JPS6298455A - Character string processor - Google Patents

Abstract

PURPOSE:To select and display plural converted character strings by storing one outer mark of the second character string, also storing a data for indicating a candidate to be displayed and a data for indicating the like liest candidata among the candidates to be displayed, and generating a candidate to be displayed, from the indicated data. CONSTITUTION:From an input device 2, a reading input which is generated in reaction to a scan of a reading character key is inputted, and also an instruction by a conversion key for instructing a KANA (Japanese syllabary) to KANJI (Chinese character) conversion, and an instruction by a candidate selecting key for selecting one candidate from in plural displayed candidates are inputted. These inputs are sent to an input data buffer through a key input taking-in part 4. A conversion key detecting part 8 detects whether an instruction of the KANA to KANJI conversion by the conversion key exists in an input from the input data buffer 6 or not, and in case when the instruction exists, read-out from a dictionary 12 is executed by a dictionary read-out part 10, and in case when no instruction exists, an input data of a KANA character is outputted to an input data output part 18. The input data output part 18 outputs the input data of the KANA character to a display memory 20.

Description

DETAILED DESCRIPTION OF THE INVENTION Technical Field The present invention relates to a character string processing device, particularly a Japanese word processor, a memory typewriter, an office computer, a translation machine, etc., which converts an input character string into a corresponding character string. This invention relates to a character string processing device that derives and outputs character strings.

Conventional technology For example, in the 1 Kana-Kanji conversion method widely adopted in Japanese word processors, the input sentence 7 and 1
from! There is a method of generating a converted character string at 't\ word middle word middle line sentence m rrt- position. The kanji-kana mixed character string that corresponds to the input reading character string depends on the way the okurikana is added.
There may be multiple types. For example, if you enter "uketsuke" as an input string and convert it to kana-kanji, the candidates for the string containing kanji and kana will be "reception", "reception", and "reception".
There is a "Ukeha Tsuke". In Japanese, okurikana is often used differently in this way, with or without okurigana.

Conventionally, it has been necessary to register all of the candidates for a plurality of character strings containing kanji and kana based on different okurikana in a dictionary stored in internal memory or external storage.

Therefore, there is a drawback that the dictionary has a large capacity.

In order to solve this drawback, it is conceivable to register in a dictionary one notation of a candidate for a plurality of character strings containing kanji and kana, and data for creating other candidates from this one notation. However, in this case, although a plurality of candidates for character strings containing kanji and kana can be displayed, there is a drawback that a learning function for memorizing the most likely candidates cannot be provided.

Purpose The present invention solves these drawbacks, and makes it possible to select and display a plurality of converted character strings without having a large-capacity dictionary, and to display the most likely candidate character string. The purpose is to provide a character string processing device.

Structure In order to achieve the object stated in -1-, the present invention includes an input means for inputting a first character string and an instruction, and a storage means for storing a second character string candidate corresponding to the first character string. , candidate retrieval means for reading out candidates for a second character string corresponding to the first character string inputted from the input means from the storage means, wherein the storage means is configured to retrieve candidates for the second character string corresponding to the first character string inputted from the input means; The device further stores one notation read from the storage means and data instructing a candidate to be displayed and data instructing a most likely candidate among the candidates to be displayed. Candidate processing means for creating candidates to be displayed from data instructing candidates to be displayed, and most likely candidate processing means for creating the most likely candidate among the candidates created by the candidate processing means from data instructing the most likely candidate. It is characterized by having the following.

Hereinafter, one embodiment of the present invention will be described in detail.

Referring to FIG. 1, an embodiment in which the present invention is applied to a kana-kanji conversion processing device is shown in a conceptual functional block diagram.

As shown in the figure, this embodiment has an input device 2 such as a keyboard. In this embodiment, the input device 2 includes a character key for inputting a kana character, a conversion key for instructing display of a candidate kanji obtained as a result of kana-kanji conversion, and a key for selecting one of the displayed candidate kanji. and a candidate selection key for selection.

The device also has a display 24, which can be displayed, for example by C.
An RT character display device may be used, and the reading character string input into this system, the character string resulting from the kana-kanji conversion, the selected candidate kanji, etc. are visually displayed.

The device also has a dictionary 12, which is a file storage device in which conversion dictionary data to be indexed during kana-kanji conversion is stored, such as an internal memory such as RAM or ROM, a fixed disk, or a floppy disk. External storage devices such as disks may be used. Elements other than item 1 shown in FIG. 1 are constructed by a processing system such as a microprocessor, and are shown in blocks according to their functions in the figure.

