JPH07319864A - Kana-Kanji converter - Google Patents

Abstract

(57) [Abstract] [Purpose] In a kana-kanji conversion device for continuous sentence clause conversion, which is used for personal computers, Japanese word processors, etc., input character strings can be stored without increasing dictionary storage capacity, device complexity, or size. The conversion of Katakana is performed regardless of the presence or absence of the diacritical mark or the delimiter code and the position, and the conversion result in which the masonry mark or the delimiter code is added to an appropriate position can be obtained. When a character string containing a code is input, the character string with the code removed by the code deleting means 9 is sent to the dictionary searching means 10 and a dictionary search is performed. The dictionary storage device 4 has information on the number of long sounds contained in the word, the long sound symbol whose presence / absence is unstable, and the position of the midpoint, and if the information matches the information of the input character string, the long sound or the midpoint is detected according to the information. The assigned conversion result is output.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a kana-kanji conversion device for use in personal computers, Japanese word processors and the like, and more particularly to a kana-kanji conversion device of the continuous clause conversion method including foreign words in katakana notation.

[0002]

2. Description of the Related Art Generally, a kana-kanji conversion device converts a kana character string input from an input device into an appropriate Japanese notation such as kanji, hiragana, katakana, etc. based on information obtained by a dictionary search and outputs it. It is a device that outputs to. There are two types of methods for converting into Japanese notation, single phrase conversion and continuous phrase conversion. The single phrase conversion consists of an independent word such as a noun or adverb, an adjunct such as a particle or auxiliary verb, or an independent word. Separate each phrase and search the dictionary each time, and convert it to a word that has the same reading as the input character string, whereas continuous phrase conversion collectively converts long character strings that are continuously input without being aware of the phrase The algorithm is set in advance based on the rules for connecting independent words and adjunct words, or the connection between adjunct words, and if there is a conversion instruction, it is input according to the algorithm based on the word information obtained by the dictionary search. It automatically divides the character string into bunsetsu and converts it to the kanji notation stored in the dictionary according to the delimiter. It is widely used because the input / conversion operation is smoother than the single bunsetsu conversion. To have.

This conventional kana-kanji conversion device is widely used, and its basic functional block is shown in FIG. Figure 4
In, the input device 1 is a keyboard, and the character string input from this is sent to the input control device 35 and stored. When the input character string is sent through the input control device 35, the dictionary search device 36 searches the dictionary 37 accordingly. As shown in FIG. 5, an example of the dictionary 37 is a kana heading 39 which is a word reading and is used for matching with an input character string, and a notation output as a conversion result when the input character string and the kana heading match. 40, further part-of-speech information 13
The information about the word is described. In the case of FIG. 5, for the input character string "gairaigo", the notation 40 "foreign word" in which the kana heading 39 is "Guairaigo"
Is obtained as the conversion result. The notation obtained by the dictionary search is sent from the dictionary search device 36 to the input control device 35,
The input control device 35 sends and displays the notation on the display device 5, and stores the notation in the storage device 38.

However, in such a conventional kana-kanji conversion device, since a dictionary is searched according to an input character string, especially in katakana words such as foreign words, for example, "transistor / transistor" or "floppy disk / floppy disk". If the presence or absence of a long-sound mark (-) or midpoint (・) is uncertain, as in
Depending on the input character string, it may not be possible to convert it, such as when the dictionary contains only "Transistor" or "Floppy disk", it cannot support the character string "Transistor" or "Floppy disk". It was A countermeasure against such fluctuations in readings and notations is to register all possible readings and notations in the dictionary, but there is a problem that the dictionary capacity is limited.

Therefore, Japanese Unexamined Patent Publication No. 3-40060,
Japanese Laid-Open Patent Publication No. 1-21069 has been proposed. These methods will be described below. FIG. 6 shows the former (Japanese Patent Laid-Open No.
40060) is a functional block diagram. The same keyboard 1, input control device 41, and dictionary search device 3 as in the basic form
In addition to 6, the dictionary 37, the display device 5, and the storage device 38, there are a long sound search device 42 connected to the input control device 41, a long sound search device 42, and a long sound addition device 43 connected to the input control device 41.

