JPH0430247A - Automatic unknown word registration system in kana-kanji conversion - Google Patents

Abstract

PURPOSE:To reduce the burden of a user by accessing an unknown word dictionary based on the number of characters from the head of the 'reading' of an unknown word, obtaining the unknown word concerned and transferring it to a KANA(Japanese syllabary)-KANJI(Chinese character) conversion processing mainbody as a candidate character. CONSTITUTION:When the 'reading' of the unknown word which cannot be converted is inputted in the retrieval of a regular dictionary (general dictionary and user dictionary), the 'reading' is transferred to an unknown word processing means 13 and a call request is given. Consequently, the unknown word concerned is fetched from the unknown word dictionary 17 based on (n) characters from the head of the 'reading' (the 'reading' itself can be (n) characters from the head of original reading) and the candidate word can be obtained. Thus, it becomes unnecessary to previously register the word or a short sentence, which are predicted to be frequently used, and it also becomes unnecessary to register the word and the short sentence which is to be used or the used word or the short sentence by interruption work in the middle of document generation. Thus, the burden of the user at the time of input can be reduced.

Description

[Detailed Description of the Invention] [Purpose of the Invention (Field of Industrial Application) The present invention provides automatic unknown word registration in kana-kanji conversion, which is suitable for use in word processors, personal computers, etc. that perform Japanese input through kana-kanji conversion. Regarding the method.

(Prior Art) Conventionally, when creating sentences using kana-kanji conversion, the user has to register words and short sentences that are frequently entered manually in a dictionary by registering preliminary words or short sentences. By doing this, the word (or short sentence)
Words (or short sentences) without having to enter the entire reading.
It is now possible to call. With this method, conventional kana-kanji conversion is designed to improve the efficiency of text creation.

(Problems to be Solved by the Invention) As described above, according to the conventional example, when using the word function or short sentence function, it was necessary to perform an operation for word registration or short sentence registration. Furthermore, when creating various sentences, it is difficult to predict what kind of words (or short sentences) will appear frequently and register the words (or short sentences) in advance. Furthermore, it is possible to register words (or short sentences) while creating sentences, but it is necessary to specify the corresponding words (or short sentences) and "yomi".
I have to stop writing the text.

The present invention has been made in view of the above circumstances, and is an automatic kana-kanji conversion system that makes it possible to automatically register words and short sentences that are not registered in the dictionary, so-called unknown words, without interrupting the writing process. The purpose is to provide an unknown word registration method.

[Structure of the Invention] (Means for Solving the Problems) The present invention provides a general dictionary in which "yomi" and corresponding words are registered in advance, and a user-specified "yomi" and corresponding words or short sentences are registered. The user dictionary that is
An unknown word dictionary in which "Yomi" and corresponding unknown words are registered, and registration, deletion, and deletion of unknown words for this unknown word dictionary.
and an unknown word processing means that calls up an unknown word from this unknown word dictionary, converts the input "yomi" into kanji etc. by accessing the above general dictionary and user dictionary, and stores the conversion result in a kana-kanji conversion output buffer. If the conversion result is confirmed or if it is an unknown word, the "yomi" and the conversion result are passed to the unknown word processing means, and if the "yomi" of an unknown word that cannot be converted is manually written, the "yomi" is reading"
and a kana-kanji conversion processing body that passes a call request to the unknown word processing means, and when an unknown word is determined in the kana-kanji conversion processing, the unknown word and the conversion result are automatically transferred to the unknown word processing means. Register it in the unknown word dictionary, and from the next time onwards, simply enter the n characters from the beginning of the reading of the unknown word (n is a predetermined value), and the previously determined result will be used by the unknown word processing means. It is characterized by being able to be converted.

