JPS6336551B2 - - Google Patents

Description

[Detailed description of the invention] The present invention relates to a character conversion device.

In a character conversion device, for example, a kana-kanji conversion input device that inputs kana and converts it into kanji, when processing so-called homonyms that have the same pronunciation but different meanings, input kana and display one corresponding homonym, If the displayed kanji is not the desired kanji, a method is used in which different homonyms are displayed one after another and this process is repeated until the desired kanji appears.

In such cases, the order in which homophones are displayed directly affects the input speed of Kana-Kanji conversion, so the following two typical methods of ordering to speed up the conversion input speed have been used: is taken.

The first method is to count the number of times each homonym has been input and used, and to rank the words in descending order of the cumulative number of times they have been input and used in the past. In this method, if you happen to input a homophone that has a small number of cumulative input usages in the past, the desired homophone will be used as the first homophone until the cumulative number of usages of the desired homophone exceeds that of other homonyms. , and the display operation is forced to be repeated, slowing down the conversion input speed. In other words, it depends on the device's quick response to data such as documents to be input.

The second method is the so-called last in first out method, in which the last homonym input and used is placed first in the next display. With this method, if multiple homonyms are frequently input and used in the same document, the first word displayed will change each time, making input more cumbersome and slowing down the conversion input speed. There is a drawback.

SUMMARY OF THE INVENTION An object of the present invention is to provide a character conversion device that eliminates the drawbacks of the conventional devices as described above and can quickly respond to documents and increase the speed of character conversion.

The device of the present invention is a character conversion device that converts a first type of character into one of a plurality of characters of a second type that corresponds to the first type of character, and inputs the first type of character as heading data. an input that generates a conversion instruction signal that instructs to start converting one of the plurality of characters of the second type corresponding to the character, and generates a character selection instruction signal that instructs to select a desired corresponding character of the second type; and each of the plurality of characters of the second type corresponding to the characters of the first type,
a first storage means for storing a corresponding first type of character as header data and as first data together with an area for storing a flag instructing to read the first type of character among the characters of the same header data;
Each of the second type characters most recently selected among the second type characters of the same heading data is set as the corresponding first type character as the heading data,
and a second storage means for storing as second data together with number data indicating the number of consecutive selections, the inputted heading data, and the first
a third memory that stores first data read from the memory means, second data read from the second memory means, and a preset specific value having a value of 2 or more; means, and in response to the supply of the conversion start instruction signal, sequentially reads first data corresponding to the heading data in the third storage means from the first storage means, starting with data with a flag. Displaying the second type of character, and selecting the second type of character being displayed in response to supply of the character selection instruction signal inputted when a desired second type of character is displayed. In addition, when the first data including the selected second type of character is flagged and there is second data of the same heading data as the heading data in the third storage means, this second data is output. 2, and when the first data containing the selected second type of character is not flagged, the selected second type of character is the same as the heading data in the third storage means. The same second data is read from the second storage means, and when the number of times data is equal to (specific value - 1), this second data is erased from the second storage means, and the erased second data is read out from the second storage means. A flag is added to the first data of the first storage means corresponding to the data, and when there is a flag in other first data of the same heading data, the latest flag is made valid and the previous flag is deleted, and the When the data is smaller than (specific value - 1), 1 is added to the number of times data and stored again, and a second type of character that is the same as the selected second type of character is stored in the second storage means. If the same heading data is not available or the same heading data is not available, count data 1 is added to the first data containing the selected and outputted second type of character, and if the same heading data is in the second storage means, this second data is added. and a comparison control means for replacing the data with the data and storing it as new second data when the data is missing.

Next, the present invention will be explained in detail with reference to the drawings.

In the present invention, taking a kana-kanji conversion input device as an example, the number of times one word among homophones is inputted consecutively (hereinafter referred to as the learning degree and indicated by α) is counted, and this number of times is counted. The word for which the value exceeds a certain value (hereinafter referred to as the inertia value and indicated by A) is given the first rank of display among homophones. If a word whose learning level α is greater than or equal to the inertia value A becomes 2 or more homophones, the word whose learning level α is greater than or equal to the inertia value A will be given the first ranking in the display. There is.