From the input device 2, the reading input generated in response to the scanning of the reading character key is input, as well as an instruction using the conversion key to instruct kana-kanji conversion, and one candidate from among the displayed candidates. An instruction using a candidate selection key to select is input. These inputs are sent to the input data buffer γB through the key input input unit 4.

The conversion key detection section 8 detects whether or not there is an instruction for kana-kanji conversion using a conversion key in the input from the input data puffer 6, and if there is an instruction, the dictionary reading section IO sends the dictionary 12 to the dictionary reading section IO.
If there is no instruction, manual data of kana characters is output to the input data output unit 18. The human data output unit 18 outputs input data of kana characters to the display memory 20.

When the conversion key detection unit 8 detects a kana-kanji conversion instruction, the dictionary reading unit 10 generates an internal address, that is, a physical address, of the dictionary 12 based on the reading data from the input data packer 6, and inputs the internal address to the dictionary 12. It requests the search for the kanji corresponding to the reading data.

In the dictionary 12, as shown in FIG. 2, at the storage location of each address, there is one representation of a plurality of candidates for the kanji and okurikana corresponding to the reading, a flag for displaying other representations, and the most likely representation. A flag has been registered to display candidates.

For example, for the reading ``uketsuke,'' there are three types of okurikana: ``ukewatsuke,''``reception,'' and ``reception.'' However, as shown in Figure 2, the dictionary 12 has the reading ``uketsuke,'' and the above. Of the three types, the notation “Ukeha Tsuke”,
For example, a 5-bit flag, part-of-speech information, etc. for displaying other notations and the most likely candidates are registered.

The 5-bit flag consists of a 3-bit flag for creating another notation from the notation "receiving" and a 2-bit flag for indicating the most likely notation from among multiple notations.

In the dictionary 12, instead of registering all of the three types of okurikana listed in 1., in this embodiment, only the notation ``Uke wa Tsuke'', which has the most okurigana, is registered, and the other okurikana are - It is indicated by the 3-bit flag shown below.

The display using 3-bit flags is shown in Table 1, for example.

Table 1 When the 3-bit flag bito~2 is all 0, it is displayed as shown. In other words, it is displayed as "Uke wa Tsuke". When bitO of the 3-bit flag is 1, all okurikana are added as shown, that is, "Uke wa Tsuke" is displayed. When bitl is 1, all the okurikana between the blank characters are removed, that is, "reception" is displayed.

When bit 2 is 1, all okurikana characters are removed, that is, "reception" is displayed.

Therefore, it is displayed as shown in Table 2, for example.

Table 2 When bits O to 2 are all l, all of the three types of okurikana described above are displayed as candidates. In other words, ``Uke wa Ikeke'', ``Reception'', and ``Reception'' are displayed. Also, bi
When tO is O and bits 1 and 2 are 1, candidates are displayed that have okrigana between kanji characters and those that have all okurikana.

That is, "reception" and "reception" are displayed as candidates.

The okurikana processing section I4 performs okurikana processing to create one or more okurikana from one notation read from the dictionary 12 and the 3-bit flag, and sends them to the candidate output section 16. The candidate output unit 16 is the okurikana processing unit 14
The candidates for the okurikana sent from are sent to the display memory 20.

The most likely candidate processing unit 26 displays the most likely candidate from the plurality of okurikana notations created in the okurikana processing unit 14 using a 2-bit flag. The display by the flag of 2-pitsu I is as shown in Table 3, for example.

Table 3: When the 2-bit flag bits 3 and 4 are both 0, the most likely candidate is not displayed. In other words, the learning function that is the most likely candidate is not displayed.

When bit 3 is 1 and bit 4 is 0, the most likely candidate is displayed with all okurikana added, that is, ``Uke wa Tsuke''. When bit 3 is 0 and bit 4 is 1, all okurigana between kanji are removed, that is, "
"Reception" is displayed as the most likely candidate.

When bit 3 and bit 4 are both l, the most likely candidate is displayed with all the okurikana characters removed, that is, "reception". In this way, the most likely candidate is displayed using a 2-bit flag.

Therefore, if the 5-bit flag is, for example, bit 43210 = 01111, "Ukeha I4ke", "Reception", and "Reception" are all displayed as candidates, and "Ukeha Tsuke" is displayed as the most likely candidate. select.

The most likely candidate processing unit 26 processes the sending kana processing unit 14 in this way.
Processing is performed to display the most likely candidate from the plurality of okurikana notations created in , and the most likely candidate is sent to the most likely candidate output section 28 . The most likely candidate output unit 28 sends the most likely candidate for the kana to be sent from the most likely candidate processing unit 26 to the display memory 20.