The operation of this kana-kanji conversion device is as follows according to the description of the embodiment. When a character string that needs to be converted into katakana notation is input from the keyboard 1, it is sent to and stored in the input control device 41, and a long sound symbol is included in the long sound search device 42. You can find out. If it is included, the character string "Tranj Star" with the long-sound mark deleted is input control device 4
1, and the position of the deleted long sound symbol is notified to the long sound providing device 43. In this case, it is notified that the long sound symbol is the seventh character. Next, when the character string "Tranj Star" in which the long-sound symbol is deleted is sent to the dictionary search device 36, the dictionary search is performed according to it. In the dictionary, as shown in the correlation diagram in FIG. 7, word reading and notation are registered as dictionary reading 44 and dictionary notation 45 in the form of omitting the long syllabary. As a result of the search, the notation 4 of the dictionary having the reading 44 of the dictionary corresponding to the character string in which the long sound mark is deleted from the input reading 46 "Tranj Star"
Since 5 is a “transistor”, the “transistor” is sent to the input control device 41. The sent notation is further transferred to the long sound imparting device 43, and a long sound symbol is added according to the deleted information 47 notified to the long sound imparting device 43, and becomes the long sound imparted notation 48 “transistor” and becomes the input control device 41. Sent to. The input control device 41 sends the “transistor” to the display device 5 and the storage device 38 to display and store the notation.

In this way, by the operations of the long sound retrieval device and the long sound adding device, it is possible to convert Katakana words with or without constant long sound symbols without increasing the dictionary storage capacity.

The latter (Japanese Patent Laid-Open No. 1-21069)
Although the detailed structure is not shown in the publication,
Use a dictionary such as the one shown in Figure 8, and use the midpoints and hyphens (-).
When inputting a foreign word that may contain delimiters such as, the input person's kana character string excluding the middle point and hyphen,
For example, by entering "Floppy Disk",
If you want to display the standard standard notation "floppy disk" and want to use a delimiter other than the midpoint or omit the delimiter, press a specific key (for example, the midpoint key) 1-2. By pressing the button twice, the conversion result of "floppy disk" or "floppy disk" is displayed. This also increases the storage capacity of the dictionary regardless of the presence of separators such as midpoint and hyphen. It is thought that it was intended to obtain an appropriate conversion result without any trouble.

[0009]

In the above-mentioned conventional kana-kanji conversion device, in the former case (Japanese Patent Laid-Open No. 3-40060), it is necessary to increase the number of words registered in the dictionary for conversion of katakana notation words including long sound symbols. It became possible to perform conversion corresponding to this, but in order to add the long punctuation that was deleted once in the conversion process based on the position in the input character string, this method is used in the consecutive phrase conversion. When introduced, when processing a sentence in which Kanji and Katakana are mixed, the number of characters in the input character string and the conversion result are not necessarily the same, so it is not possible to add a long sound symbol at the same position as the input character string. Without being limited to this, the conversion result intended by the user may be significantly different. The latter (Japanese Patent Laid-Open No. 1-211)
No. 069 publication), it is not necessary to increase the number of registered words in the dictionary for conversion of katakana notation words including delimiter codes, and conversion corresponding to that is possible, but the user does not omit the delimiter codes. For example, when the character string "Floppy disk" is input, the dictionary search is performed because the reading 49 that matches "Floppy disk" is not registered in the dictionary of FIG. There is a problem that conversion processing cannot be performed without fail.

Further, in such a foreign word in katakana notation, the above-mentioned long-sound symbols and delimiter codes are mixed, so in order to deal with this, both the former and latter expressions are combined. Therefore, there is a problem that the device becomes complicated and large.

Therefore, the object of the present invention is to overcome the above drawbacks and to convert Katakana words regardless of the presence or absence of a long sound symbol or the presence or absence of a delimiter without increasing the dictionary storage capacity, complicating the apparatus, and increasing the size. Another object of the present invention is to provide a kana-kanji conversion device capable of obtaining a conversion result in which a long sound symbol or a delimiter code is added to a position intended by the user.

[0012]

A kana-kanji conversion device of the present invention searches an dictionary by an input device for inputting a character string, an input control device for transferring a character string or a conversion result, and a dictionary search means. A conversion processing device that creates a phrase candidate and an output device that outputs a conversion result, and the input control device can receive information from the input device and send information to the output device, and the conversion processing device The conversion processing device is connected to the input control device and can access the dictionary storage device.The kana heading that omits the long sound symbol and the delimiter code from the reading of the words and the unstable long sound symbol and the delimiter code are omitted. A notation, a dictionary storage device that stores information about words, a code search device, a code deletion device, and a code addition device that constitute a conversion processing device together with a dictionary search device, The code search means is connected to the input control device, the code addition means and the code deletion means, the code deletion means is connected to the code search means and the dictionary search means, and the code addition means is connected to the input control device and the code search means. It is characterized by doing.

[0013]

Embodiments of the present invention will now be described in detail with reference to the drawings.