In this way, the shortcoming of the conventional method of ordering according to the past cumulative number of used inputs, which is the non-immediate response to input documents, can be eliminated, and the last-in-first-out method, which is another conventional method, can be eliminated. We focused on the fact that when one word among homonyms is used consecutively, there is a high probability that that word will be used next. give the first rank of display to
In addition, priority can be obtained only when the number of consecutive uses, which is the criterion for giving the first priority, exceeds an arbitrarily set inertia value, thereby improving responsiveness and adaptability to input documents.

FIG. 1 is a block diagram of one embodiment of the present invention. The keyboard section 1 has a large number of keys, and has functions such as inputting kana that needs to be converted to kanji, instructing the start of conversion to kanji, and selecting a desired kanji. Dictionary memory 2 stores a dictionary for kana-kanji conversion, and includes, for search purposes, a heading area arranged in alphabetical order, a homophone number area indicating the number of homophones, and a kanji code for the kanji corresponding to the heading. It consists of a Kanji area and a flag area indicating the first rank. Generally, the order of the kanji codes for the homophone kanji corresponding to the headings will be in order of statistical frequency. The appearing word memory 3 stores kanji that have been input into kana-kanji conversion, and consists of a heading area, a kanji area, and a learning level area indicating the learning level α. The register 4 has a function of temporarily storing data from each part, and the register 4-1 (register 4-1) stores data from the dictionary memory 2.
(also has a learning degree storage location corresponding to the learning degree area of the appearing word memory 3) Register 4-2 stores the heading input from the keyboard section 1, and stores the inertia value input from the keyboard section 1. It consists of a register 4-3 and a register 4-4 that stores data input from the occurring word memory 3. Adder 5 has the function of adding the number 1 to the learning degree indicated in the learning degree area of registers 4-1 and 4-4. The comparison unit 6 compares and inspects the contents of each register in the register unit 4,
It has the function of issuing various commands depending on the results. The display section 7 has a function of displaying the kanji selected by the comparison section 6.

FIG. 2 is a flowchart showing the operation of the embodiment of FIG. 1, which is made up of 33 steps from step 31 to step 63. "Clear words with the same heading" in steps 44 and 59 in FIG. "Replace set" means to "set a new flag in the flag area of the word in the dictionary memory 2 that corresponds to the selected word, and if a flag was previously set for the same heading as the selected word," If there is another word, the previous flag is deleted and the flag is replaced (replaced and set) with the word corresponding to the newly selected word.

The operation of one embodiment of the present invention will be described in detail with reference to FIGS. 1 and 2. The first explanation of the operation is
The explanation will be divided into three cases: initial input, second, when there are no words that have reached the inertia value, and third, when there are words that have reached the inertia value.

In the initial input in the first case, no flag is attached to the flag area of the dictionary memory 2, and there is no appearing word in the appearing word memory 3, so the storage is not performed. When using, first set the inertia value A by operating the keys on the keyboard section 1.
The inertia value A is stored in the register 4-3. Does it need to be converted to kanji (hereinafter referred to as headings)?
If there is, enter this kana from keyboard section 1,
A command for conversion is issued by pressing a conversion key on the keyboard section 1. These are steps 32 and 33, where the heading is input from the keyboard section 1 to the register 4-2, and when the conversion key is pressed, reading of the dictionary memory 2, which is the first operation of conversion, begins, and the read word is stored in the register. 4-1, and its heading is compared with the contents of register 4-2 by comparison unit 6,
The word stored at the lowest address in the dictionary memory 2 under the same heading is searched and displayed on the display section 7 (steps 34, 35, 45 and 46). The operator checks whether the word displayed on the display section 7 is the desired word (step 47), and if it is not the desired word, the operator checks whether it is the final record or not (step 48).
After that, the operator presses down the NEXT key on the keyboard section 1 (step 39) to read and display the next word under the same heading from the dictionary memory 2, and repeats the above steps to find the desired word. If the desired word is not found, the process goes to step 48, displays that conversion is not possible (step 49), and ends the process. When the desired word is displayed on the display section 7, the operator depresses the input key on the keyboard section 1 in order to input the kanji into a predetermined device (step 50). This completes the input operation of converting and displaying the desired kanji, but the operation of specifying the word to be converted and displayed first based on the learning level of homophones and the learning of homophones. is about to begin. In step 51, the comparator 6 checks the number of homophones in the data stored in the register 4-1, and if it is "0", there is no homophone, so there is no need to order the conversion display, and the operation ends here. do. If the number of homophones is not “0”, step 5
Proceed to step 2 and inertia value A = 1 (register 4-
3), this is the ranking in which the word used now has the highest probability of being used next. This is a case where one rank is given to this,
The dictionary memory 2 attaches a flag to the flag area of the register 4-1 based on the flag change set signal from the comparator 6 (specifically, changes the flag bit from 0 to 1).
(step 60). If the inertia value is greater than 1 in step 52, the data with the same heading as the data stored in register 4-1 is read out from the occurrence word memory 3 in response to the occurrence word memory read signal from the comparing section 6 (step 52). 53→Step 54). In the initial state, as mentioned above, nothing is stored in the appearing word memory 3, so the process naturally proceeds to step 61, where the adder 5 is operated by the appearing word memory write signal from the comparator 6.
Learning degree of data stored in register 4-1 (The learning degree area of register 4-1 is normally “0” and becomes “1” only when receiving the addition signal from the adder)
(The learning level changes from "0" to "1"), and the data in the heading area, kanji area, and learning level area of the contents of register 4-1 is newly written to the appearing word memory 3. The operation ends.