(Operation) As described above, according to the present invention, there is an unknown word dictionary in which "yomi" and corresponding unknown words are registered, and registration and deletion of unknown words in this unknown word dictionary, as well as this unknown word. An unknown word processing means is provided to retrieve an unknown word from a dictionary,
If an unknown word (unknown word or short sentence) is determined by the kana-kanji conversion process in the kana-kanji conversion process itself,
By passing the reading and the conversion result to the unknown word processing means, it becomes possible to automatically register the confirmed unknown word as a word or short sentence in the unknown word dictionary. Dictionary 1!F) When the ``yomi'' of an unknown word that cannot be converted in a search is input, the ``yomi'' is passed to the unknown word processing means and a call request is made.
Based on the n characters from the beginning of the reading (it may be the n characters from the beginning of the reading itself or the original reading), extract the corresponding unknown word from the unknown word dictionary and create a candidate word. It becomes possible to obtain.

This eliminates the need to pre-register words and short sentences that are expected to be used frequently, and allows you to interrupt work while creating a document. There is no need to register the information, and therefore the burden on the user at the time of input can be reduced. Furthermore, just by entering the first n characters of the "yomi",
Since words or short sentences registered in the unknown word dictionary can be recalled, the efficiency of writing sentences is improved.

(Example) Embodiments of the present invention will be described below using the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention. In the figure, the reference numeral 11 is the main body of the kana-kanji conversion process, and the input "yomi" is converted into the general dictionary 15 and the user dictionary 16.
Convert to kanji or katakana by accessing . The converted results are stored in the kana-kanji conversion output buffer 14. When the main body 11 determines the conversion result, if it is an unknown word, it passes the "yomi" and the conversion result to the unknown word processing unit 13. Also, when converting to kanji, unknown words are
If “Yomi” is input, the main body 11 uses the unknown word processing unit 1.
A call request is made to 3 to obtain candidate characters. Reference numeral 12 denotes a word/short sentence processing section, which accesses the user dictionary 16 to register and call up words/short sentences. This word/short sentence processing section 12 normally performs registration processing based on "yomi" and words, but it also processes the unknown word dictionary 17 via the unknown word processing section 13.
It is also possible to register unknown words in the user dictionary 16. Reference numeral 13 is an unknown word processing section, which processes data based on the data received from the kana-kanji conversion processing main body 11.
Conversion candidate characters for "yomi" input as an unknown into the unknown word dictionary 17 are stored. Reference numeral 15 is a general dictionary, and 16 is a user dictionary, in which "yomi" and corresponding words and short sentences are stored. In the user dictionary 1F16, the word/short sentence processing unit 12 registers and deletes words/short sentences. Reference numeral 17 is an unknown word dictionary in which "yomi" and corresponding unknown words are stored.

The unknown word dictionary 17 is used by the unknown word processing unit 13 to register and delete unknown words.

FIG. 2 is a diagram showing the internal structure of the unknown word dictionary 17 shown in FIG. 1. In the figure, reference numeral 21 is an unknown word dictionary information table, in which information necessary for registering and managing unknown words, such as dictionary size, number of registered unknown words, and number of used blocks, is set. The code 22 is an unknown word reading directory, and the code Bl
As shown in , it consists of an unknown word header, an unknown word block number indicating the registered block of the unknown word, a record having the number of calls, and the "yomi" of the unknown word (referred to as a "yomi record"). The reading records (Bl) are set in the unknown word reading directory 22 in alphabetical order.

If they have the same reading, it will be the one that was recently called and used. The unknown word block number contains a value corresponding to the unknown word registration section empty map table 23 described below. The number of calls refers to the number of times the unknown word corresponding to the "reading" is called and used. The size of the unknown word reading directory 22 is fixed, and the size of the reading record is variable.

Reference numeral 23 denotes an unknown word registration section empty map table, which has the configuration shown by 01 in the figure, and has an unknown word registration section 24 which will be described next.
represents the usage status in bits. When a new unknown word is registered, the bit corresponding to the block of the secured unknown word registration section 24 is turned on. Conversely, if a registered unknown word is deleted, the bits in the corresponding block are turned off. The size of the table 23 is a fixed size depending on the size of the unknown word registration section 24. Note that since the size of the unknown word record registered in the unknown word registration section 24 is variable, the number of blocks is naturally also variable. Therefore, all bits of unused blocks are set to 1.

Reference numeral 24 denotes an unknown word registration section, which consists of records (referred to as "unknown word records") having the structure shown by Dl in the figure. The size of the unknown word registration section 24 is fixed, and the size of the unknown word record is variable.