The candidate selection key detection unit 30 detects whether or not there is an instruction to select a candidate using a candidate selection key in the input from the input data buffer 6, and if there is an instruction to select a candidate using a candidate selection key, the Among the kana candidates, candidates other than the one selected by the candidate selection key are deleted, and the candidate selected by the candidate selection key is registered in the dictionary 12 as the most likely candidate. That is, the 2-bit flag marked "-" is rewritten.

The display memory 20 stores input data from the input data output section 18, data from the candidate output section 16, and data from the most likely candidate output section 28, and sends them to the display device 24 through the display circuit 22.

Explain the operation.

The reading is input from the input device 2. For example, "uketsuke" is input. This input is sent from the key manual input section 4 to the input data buffer 6. The input data from the input data buffer 6 is detected by the conversion key detection unit 8 as to whether or not there is an input using the conversion key.If there is no input using the conversion key, the input data is input data, that is, "receive".
is sent to the input data output section 18. Input data output section 18
The input data "receipt" sent to is stored in the display memory 20.

When the conversion key detection unit 8 detects that there is an input using the conversion key, the input data is sent to the dictionary reading unit 10, and the dictionary reading unit 10 generates a kanji corresponding to the reading data input from the dictionary 12. I do. As described above, one type of notation and the 5-bit flag information tυ are read from the dictionary 12. In this example, the notation "Uke wa ikike" and 5-bit flag information are read out.

Of the 5-bit flags, 3-bit flags indicate data for creating okurikana candidates including other notations, and 2-bit flags indicate, for example, the most recently used okurikana. Information to be displayed as candidates is shown.

1 read out from the dictionary 12 by the dictionary reading unit 10
The notation of the type and the 3-bit flag are shown in the okurikana processing section 1.
4, and is sent to the display device 24 in the sending kana processing section 14.
Candidates for the okurigana notation to be displayed are created. This okurikana processing will be explained with reference to the flowchart shown in FIG. 3A.

Detect whether or not all three bits bits ~ 2 of the flag are O (1oo), and if all are 0, output as indicated, that is, "acceptance accepted" as a candidate to the candidate output unit IB. 114), the sending kana processing ends. If any of the three bits (bits Q~2) is not O, judge whether the first bit O is 1 or not, and if it is l, add all the okurigana as shown, That is, the display ``Reception is accepted'' is output as a candidate to the candidate output unit 16 (104). If it is not 1, do nothing and proceed to step 106.

Next, it is judged whether the second bitl is 1 or not (10f()). If it is l, the candidate output section 1B takes all the kana between the kanji characters, that is, the display of "reception" as a candidate. (ioa). If it is not 1, do nothing and proceed to step 110. Next, determine whether the third bit bit2 is 1 (110)
, 1, the one with all the okurikana removed, that is, the display of "reception" is output as a candidate to the candidate output unit 16 (112). ■If not, do nothing and end the Okukana processing.

In this way, candidates for displaying okurigana are created from one notation registered in the dictionary 12. The candidate output unit 16 sends this display to the display memory 20. The display circuit 22 displays one or both of reading input data stored in the display memory 20, that is, a display of "uketsuke" and a display of okurikana candidates, for example, a display of "reception."
4, and the data sent from the display memory 20 is displayed on the display device 24.

When displaying all candidate okurikana in the area (not shown) for displaying kanji candidates on the display device 24, a plurality of okurikana candidates are displayed as described above. For example, if all 3 bits) bitO~2 are set to 1, "Uke is attached"
Three types of candidates are displayed: "Reception" and "Reception". If there are only two types of kanji, for example, bit.
Two types of okurigana can be displayed by setting O and 1 to 1 and bit 2 to O.

Next, the process of selecting the most likely candidate will be explained with reference to the flowchart shown in figure f53B.

It is determined whether or not the two bits 3 and 4 of the flag are both O (11B), and when both are O, none of the candidates is determined as the most likely candidate (118). That is,
Depending on the learning function, steps lOO to 114 in -1-
None of the multiple notation candidates displayed by the processing of the okurikana is selected as the most likely candidate.

When bits 3 and 4 are not O, bit
Please judge whether t3 is 1 and bit4 is 0.
20) When bit3=1 and bit4=0, the one with all the okurikana added is displayed as the most likely candidate (122). When bit 3 = 1 and bit ← is not 0, judge whether bit 3 = 0 and bit 4 = 1 (124), and when bit 3 = 0 and bit 4 = 1, remove all kana between kanji. Displayed as a likely candidate (12B).