The above is the operation of the embodiment corresponding to a heading input for use for the first time.

Next, a second case will be described in which the inertia value A is greater than 1 and a heading that has been input for use once in the past is input for use again. In this case, since the learning level is still smaller than the inertia value, no flag is attached to the corresponding heading word in the dictionary memory 2. Therefore, up to step 53, the operation proceeds in the same manner as described above. In the operation of step 54, data with the same heading is read from the occurrence word memory 3 to the register 4-4 by the occurrence word memory read signal from the comparison section 6, and the data of the same heading is read out from the occurrence word memory 3 to the register 4-4.
and register 4-1 are compared. This time, since words with the same heading are stored in the appearing word memory 3, the process advances to step 55, and registers 4-1 and 4 are stored.
The comparison unit 6 compares the data of the kanji area 4. If they are not the same, the adder 5 operates in response to the appearing word memory write signal from the comparison unit 6, and adds "1" to the learning degree of the data stored in the register 4-1 (the learning degree is " (changes from "0" to "1") The data in the index area, kanji area, and learning level area of register 4-1 is written to the location where words with the same heading are stored in the occurrence word memory 3, that is, the same heading. The data is rewritten and the operation ends (step 61→step 62→step 63). The learning level α in the present invention means the number of times a homophone homophone word is used consecutively in a homophone, so words that are not used consecutively are deleted from the occurrence word memory 3. . As a result, if words having the same sound as any of the words are stored in the appearing word memory 3, only one word is stored.

Next, as a result of the comparison of the kanji areas in step 55 in the comparing section 6, if they are the same, the register 4--
The learning degree α of the data stored in 4 is the inertia value A
The comparator 6 checks whether the value is equal to the value obtained by subtracting 1 from the above (step 56). This is to check whether the learning degree has just become equal to the inertia value in the current use of this data. In the case of A-1≠α, the adder 5 is operated by the appearing word memory rewrite signal from the comparison unit 6, and this time the register 4-
4 is added to the learning degree α of the data stored in register 4-4, and the content stored in the same heading of the occurrence word memory 3 is replaced with the resultant contents of register 4-4, and the operation is completed (step 4). 57→Step 58→Step 63). This indicates that in the case of continuous use of a word with the same sound and same kanji that does not reach the inertia value, the number of times the word has been used is stored in the word memory. If A-1=α in step 56, the words with the same heading in the occurrence word memory 3 are cleared by the occurrence word memory clear signal from the comparison section 6, and at the same time, the words with the same heading are cleared by the flag change set signal from the comparison section 6. , attaches a flag to the data in register 4-1 and stores it in dictionary memory 2, and also sets the flag bits of other words with the same heading in dictionary memory to "0" and ends (step 59 → step 60 → step 63) . This is to position words whose learning level has reached the inertia value in the first rank in the conversion display from now on, and the words that were previously in the first rank are no longer in the first rank. The above operations, except for step 60, have been explained in the case where there is no flagged word.