FIG. 3 is a flowchart showing the operation of the embodiment of the present invention, and shows the processing procedure by the unknown word processing section 13.

In the figure, reference numeral "a" indicates a step for initializing the unknown word dictionary 17, in which necessary tables are initialized and the like. This is a step of registering the "yomi" and unknown words received from the code Sukanji conversion processing main body 11 into the unknown word dictionary 17. In this step b, the unknown word dictionary 17 is first accessed, and empty blocks in the unknown word registration section 24 are found.

If there is a free block, that block is secured and the unknown word is registered. Furthermore, in accordance with the rules of the reading directory 22, the "yomi" and unknown word block number are also registered.

Code C is a step of calling a registered unknown word,
Based on the "reading" received from the kana-kanji conversion processing main body 11, the registered unknown word corresponding to the "reading" is searched for and returned to the kana-kanji conversion processing main body 11. Reference numeral d is a step for deleting registered unknown words from the unknown word dictionary 17, and unknown words whose usage frequency is less than a specified value are deleted from the unknown word dictionary 17.

The operation of the embodiment of the present invention will be explained below.

First, the operation of registering an unknown word in the unknown word dictionary 17 will be explained. If the word/short sentence (word or short sentence) determined by the user's selection operation etc. is an unknown word that is not registered in either the general dictionary 15 or the user dictionary 16, the kana-kanji conversion processing main body 11 converts the “reading” ” and passes the word/short sentence to the unknown word processing unit 13. This process is performed for each confirmed character string. It is well known that an unknown word (a string of characters) can be input by repeatedly inputting a code or inputting one kanji character (single kanji input).

The unknown word processing section 13 checks the length of the "yomi" received from the kana-kanji conversion processing main body 11, and returns an error code to the kana-kanji conversion processing main body 11 if it is shorter than the specified value n. On the other hand, for the above bone, the length of the "reading" taken is the specified value n
If this is the case, the unknown word processing section 13 sequentially refers to the unknown word registration section empty map table 23 of the unknown word dictionary 17 starting from the first bit, and searches for the bi-soto that is in the OFF state. Here, it is either the bit position number of the off bit that was searched for (the bit position number when the bit position of the first bit is 0) or the number of the empty block to be sought (abno). If all the bits in the unknown word registration section free map table 23 are on, there are no free blocks, so an error code is returned to the kana-kanji conversion processing main body 11 without registering the unknown word. In addition, to secure a vacant program, use the process of step d shown in Figure 3 to store unknown words with a specified usage frequency or less in an unknown word dictionary]
This is done by deleting from 7.

Next, the unknown word processing unit 13 updates the reading directory 22 of the unknown word dictionary 17. That is, the unknown word processing unit 13
First, the reading directory 22 is checked to determine the position where the reading record of the unknown word to be registered should be inserted. The reading records are stored in alphabetical order. If the "yomi" is the same, it will be inserted at the end of the same "yomi". The unknown word processing unit 13 is
The reading record after the determined position is moved back by the registration record size, and the reading record of the unknown word to be registered is stored. A 1-byte unknown word header (OXFF) is added to the beginning of this reading record. And unknown word processing unit 13
searches for an empty block in the unknown word registration section 24 of the unknown word dictionary 17 in which the unknown word should be registered. This empty block is the above-mentioned abn where the number of unknown word headers in the unknown word registration section 24 matches the number indicated by the previously calculated empty block number.
This is the o-th position in the unknown word registration section 24.

When the unknown word processing unit 13 searches for an empty block in the unknown word registration unit 24 as described above, it registers the unknown word (word/short sentence) in the same block in the same way as updating the reading directory 22. At this time, the unknown word processing unit 13 also adds a 1-byte unknown word header (OXFF) to the beginning of the unknown word to be registered, and then returns a normal completion code to the kana-kanji conversion processing main body 11 to end the process.

Next, the calling operation will be explained. First, if the word/short sentence corresponding to the input "yomi" is an unknown word that is not registered in either the general dictionary 15 or the user dictionary 16, the kana-kanji conversion processing body] 1 converts the "yomi" It is passed to the unknown word processing unit 13. The unknown word processing unit 13 accesses the reading directory 22 of the unknown word dictionary 17 and extracts n characters from the beginning of the "reading" received from the kana-kanji conversion processing main body 11 (n is a predetermined value, For example, check whether there is a match with 3).

