JPH05314133A - Kanji text input device - Google Patents

Abstract

(57) [Summary] [Structure] Input the reading character string into the dictionary (10
2) Create a part-of-speech sequence (103). If there is an undetermined part-of-speech, a pattern candidate is created (121), it is matched with a typical part-of-speech pattern having priority information such as frequency information (124), and if matched, the pattern candidate is prioritized with the matched typical pattern. It is saved together with the information (125), and the parts of speech are stored in the part-of-speech order storage table in the order of higher priority patterns. The part-of-speech of the first part of the part-of-speech order storage table is set as the determined part-of-speech, and among the kanji of the part-of-speech, the first-ranked kanji is displayed. When the kanji is not the target kanji and the kanji is changed and input by moving the cursor, the kanji of the part of speech of the kanji is displayed in order of priority. [Effect] It is possible to efficiently perform the conversion work of the same kanji and different characters into the target kanji character string.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a text input device,
In particular, it relates to a device for selecting a desired kanji from homophones.

[0002]

2. Description of the Related Art Regarding the selection of homophones and kanji, the most recently used kanji is selected (Japanese Patent Publication No. 58-39334).
Or a word is given a priority, and if a homonym occurs, it is checked whether or not a word with a priority is present in the storage means, and if there is, a homonym with a priority is given. There is a selection method (Japanese Examined Patent Publication No. 58-11928), but there is a problem that it is often necessary to change the page of the kanji option display many times in order to obtain the desired kanji for kanji with many homophones.

[0003]

However, in the case of kanji with many homophones and different characters, it is often necessary to change the page display of the kanji choices many times in order to obtain the desired kanji, which is troublesome. In some cases, homomorphic characters have similar character forms, often overlooking the desired Kanji, which is a cause of double labor.

The present invention has been made for the purpose of facilitating selection of homophones and kanji.

[0005]

SUMMARY OF THE INVENTION In order to achieve the above object, the present invention provides means for narrowing an input reading character string to a highly likely part-of-speech sequence and homophones in order of most likely part-of-speech. It is to provide a means for grouping and displaying and a means for updating the part of speech.

[0006]

[Function] The input phonetic character string is searched for in a dictionary to create a part-of-speech string. If there is an undetermined part of speech, create a pattern candidate, match it with a typical part-of-speech pattern that has priority information such as frequency information, and if it matches, save the pattern candidate together with the priority information of the matched typical pattern, and give priority. Select the part of speech in the order of the highest ranking pattern. The selected part-of-speech name and the kanji of that part-of-speech are grouped and displayed. If there is a target Kanji in the part of speech, the priority of the typical part-of-speech pattern is updated. If it is not the displayed part of speech, the part of speech is updated.

[0007]

1 to 3 are flowcharts of the whole processing of the present invention, and FIG. 4 is a block diagram of a kanji sentence input device. In FIG. 4, 1 is a display device, 2 is a keyboard for input, 3 is a main memory, 4 is an analysis processor,
Reference numeral 5 is a document file. Reference numeral 6 is a term dictionary file, which is made up of records in which reading character strings correspond to Chinese characters. A part-of-speech pattern file 7 stores typical part-of-speech patterns. Reference numeral 8 denotes a part-of-speech pattern matching memory, which includes an 81 input table, an 82-dictionary result storage table, an undecided part-of-speech storage table 84, a pattern candidate storage table 84, and a part-of-speech pattern comparison result storage table 85.
Consists of. Reference numeral 9 denotes a part-of-speech display order control memory, and 91
It includes a part-of-speech order storage table and a 92 default part-of-speech order table.

FIG. 5 shows an example of a dictionary record. A record consists of a phonetic string (pinyin notation), a kanji character string, and a priority. FIG. 6 shows an input table. FIG. 7 is a dictionary search result storage table. FIG. 8 is an undecided part-of-speech storage table.

FIG. 9 shows a pattern candidate table.

FIG. 10 shows a part-of-speech pattern table. The part-of-speech pattern table consists of typical part-of-speech patterns and the priority of each pattern.

FIG. 11 shows a part-of-speech pattern matching result storage table. The part-of-speech pattern matching result storage table also has a pattern priority.

FIG. 12 shows a part-of-speech order storage table.

FIG. 13 is a default part-of-speech order table. An asterisk is added to the part of speech that was in the part-of-speech order storage table in the displayed identification part of the default part-of-speech order table.
(*) Is displayed.

FIG. 14 shows an example of a Kanji selection screen. The part-of-speech name corresponding to the part-of-speech update key is displayed on the first row.