When bit3=0 and bit4=1, display the one with all kana removed as the most likely candidate (12B)
.

In this way, the most likely candidate is selected and displayed on the display device 24.

Next, the process when the operator selects a candidate using the candidate selection key will be explained with reference to the flowchart shown in FIG. 3C.

It waits for an input using a key (200), and if there is an input, the candidate selection key detection unit 30 detects whether or not the input is from a candidate selection key (202). If there is no input using the keys, the process returns to step 200 and waits for manual input using the keys again. If the input is not by the candidate selection key, this process ends.

In the case of input using the candidate selection key, the selected candidate is stored in the display memory 20, and other candidates stored in the display memory 20 are deleted (204). Furthermore, it is determined whether the selected candidate has all the okuri kana added (2013), and if it has all the okurigana added, bit 3.4 of the flag is set to bit 3=1. bit4=
Rewrite to 0 (208.). If not all the okurikana have been added, the process proceeds to step 210 without doing anything.

Next, it is determined whether or not the selected candidate takes all the Okurigana between kanji (210), and if it takes all the Okurigana between 9 characters, bit 3 of the flag is selected.
Rewrite 4 to bit3=o, bit4=1 (212
). If all the okurigana between kanji are not included, the process proceeds to step 214 without even asking.

Next, it is determined whether the selected candidate has all the okuri kana (214), and if it has all the okurikana, set bit 3.4 of the flag to bit 3=I,
Rewrite bit 4 = 1 (21 [(). If all okurigana are not removed, do nothing and end.

According to this embodiment, one type of notation and a 5-bit flag are registered in the dictionary 12, and three types of okurikana notations are created and displayed using 3-bit flags among the 5-bit flags. Therefore, the capacity of the dictionary 12 can be reduced compared to the conventional method in which all three types of expressions are registered in the dictionary 12.

Further, the most likely candidate is displayed using a 2-bit flag, and the displayed most likely candidate can be replaced with a recently used one through learning based on the operator's selection.

Effects According to the present invention, when there are multiple candidates for the second character string corresponding to the first character string, the notation for one of them and data for creating other candidates are stored in the storage means. By storing the information in advance, necessary candidates can be selected and created, and the capacity of the storage means can be reduced.

Furthermore, the most likely candidates learned from the data stored in the storage means can be displayed.

[Brief explanation of drawings]

FIG. 951 is a functional block diagram schematically showing an embodiment in which the present invention is applied to a kanji-kana conversion processing device, and FIG.
FIG. 3A is a flowchart showing the processing flow of the okurikana processing in the device shown in FIG. Flowchart Showing Flow of Most Promising Candidate Selection FIG. 3C is a flowchart showing the flow of processing performed by operating the candidate selection key of the device shown in FIG. Explanation of symbols of main parts 210. Input device 801. Conversion key detection unit 10, . . . dictionary reading unit +2. ,,Dictionary 14, ,Forward kana processing unit 1B, ,Candidate output unit 24, ,Display device 2B, ,Most likely candidate processing unit 2B, ,Most likely candidate output unit 30, ,Candidate selection key detection. Department

Claims (1)

[Claims] 1. An input means for inputting a first character string and an instruction; a storage means for storing second character string candidates corresponding to the first character string; and candidate retrieval means for reading out candidates for a second character string corresponding to the first character string from the storage means, wherein the storage means reads one representation of the second character string. and data indicating the candidates to be displayed and data indicating the most likely candidate among the candidates to be displayed;
candidate processing means for creating candidates to be displayed from the notation and data instructing the candidate to be displayed; and creating the most likely candidate among the candidates created by the candidate processing means from the data instructing the most likely candidate. A character string processing device characterized in that it has a most likely candidate processing means. 2. In the device according to claim 1, the first character string is a kana, the instruction is an instruction for converting kana to kanji, and the second character string is a kana character string, and the second character string is a kana character string. A character string processing device characterized in that the character string is a character string. 3. A character string processing device according to claim 2, wherein one representation of the second character string stored in the storage means is a character string with the most okurikana characters. 4. In the apparatus according to claim 3, the data indicating the candidate to be displayed is indicated by a 3-bit flag, and the data indicating the most likely candidate is indicated by a 2-bit flag. A string processing device characterized by: 5. The apparatus according to claim 4, wherein the 2-bit flag indicating the most likely candidate indicates a character string used most recently among the displayed candidates. Column processing device.