-22.   A computer-readable recording medium on which the control program according to claim 15 is recorded. The conversion character identification information is associated with display auxiliary information about the conversion character indicated by the conversion character identification information,
Order determining means for determining the display order of the head conversion characters indicated by the head conversion character identification information acquired by the head conversion character acquisition means based on the display auxiliary information;
2. The character conversion apparatus according to claim 1, further comprising display control means for displaying the first converted character on the display unit so that the user can select the first converted character based on the display order determined by the order determining means. Learning control means for recording the first conversion character selected by the user as a learning character in a learning word recording unit,
The order determination means determines whether the first conversion character acquired by the first conversion character acquisition means is a learning character based on display auxiliary information indicating whether the first conversion character is a learning character recorded in the learning word recording unit. 18. The character according to claim 17, wherein the display order is determined so that the first conversion character as a learning character is displayed with priority over other first conversion characters. Conversion device. In the learning word recording unit, a learning time indicating a time when the first conversion character is recorded as a learning character by the learning control unit is stored in association with each first conversion character as the display auxiliary information,
19. The character conversion apparatus according to claim 18, wherein the order determination means determines the display order so that the first conversion character with the new learning time is displayed with priority. The word identification information is associated with display auxiliary information about the word indicated by the word identification information,
Order determination means for determining the display order of the words acquired by the word acquisition means based on the display auxiliary information;
2. The character conversion device according to claim 1, further comprising display control means for displaying the word on the display unit so that the user can select the word based on the display order determined by the order determination means. Learning control means for recording the word selected by the user as a learning word in a learning word recording unit;
The order determination means determines whether the word acquired by the word acquisition means is a learning word based on display auxiliary information indicating whether the word is a learning word recorded in the learning word recording unit. 21. The character conversion device according to claim 20, wherein the display order is determined so that the learning word is displayed with priority over other words. The learning word recording unit stores a learning time indicating the time when the word was recorded as a learning word by the learning control unit in association with each word as the display auxiliary information,
The character conversion apparatus according to claim 21, wherein the order determination means determines a display order so that a word with a newer learning time is displayed with higher priority.

Images