JPS5831434A - Japanese sentence preparing device - Google Patents

Abstract

PURPOSE:To increase the probability of displaying only one ''Kanji'' (Chinese character) by input data by using the phonetical reading and Japanese pronounciation of the ''Kanji'' and the reading of the Chinese compound word including the ''Kanji'' as the input data for indexing the ''Kanji'' CONSTITUTION:A ''Kana'' (Japanese syllabary) code inputted by a ''Kana'' code inputting device 1 is temporally stored by a temporally storing part 2 and converted into ''Kana'' and ''Kanji'' mixing description by a ''Kana''/''Kanji'' converting part 3. An indexing part 6 retrieves Chinese compound words coincident with a ''Kana'' code string by using a word dictionary 4 and a ''Kanji'' dictionary 5 and output a ''Kanji'' code included in both the dictionaries in common. An output display part 7 displays the outputs of the ''Kana''/''Kanji'' converting part 3 and the indexing part 6 as ''Kanji'' description.

Description

[Detailed Description of the Invention] The present invention converts a Japanese sentence input in kana code into a sentence mixed with kana and kanji that is appropriate for Japanese and outputs it, or inputs multiple kana code strings and outputs the resulting sentence. 0Concerning a Japanese sentence creation device that outputs kanji specified simultaneously by
A well-known method is to convert this into a sentence mixed with kana, kanji, and kanji that is appropriate for Japanese and output it. At this time, if you make a mistake in entering the kana code, or if you enter the pronunciation of a word that is not registered in the word dictionary for kana-kanji conversion, an undesired kanji will be output and displayed, and the kanji displayed in the output will be the one you want. It is necessary to correct the kanji to the correct kanji. Generally, in order to correct the kanji, a kanji dictionary is set up that stores the correspondence between the kanji code for the onyomi and kunyomi of the kanji, and the onyomi and kunyomi of the kanji are input in kana code, and the characters specified by these are input. A method is used in which a group of kanji characters is displayed all at once on a display or the like, and the operator selects one of the kanji characters.

However, according to the above method, the kanji dictionary becomes huge, and there are many corresponding kanji when specifying on-yomi and kun-yomi, and the burden on the operator to select one from the group of kanji displayed in the output is heavy. There is something.

Also, instead of having a sound 1tll + dictionary, we have an idiom dictionary,
A method has been proposed in which, when specifying a certain kanji, the pronunciation of an idiom containing the kanji is input, and after the idiom is output and displayed, a desired i-character position in the idiom is selected. However, this method also requires two operations to specify a kanji: inputting the reading and selecting it, and has the disadvantage that it is very difficult to use when correcting kanji.

In view of the above-mentioned drawbacks, the present invention provides a Japanese sentence creation device that reduces the size of a kanji dictionary for searching kanji characters and reduces the burden on the operator for kanji indexing. The present invention will be described below along with one embodiment thereof.

In FIG. 1, 1 is a kana code input device such as a kana keyboard. Reference numeral 2 denotes an input input character code temporary storage unit that temporarily stores the character code input by the character code input device 1. Reference numeral 3 denotes a kana-kanji conversion unit which converts the kana code string inputted by the kana code input device 1 into a proper kana-kanji mixed-shi notation as a Japanese sentence. 4 stores kana codes indicating the pronunciation of words such as Japanese pronunciations and nominals, kanji codes indicating kanji notation, codes indicating parts of speech, etc., and the kana-kanji conversion unit 3 performs appropriate conversion as Japanese. This is a word dictionary that searches for useless words. Reference numeral 5 denotes a kanji dictionary in which kana codes and kanji codes representing the phonetic readings of kanji used in Japanese sentences are stored in correspondence. Reference numeral 6 is an index section that searches for kanji with on-yomi or kun-yomi that matches the kana code string input from the kana code input device, or compound words with the reading that matches the input kana code string, using the word dictionary 4 and the kanji dictionary. 6 and outputs the kanji code that is commonly included in all of these. 7 outputs and displays the outputs of the kana-kanji converting part tan and the indexing part 6 as kanji notation. For example, it is an output display section of a display device or the like. 8 is a control unit that controls the operation of the entire Japanese language creation device, and 9 is a control unit 8 that is controlled by keystrokes.
This is a function key that instructs the operation of the

Next, the detailed configuration of the kana-kanji converter 3 will be explained using FIG. In the figure, reference numeral 10 denotes an independent word dictionary 4 that searches for an independent word (nominal or disjunctive) that has the same pronunciation as a partial sequence from the beginning of the kana code string stored in the input kana code temporary storage unit 2. This is the word search section. 12 is a conversion candidate temporary storage unit that temporarily stores independent word candidates, which will be described later;
Reference numeral 13 denotes an adjunct word search unit that searches the adjunct word dictionary 14 for an adjunct word of an independent word candidate, and reference numeral 15 denotes a connection inspection unit that examines the search results of the adjunct word search unit 13.