FIG. 1 is a functional block diagram showing an embodiment of the present invention. In FIG. 1, an embodiment of the present invention includes an input device 1 such as a keyboard, an input control device 2 connected to the input device 1, and a conversion processing device 3 connected to the input control device 2.
And a dictionary storage device 4 connected to this conversion processing device,
And an output device 5 such as a display device connected to the input control device. The conversion processing device 3 is composed of a code searching means 6, a long sound adding means 7 as a code adding means, a midpoint adding means 8, a code deleting means 9 and a dictionary searching means 10. The input control device 2 is connected to the code retrieval means 6, the long sound imparting means 7, the midpoint imparting means 8 and the dictionary retrieval means 10 of the conversion processing device 3. The code retrieval means 6 is connected to the long sound imparting means 7, the midpoint imparting means 8, and the code deleting means 9. The dictionary search means 10 is connected to the dictionary storage device 4 and is accessible.

The keyboard 1 and the display device 5 are the same as those described in the prior art. The input control device 2 stores the character string input from the keyboard 1 and stores it in the code search means 6
And receives the information obtained as a result of the dictionary search from the dictionary search means 10, and also sends the information to the long sound imparting means 7 and the midpoint adding means 8 to receive and display the notation with the long sound symbol and the midpoint. It is sent to the device 5 and displayed.

The code retrieving means 7 of the conversion processing device 3 retrieves the input character string sent from the input control device 2, and if a long sound symbol is included therein, the total number of the long sound symbols included therein is calculated. The long sound imparting means 7 is notified, and when the midpoint is included, the midpoint imparting means 8 is operated to set a flag and the input character string is sent to the code deleting means 9. The code deleting means 9 sends to the dictionary searching means 10 a character string in which the long-sound symbols and midpoints have been deleted from the input character string sent from the code searching means 6. The long sound adding means 7 adds long sound symbols to the notation based on the information sent from the long sound searching means 6 and the input control device 2.
The midpoint assigning means 8 sets a flag indicating that the input character string includes the midpoint when the midpoint assigning means 8 receives an action from the code retrieving means 6, and based on the information sent from the input control device 2, the midpoint adding means 8 Points are given and sent back to the input control device 2. The dictionary search means 10 searches the dictionary storage device 4, and sends to the input control device 2 the notation and word information of a word having a kana heading that matches the character string sent from the code deletion means.

A dictionary is stored in the dictionary storage device 4 in the same manner as that described in the prior art, the contents of which are shown in FIG.
In addition to the kana heading 11 in the form of removing the long sound mark and the midpoint from the notation of the word, the notation 12 of the word including the long sound symbol with stable presence / absence, the part-of-speech information 13, and may be included in the word. If the number of long-sounds that indicates the maximum number of long-sounds is 14 and long-sounds that may be included are not constant, the long-sounds information 15 that indicates the position in the notation of the long-sounds, and the midpoint is The midpoint information 16 indicating the position in the notation of the midpoint when included is described. In FIG. 2, since the “keyboard” does not include unstable long-sound symbols or midpoints, the number of long-sounds is 2, and long-sound information and midpoint information are undefined.
Considering that "computer" is also called "computer", the maximum number of long sounds that can be included is 2, so the number of long sounds is 2, and the presence or absence of the long sound symbol in the last 7th character is unstable. The information is 7, and since the midpoint is not included, the midpoint information is undefined. "Eleninho phenomenon"
, There is no unstable long-sound symbol, and the long-sound information is undefined, and the long-sound information is undefined, and the mid-point information is 3 because the notation "El Niño phenomenon" in which the mid-point is included in the third character is also considered. Becomes

Next, the operation of the kana-kanji conversion device will be described with reference to the flowchart of FIG. As an example, assume that the character string "konpyuta" is entered. When a character string is input (17), it is sent to the input control device 2 and stored (18), and sent to the code retrieval means 6. The code retrieving means 6 checks whether or not the sent "konpyu-ta" contains a long sound symbol or a midpoint (20), and if the long sound symbol is included (21), the total length is given to the long sound adding means 7. Notify (22). In this case, it becomes 2. If the midpoint is included (23), the midpoint assigning means 8 is operated to set a flag (24). However, in this case, the flag is not included, so the process is not performed. When the above processing is completed, the code searching means 6 sends the input character string to the code deleting means 9, deletes the long sound symbol and the midpoint from the input character string, and sends the character string "Konpyuta" to the dictionary searching means 10 ( 25). When the dictionary search means 10 receives a character string, the dictionary storage device 4 follows the character string.
The dictionary stored in is searched (26). In this case, since the word having the kana heading “Computer” is registered, the notation “computer” of the word, the part-of-speech information, the number of long sounds and the long sound information are also read and sent to the input control device 2. When the input control device 2 receives the information, it checks whether the long sound adding means 7 has information on the total number of long sounds in the input character string sent from the code searching means 6 (27). Send the information received from.
The long sound adding means 7 determines the total number of long sounds in the input character string from the code searching means 6, the number of long sounds written in the dictionary sent from the input control device 2, and when the input character string consists of a plurality of words. It is checked if the sum of the numbers of long sounds of those words is the same (28), and in this case, there is a match, so if there is long sound information (29), the long sound symbol is added to the notation and the notation "computer" is added to the input control device. Send it back to 2 (30). Next, the input control device 2 checks whether the flag is set in the midpoint assigning means 8 (31) If the flag is set and the midpoint information is described in the dictionary (32), the midpoint is set according to the midpoint information. However, since the flag is not set in this case, the midpoint is not given, and the conversion result "computer" is sent to the display device 5 and output (34). As a second example, the same processing is performed when the character string "konpyuta no kibodou" is input, and the long sound adding means 7 describes the total number of long sounds in the input character string and the dictionary. The numbers of long notes are compared, but in this case, the total number of long notes in the input character string is three. On the other hand, as a result of the dictionary search, the information of the two words “computer” and “keyboard” is read out, and the total number of long notes is 4 because the number of long notes of “computer” is 2 and the number of long notes of “keyboard” is 2. , Does not match the total number of long sounds in the input string. Therefore, the long sound is not added, and the conversion result of "computer keyboard" is output.