FIG. 15 shows a part-of-speech update key and a Chinese character selection key. The first step corresponds to the part-of-speech update key, the second and third steps correspond to the kanji selection key, and the fourth step corresponds to the option screen return key and the page break key.

FIG. 16 shows an example of updating the part of speech on the kanji selection screen.

FIG. 17 is an example of a page break of the Chinese character selection screen.

Next, the processing method of the present invention will be described with reference to the flowcharts shown in FIGS. First, input the reading character string input from the input device 1 (101),
When the conversion instruction key is pressed, the dictionary is searched (10
2) Create a pattern that is a part-of-speech sequence (103). When there is only one corresponding part-of-speech for each input reading word and there is no undetermined part-of-speech in the pattern, the kanji character string is displayed using the kanji in the first place of the group of the corresponding determined part-of-speech of the homophone kanji ( 105).

When the displayed kanji is the desired kanji, the process moves to the next input or ends. If the displayed kanji is not the desired kanji and there is a kanji change input by an instruction by moving the cursor (106), the part of speech name of the kanji is displayed and the same kanji of the displayed part of speech is displayed as an option (10).
7). If there is a kanji selection input (108), the kanji change input of another kanji or the next input is entered or terminated. If there is no kanji selection input and a page feed is input (109), the same-sound kanji group on the next page within the same part of speech is displayed (110). If there is a Kanji selection input, change the Kanji of another Kanji, or move to the next input or end.

If there is a part-of-speech update input (111), the part-of-speech order storage table is checked, and if there is an undisplayed part-of-speech, it is displayed in order of priority (112). If there is no part of speech that is not displayed in the part-of-speech order storage table, the part of speech that is not displayed in the default part-of-speech order table is checked and stored in the pattern area of the dictionary result storage table in the default order (113), and the kanji character string is displayed (105). ..

For each input phonetic word, when there are a plurality of corresponding parts of speech and there is an undetermined part of speech in the pattern (10
4) Create a pattern candidate with the undecided part of speech replaced
(121), the pattern candidate is collated with the part-of-speech pattern (124). If it matches the part-of-speech pattern, the matched pattern candidate is saved (125). When all the pattern candidates have been checked, it is checked whether there is a save pattern (126),
If there is a save pattern, the save patterns are sorted in descending order of priority (127), and the corresponding parts of speech are stored in the part-of-speech order storage table in order of priority. The part-of-speech of the first rank of each reading character is set as the determined part-of-speech, and is stored in the pattern area of the dictionary search result storage table (129). If there is no evacuation pattern, the part of speech having the first priority in the reading character corresponding to the undecided part of speech is set as the determined part of speech (140). When the part of speech is determined, the subsequent processing is the same as the processing when there is no undetermined part of speech.

As an example, the Chinese reading character string "x
Consider the case where "uesheng shi ribenren" is entered and the conversion instruction key is pressed.

First, spaces are stored in the input area with spaces as shown in FIG. Referring to the dictionary as shown in FIG.
To create a pattern like. For a phoneme whose part-of-speech is uniquely determined, the Kanji whose reading priority is 1st is stored in the kanji area, and the part-of-speech name uniquely determined is stored in the pattern area. Since xuesheng is uniquely determined to be a noun and ribenren is also a noun, the first kanji “student” and “Japanese” of each reading are stored, and the nouns are stored in the pattern area. If there is a reading whose part of speech is not uniquely determined, it is displayed as it is in the kanji area, and a plurality of parts of speech names are stored in the undetermined part-of-speech candidate storage table.

In the example sentence, there are four types of part of speech corresponding to shi, and a judgment area, a verb, a numeral, and a noun are stored in the pattern area (FIG. 8). A pattern candidate in which the undetermined part of speech is replaced is created in the pattern candidate table. Since there are four types of parts of speech here, there are four types of pattern candidates. (FIG. 9) The pattern candidate is collated with a part-of-speech pattern table (FIG. 10) storing typical part-of-speech patterns. When it matches the part-of-speech pattern, the matched pattern candidates are saved in the matching result storage table together with the frequency. Here, two out of four types of pattern candidates match (FIG. 11). When the collation of all pattern candidates is completed, the saved patterns are sorted in descending order of priority and the corresponding parts of speech are stored in the part-of-speech order storage table. Here, the discriminant and the verb are stored (Fig. 1
2).