Next, the index portion 6 will be explained in detail with reference to FIG. In the figure, 16 is an input kana code temporary storage section 2.
A kana code that divides the kana code string stored in by the blank code described later.

The division section 17 is a syllable number calculation section that calculates the number of syllables of the kana code divided by the kana code division section 16.

A syllable is the sound or most specific unit of a single Japanese word, and the Japanese sound is Kiyon (power, ki,).

O19 is a search control unit that determines whether to search for the input kana code in the word dictionary 4 or the kanji dictionary 6 based on the calculated number of syllables. Kanji candidate temporary = self-memory section that temporarily stores candidates, 2゜ is a match practice section that searches for kanji codes, which will be described later.Next, regarding the operation of the Japanese sentence creation device with the above configuration, (1) Japanese sentence ■ When creating a ``Japanese character'' and when inputting kanji, we will explain this separately.

(1) When creating Japanese text, first press the function key 9 to send a command to the control unit 8 that the kana code input from the kana code input device 1 corresponds to the reading of the Japanese sentence. . Therefore, the control section 8 puts the kana-kanji conversion section 3 into an operating state and then performs the following steps: 1. A kana code is input using a kana code entry device 1, and the input kana code is temporarily stored in a large sword kana code temporary storage section 2 until the input is completed.

Meanwhile, the independent word search unit 1o searches the word dictionary 4 for an independent word (nominal or disjunctive) that has the same pronunciation as the substring from the beginning among the kana code strings stored in the input force decoding temporary storage unit 2. Then, the part-of-speech determining unit 11 uses the searched independent words.
I will appear on.

The part-of-speech determination unit 11 reads the part-of-speech information (information indicating whether it is a noun, adjective, adverb, verb, etc.) of the independent word that is stored in association with the independent word searched by the word dictionary 4, and reads the part-of-speech information (information indicating whether it is a noun, adjective, adverb, verb, etc.) According to the processing rules (grammatical conventions and rules for Japanese), it is determined whether the word can form a clause on its own. Next, if the independent word cannot form a clause by itself, it is interpreted as a 1-clause as described below.

In other words, the independent word candidates obtained as above are
Among the input kana code strings temporarily stored in the conversion candidate temporary storage section 12 and then stored in the input force decoding temporary storage section 2, for the sequence after the position next to the previous independent word candidate,
The adjunct word search unit 13 performs an adjunct word search. In other words, a match between the input kana code sequence and the adjunct word in the adjunct word dictionary 14 is searched.

Then, the connection checking unit 16 checks the grammatical connection conditions between the retrieved adjunct word and its independent word or the immediately preceding adjunct word,
Only pairs of independent words and attached words that satisfy preset and stored grammatical connection conditions are stored in the clause candidate temporary storage unit 12.

Next, it is determined whether or not each phrase candidate in this phrase candidate temporary storage unit 12 has the same length as the input kana code sequence length, and phrase candidates having the same length are output and displayed as kana-kanji conversion output. It is sent to the output display section 7 and displayed on the output display section 7, and the creation of the Japanese sentence is completed. In addition, if the length of any clause candidate is less than the length of the input kanacode series, repeat the clause search operation described above until a grammatically valid 4-kana, kanji-mixed sentence is obtained for all input kanacode series. Repeatedly, when a grammatically valid sentence containing kana and kanji is obtained, it is displayed on the output display section 7, and the creation of the Japanese sentence is completed.

l When inputting kanji For example, when creating a printed Japanese text, if you want to correct one character in the kanji notation of the input Japanese text, the next kana code input to the kana code input device 1 will be changed to the kanji character. It sends information to the control unit 8 via the function key 9 that the search is not allowed. Upon receiving this information signal, the control section 8 puts the index section 6 into an operating state. □For example, a kana code string such as shi, kyuu kiwa meruru kenkyu (where the mouth represents a blank code) is input from the kana code input device 1, and this kana code is stored in the input input decode temporary storage unit 2. shall be. The kana code dividing unit 16 divides the kana code string stored in the input kana code temporary storage unit 2 by blank codes, and sends the kana codes kyu, kiwameru, and kenkyu to the syllable count calculation unit 17, respectively. A syllable number calculation unit 17 calculates the number of syllables of each divided kana code, and outputs the value to the search control unit 18. If we assume that the consonant (tsu) is one syllable by itself, then the ``kyu'' has two syllables, the ``kiwameru'' has four syllables, and the ``kenkyu'' has four syllables.

Next, the search control unit 18 determines whether to search in the syllable number calculation unit 17 and the kanji dictionary 6. That is, when the number of syllables is 2 or less, both the kanji dictionary 5 and the word dictionary 4 are searched, and when the number of syllables is 3 or more, only the word dictionary 4 is searched.

The words in the kanji dictionary 5 and the word dictionary 4 searched as described above are stored in the kanji candidate temporary storage section 19. As shown in FIG. 4, the kanji candidate temporary storage unit 19 stores the results of the kanji dictionary search and word dictionary search for "kyu'" in the first row, and the word dictionary for "kiwameru" in the second row. As for the search results, the results of the word dictionary search for "kenkyu" are stored in the form of kanji codes in the third line.