As a third example, it is assumed that a character string "Er.
The input character string is sent (17) to the input control device 2 and stored (18), and the code search means 6 searches for the total number of long-sound symbols and the presence or absence of a midpoint (20). In this case, since the total number of long notes is 1 (21), the long note providing means 7 is notified that the total number of long notes is 1 (22), and since the midpoint is also included (23), the midpoint providing means 8 is operated. Set a flag (24). Then, the input character string is sent to the code deleting means 9 to delete the chords and midpoints (24), and the character string "Eni Niyongensho" is sent to the dictionary searching means 10 for dictionary search. (26) The result is the input control device 2
Sent to. If the total number of long sounds in the input character string is notified to the long sound providing unit 7, the input control device 2 sends the search result to the long sound providing unit 7 (27). In this case, the total number of long notes and the number of long notes match (28), but since there is no long note information (29), the addition is not performed. Next, the input control device 2 checks whether the flag is set in the midpoint giving means 8 (31), and if it is set, sends the notation and the midpoint information to the midpoint giving means 8. The midpoint giving means 8 gives a midpoint to the notation according to the midpoint information (33). As a result, the notation "El Niño phenomenon" is sent back to the input control device 2 and output as a conversion result to the display device (34). In this way, Katakana can be converted regardless of the presence or absence of a long sound symbol or the presence or absence of a midpoint without increasing the dictionary storage capacity, complicating the device, or increasing the size of the device. It is possible to obtain the conversion result to which the middle point is added.

In the above description, the middle point is used as the delimiter code as an example, but the delimiter code is not limited to this, and other delimiter codes such as hyphen (-) and slash (/) can be used as the code search means and the code. The same effect can be obtained and the object of the present invention can be achieved by making the adding means compatible with each delimiter code.

[0021]

As is apparent from the above description, the kana-kanji conversion device of the present invention increases the dictionary storage capacity by performing the code search by the code search means and the code deletion by the code deletion means. It is possible to convert Katakana words with or without long-sound symbols or with or without separators, and to obtain the conversion result with long-sound symbols or separators added at the user's intended position without the need for complicated or large size The effect can be obtained.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a diagram showing dictionary contents according to an embodiment of the present invention.

FIG. 3 is a flowchart of an embodiment of the present invention.

FIG. 4 is a functional block diagram of a basic form of conventional technology.

FIG. 5 is a diagram showing a dictionary content of a basic form of a conventional technique.

FIG. 6 is a functional block diagram of an example of a conventional technique.

FIG. 7 is a correlation diagram of an example of a conventional technique.

FIG. 8 is a diagram showing a dictionary content of a second example of the related art.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 keyboard 2 input control device 3 conversion processing device 4 dictionary storage device 5 display device 6 code search means 7 long sound adding means 8 midpoint adding means 9 code deleting means 10 dictionary searching means 11 dictionary kana index 12 dictionary notation 13 dictionary Part-of-speech information 14 Number of long sounds in dictionary 15 Information on long sounds in dictionary 16 Midpoint information of dictionary

Claims (1)

[Claims] 1. An input device for inputting a character string, an input control device for transferring a character string or a conversion result, a dictionary storage device, and making the dictionary storage device accessible so that a dictionary is searched to find phrase candidates. The input control device is capable of receiving information from the input device, sending information to the output device, and accessing the dictionary storage device. In a certain kana-to-kanji conversion device, a kana heading that omits long-sound symbols and delimiters from the reading of words in advance, a notation that excludes unstable long-sound symbols and delimiters, and information about words are stored. A kana-kanji conversion device characterized in that the storage device is searched by a dictionary search means, and is provided with a code search means, a code deletion means, and a code addition means for a long sound and a midpoint. Place