Note that when the number of characters in the "yomi" is n, it is checked from the kana-kanji conversion processing main body 11 whether or not there is a word that matches the "yomi" taken from uke. If there is a match, the part corresponding to the candidate number is extracted. The candidate number means the number of characters to be extracted from the same reading.

If there is no match, the Kana-Kanji conversion processing body 1
A code indicating this is returned to 1 and the process ends.

Now, if the unknown word processing section 13 finds a match with the n characters from the beginning of the "yomi" taken from the kana-kanji conversion processing main body 11, the unknown word processing section 13 selects the corresponding unknown from the unknown word registration section 24 as follows. Extract words and perform kana-kanji conversion processing 11
Return to. That is, the unknown word processing unit 13 extracts the unknown word block number from the reading record obtained above, and determines its position in the unknown word registration unit 24 in the same manner as at the time of registration. Then, the unknown word processing section 13 accesses the corresponding block of the unknown word registration section 24 corresponding to the candidate number and extracts the designated unknown word.

For example, if the unknown word "personal computer" is confirmed and registered in the unknown word dictionary 17 through the above unknown word registration, from now on, if you enter "ha-so" and convert it (n -3), the word "personal computer" will appear. Similarly, for example, if an unknown short sentence such as ``Thank you for your continued kindness'' is confirmed and registered in the unknown word dictionary 17, from now on, if you enter ``Maido'' and convert it,
"Thank you for your continued generosity," a short message began to appear. In addition, in the above example, if the word "personal computer" is registered in the dictionary as "basonaru computer", the first three characters "basoj" The word dictionary 17 is searched and converted to "personal computer", but if the pronunciation of this conversion result matches the input pronunciation, the remaining pronunciation "Narukonputa" is ignored. .

[Effects of the Invention] As explained above, according to the present invention, there is no need to register words/short sentences that are likely to be used in advance, and it is not necessary to register the words/short sentences that are likely to be used by interrupting the work in the middle of writing a sentence. Since there is no need to do this, the burden on the user can be reduced. Also, just by entering the first n characters of the "yomi",
Since words/short sentences registered in the unknown word dictionary can be recalled, the efficiency of writing sentences is improved.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a diagram showing the internal structure of the unknown word dictionary shown in FIG. 1, and FIG. 3 is a flowchart showing the processing procedure by the unknown word processing section. It is. 11...Kana-Kanji conversion processing main body, 12-word/short sentence processing section, 13...unknown word processing section, 15...general dictionary, 16
...Unisa Dictionary, 17...Unknown word dictionary.

Claims (1)

[Claims] There is a general dictionary in which "yomi" and corresponding words are registered in advance, a user dictionary in which user-specified "yomi" and corresponding words or short sentences are registered, and "yomi" and corresponding unknown words are registered. an unknown word dictionary, an unknown word processing means for registering and deleting unknown words in this unknown word dictionary, and calling up unknown words from this unknown word dictionary; Convert it to kanji etc. by accessing the dictionary, store the conversion result in the kana-kanji conversion output buffer, and when the conversion result is confirmed and it is an unknown word, the ``yomi'' and the conversion result are sent to the above unknown word processing means. and a kana-kanji conversion processing body that passes the reading of an unknown word that cannot be converted to the unknown word processing means to issue a call request when the reading of an unknown word that cannot be converted is input, and the unknown word processing means If the above Kana-Kanji conversion processing body passes the conversion result “Yomi”, it performs processing to register the unknown word in the above unknown word dictionary, and then passes the “Yomi” of the unknown word from the above Kana-Kanji conversion processing body. If a call is requested, the unknown word dictionary is accessed based on the n characters from the beginning of the reading (n is a predetermined value), the corresponding unknown word is obtained, and the above is retrieved. An automatic unknown word registration method in kana-kanji conversion characterized by passing candidate characters for "yomi" to the above-mentioned kana-kanji conversion processing main body.