Finally, as a third case, a case where there is a word with a flag, that is, a case where there is a word whose learning level has reached an inertia value, will be summarized below. If there are flagged words,
Since that word has the highest priority in the conversion display, when reading the dictionary memory in step 34, the same headword with a flag is first searched for, and the word is found in steps 35 and 36 (register 4-2 and register 4-2).
1 in the comparison section 6, and check that the flag bit is "1" in the comparison section 6), and the display section 7
will be displayed. The operator looks at the displayed word and checks whether it is the desired word (step 38). If the word is not the desired one, the user presses the NEXT key, displays the homonyms one after another, and reads the process, which is the same operation as described above, so the explanation will be omitted. If the desired word is found in step 38, the input key on the keyboard section 1 is pressed down to input this word into a predetermined device. If the inertia value A is 1, the operation ends here (step 41 → step 63),
If A≠1, words with the same heading are read out from the appearing word memory by the appearing word memory read signal from the comparison unit 6 (step 43), and if there is no word with the same heading, the operation ends. , if there are words with the same heading, the words with the same heading are cleared in the appearing word memory 3 by the appearing word memory clear signal from the comparing section 6, and the operation ends (step 44).
→Step 63).

Although the operation of one embodiment of the present invention has been described above, the present invention is not limited to this one embodiment.

For example, in the explanation, we took up the case of converting kana to kanji, but the converted word represented by kana also includes Roman characters and other languages that are a combination of hiragana, katakana, and alphabetic characters, and is represented by kanji. The converted words include not only homonyms such as Kanji but also synonymous characters, such as foreign words that have the same meaning for one Japanese language.

In addition to the dictionary memory, an occurrence word memory is provided in the embodiment, but the present invention is not limited to this; a learning level area may be provided in the dictionary memory and may also be used as an occurrence word memory. In this case, even if a flag bit is not particularly provided, the same operation as described above can be achieved by treating the learning degree that has become equal to the inertia value A as equivalent to the flag bit "1".

As described above, in the present invention, when there are multiple characters of the second type that correspond to characters of the first type, the number of successive conversion selections (uses) is counted, and the number of times can be arbitrarily determined. By giving the first priority in character display to one second character that exceeds a certain value that can be set, it is possible to improve responsiveness in reducing character conversion time for documents, and to increase character conversion speed. The effect is that it can be done.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of an embodiment of the present invention, and FIG. 2 is a flowchart showing the operation of the embodiment of FIG. In the figure, 1...keyboard section, 2...dictionary memory, 3...occurring word memory, 4...register section, 4
-1, 4-2, 4-3, 4-4...Register, 5
... Adder, 6 ... Comparison section, 7 ... Display section, 11
~23...connection line, 31~63...flowchart steps.

Claims (1)

[Claims] 1. In a character conversion device that converts a first type of character into one of a plurality of characters of a second type corresponding thereto, the first type of character is input as heading data. A conversion instruction signal is generated to instruct one of the plurality of characters of the second type corresponding to this to start conversion, and a character selection instruction signal is generated to select and instruct a desired corresponding character of the second type. an input means; inputting each of the plurality of characters of the second type corresponding to the characters of the first type, with the corresponding first type character as heading data, and inputting the first type of characters among the characters of the same heading data; a first storage means for storing as the first data together with an area for storing a flag instructing readout; and a second type of character most recently selected from among the second type of characters of the same heading data. each,
a second storage means for storing the corresponding first type of character as index data and second data together with number data indicating the number of consecutive selections; first data read from the first storage means; second data read from the second storage means;
a third storage means for storing a preset specific value having a value of 2 or more; and a first storage means for storing a preset specific value having a value of 2 or more;
The data with flags are read from the first storage means first and the characters of the second type are displayed, and when the desired character of the second type is displayed, In response to the supply of the character selection instruction signal, the second type of character being displayed is selected and outputted, and when the first data including the selected second type of character is flagged, the third type of character is When there is second data of the same heading data as the heading data in the storage means, this second data is deleted, and when the first data including the selected second type of character is not flagged, the second data is deleted. 3. When the second data that is the same as the heading data in the storage means and the selected second type of character is read from the second storage means and the number of times data is equal to (specific value - 1), This second data is erased from the second storage means, and the first data of the first storage means corresponding to the erased second data is erased from the second storage means.
A flag is added to the data, and if there is a flag in other first data of the same heading data, the latest flag is made valid and the previous flag is deleted, and when the said number of times data is smaller than (specific value - 1), The number data is incremented by 1 and stored again, and when there is no second type of character that is the same as the selected second type of character in the second storage means or when there is no same heading data, it is selectively output. The number data 1 is added to the first data containing the second type of characters, and if the same heading data exists in the second storage means, it is replaced with this second data, and if it is not, a new one is added. A character conversion device characterized by comprising: comparison control means for storing as second data.