The first part of speech in the part-of-speech order storage table is used as the determined part of speech, and the kanji character string is displayed using the kanji in the first part of the part of speech. Here, the decision word is the determined part-of-speech, and the Chinese character “” which is the first rank of the part-of-speech is displayed.

When “” is not the intended kanji, the cursor is moved to “” and kanji options are displayed. At that time,
The part of speech stored in the part-of-speech order storage table is displayed in the part corresponding to the part-of-speech update key. When the number of parts of speech in the part-of-speech order storage table is less than 3, the default part-of-speech order table that has not been displayed is displayed in ascending order of the numbers (FIG. 13). If you want to update the part of speech to a verb, press the numeric keypad 8 in this case (FIG. 15). When the part of speech is updated to the verb, the kanji is selected and the screen changes as shown in FIG. If the input reading is gongchenshi shi xingneng, gongchensh
i is a “process teacher” and is fixed by a noun, and xingneng is also a “performance” and is fixed by a noun. shi is uncertain as in the previous example. The part-of-speech is updated, the target kanji character “trial” is obtained, and the target kanji character string “process teacher trial performance” is obtained.

In this embodiment, the kanji selection screen is formed for each part of speech, but there is also a method of displaying a sign between the part of speech and the part of speech to indicate a group and further reducing the number of page breaks on the kanji selection screen. is there.

In the above embodiment, a dictionary grouped by part of speech is used. However, if the dictionary has the part of speech information, it is not necessary to use the dictionary grouped from the beginning.

In the above-mentioned embodiment, an example in which the parts of speech are grouped and displayed is shown. Similarly, if a dictionary having information on the radicals of the first character is used, the radicals of the first character are grouped and displayed. If a dictionary having voice information is used, the voices can be grouped into four voices and displayed.

[0030]

According to the present invention, since the number of page breaks on the option display screen for homophones and different characters is reduced, it is possible to efficiently perform the conversion into the target kanji character string.

[Brief description of drawings]

FIG. 1 is a flowchart of input processing according to an embodiment of the present invention.

FIG. 2 is a flowchart for connecting to the flow of FIG.

FIG. 3 is another flowchart connected to the flow of FIG.

FIG. 4 is a block diagram of a kanji sentence input device.

FIG. 5 is an explanatory diagram of an example of a record of a dictionary.

FIG. 6 is an explanatory diagram of an example of a record of a table that stores an input result.

FIG. 7 is an explanatory diagram of an example of a record of a table that stores a dictionary search result.

FIG. 8 is an explanatory diagram of an example of a record of a table that stores undetermined part-of-speech candidates.

FIG. 9 is an explanatory diagram of an example of a record of a table that stores pattern candidates.

FIG. 10 is an explanatory diagram of an example of a record of a table that stores typical part-of-speech patterns.

FIG. 11 is an explanatory diagram of a result of matching a pattern candidate with a typical part-of-speech pattern.

FIG. 12 is an explanatory diagram of an example of a record of a table that stores the order of displayed parts of speech.

FIG. 13 is an explanatory diagram of an example of a record of a table that stores a default part-of-speech display order.

FIG. 14 is an explanatory diagram of a screen example of Kanji choices.

FIG. 15 is an explanatory diagram of an arrangement example of a part-of-speech update key and a kanji selection key.

FIG. 16 is an explanatory diagram of a screen example of Kanji choices (after updating the part of speech).

FIG. 17 is an explanatory diagram of a screen example of a kanji option screen (after page break of the kanji character selection screen).

[Explanation of symbols]

11 ... Read character string input, 102 ... Dictionary lookup, 103 ... Pattern creation.

Claims (5)

[Claims] 1. A reading dictionary is converted to a Kanji character string by searching the word dictionary for the input character string, provided with a word dictionary in which the correspondence between the reading character string and the Kanji character string is stored. In the kanji conversion device, a means for quickly selecting a desired kanji from the same-sound kanji, a means for displaying candidate kanji by grouping by part of speech, a means for rearranging the part of speech display in the order of high possibility, and a purpose A kanji sentence input device comprising: means for updating the part of speech when the kanji is not the displayed part of speech. 2. The kanji sentence input device according to claim 1, further comprising means for narrowing down a part-of-speech candidate of a homophone kanji by collating it with a typical part-of-speech pattern. 3. The kanji sentence input device according to claim 1, wherein each radical of the first character of each word is divided into groups. 4. The kanji sentence input device according to claim 1, wherein the kanji sentence is divided into groups of four voices in Chinese. 5. The kanji sentence input device according to claim 1, further comprising means for squeezing a part-of-speech candidate for a homophone and kanji by parsing.