Next, the matching search unit 2o searches for kanji codes that appear in all lines of this kanji candidate temporary storage unit 19. Sequentially, it is determined whether the kanji code is included in all other lines. Generally, a plurality of kanji codes that are determined to appear in all rows by the match recognition section 2o are output and displayed by the output display section 7.

Note that if there are multiple kanji displayed in the output, only the desired kanji can be output and displayed by pressing the desired number among the numbers attached to each kanji. I can do it.

As described above, the search control unit 18 searches the kanji dictionary 5 that includes only the on-yomi readings of kanji based on the number of syllables of the divided input kana code, but also searches the kanji dictionary 5 that includes only the on-yomi readings of kanji characters, or the kanji dictionary 5 that includes the kun-yomi of kanji characters and the pronunciations of compound words that include the kanji characters. This is to determine whether or not to search the word dictionary 4.

Here, when the number of syllables of the divided input input na code to the search control unit 18 is 2, the well-known "i" in which the second syllable kana always appears in the second syllable.

By adding a function to check whether it corresponds to any of the following, only when the above test is satisfied, the kana input sequence of the kanji dictionary 6 is added. assumes the on-yomi of kanji, the kun-yomi, and the pronunciation of idioms containing the kanji.

Therefore, in the kanji dictionary 6, the on-yomi of a kanji and its kanji code are stored as a pair. However, for example, when the kanji ``general'' is explained by zero head, it is usually said ``general bread'', and the sound ``pan'', which is not originally in the onyomi of the kanji, is used.
is usually used as the on-yomi reading of kanji. In this embodiment, in order to recognize such an onyomi with rendaku sound as an input kana sequence of the index character no dame, the match reading section 2
At ゜, as shown in Fig. 4, the kanji candidate temporary storage unit 1
9. If a kanji code appearing across all lines is not detected, a signal to this effect is sent to the search control unit 18. The detection control unit 18 checks whether the first and second kana of the input kana code series searched as on-yomi corresponds to any of "ba, pi, pu, be, po",
If applicable, change the kana to “Ba, Hi, Fu, He,
The kanji dictionary is searched again for the input kana series with the changed on-yomi. If you do this, you can index even the onyomi where rendaku occurs.

As described above, the present invention uses the on-yomi, kun-yomi, and the pronunciations of compound words containing the kanji as long input information for index characters of kanji, and displays and displays kanji that are simultaneously specified from these input information. Therefore, there is a high possibility that only one Chinese character will be displayed and output depending on the input information, which reduces the need for selection by the operator and reduces the burden on the operator.

In addition, a kanji dictionary is set up for on-yomi as a kanji index dictionary, and a word dictionary for kana-kanji conversion is shared for the conversion of kun-yomi and compound words that include the kanji, so the kanji dictionary itself is It reduces the capacity.

Furthermore, since the input information can be input in the order in which kanji are explained orally in daily life, such as "Asahiroasa" or "Sekai Lse", the input operation becomes natural.

In addition, the kun-yomi or compound words that include the kanji that are usually used to explain kanji are of course well-known words, and if the conversion rate of the kana-kanji converter is practical, they will naturally be registered.
It has the advantage that there is no need to register it specifically for the kanji index. Furthermore, even when a rendaku sound is input as an on-yomi for index input, by providing a function to automatically change it to a clear sound, it can be used as a kanji dictionary without increasing the capacity of the kanji dictionary. Its benefits are great, including the advantage of being able to grow purple sweet potatoes in a natural way.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of a Japanese sentence creation device in an embodiment of the present invention, FIG. 2 is a block diagram showing the configuration of the kana-kanji conversion section, which is the main part, and FIG. 3 is the purple part of the main part. FIG. 4 is a block diagram showing the configuration of the potato section, and is a diagram illustrating the storage mode of the kanji candidate temporary storage section. , 1... Kana code input device, 3...
Kana-Kanji Conversion Department, 4...Word Dictionary, 6...
...Kanji Dictionary, 6... Purple Sweet Potato Club, 7...
Output display section, 8...control section.

Claims (2)

[Claims] (1) An input device that inputs kana codes and a word dictionary that stores the on-yomi of kanji and their correspondence with kanji notations convert the kana code string input from the input device into sentences containing kana and kanji. a kanji conversion unit that outputs a kanji conversion unit; a determining unit that determines whether the input kana is an on-yomi reading, a kun-yomi reading, or an idiom reading; Alternatively, the Japanese sentence creation device further comprises a search unit that searches for a kanji using the word dictionary when it is determined that the pronunciation is a compound word. (2) The index section has a test means for checking whether the input kana string can be the onyomi of rendakuon, and if the test result shows that the input kana string can be the onyomi of the shirendakuon, the input kana string is The Japanese sentence creation device according to claim 1, characterized in that a part is changed.