JPS63116269A - Kana/kanji converter for japanese processing - Google Patents

Abstract

PURPOSE:To improve the conversion efficiency of homonym by checking the connecting relation of words based upon meaning information added to respective words in addition to the check of the connecting relation of words based upon the grammatical character in parts of speech. CONSTITUTION:A dictionary retrieving part 2 collates and retrieves an input character code string with a character string in a conversion dictionary 3 to retrieve the existence of 'KANJI' (Chinese character)/'KANA' (Japanese syllabary) mixing character string corresponding to the input character string. When one corresponding character string exists, an output entry word stored in an output entry word storing part 3b and a meaning code stored in a meaning code storing part 3d are stored in a register 6. When there are plural corresponding character strings, an output entry word stored in the storage part 3b corresponding to a character string most close to the meaning code already stored in the register 6 and the meaning code stored in the storage part 3d are stored in the register 6. The output entry word from the register 6 is applied to an output control part 4 and a 'KANJI'/'KANA' mixing character string is outputted.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a kana-kanji conversion device for Japanese processing.

[Background of the invention]

In Japanese document creation devices, word processors, etc., processing for converting kana characters into kanji characters is essential. Such kana-to-kanji conversion has progressed from the initial word-for-word conversion to word-for-word conversion, and from phrase conversion to batch conversion.

[Prior art]

In conventional kana-kanji conversion devices of this type, judgments are made by checking only the connections between the words ``i'' and ``go'' based on grammatical characteristics such as parts of speech. However, when there are a plurality of homophones having the same grammatical characteristics, the desired word is often not accurately obtained, and there is a drawback that it takes time for repeated conversion.

[Purpose of the invention]

In addition to checking the connection relationships between words based on grammatical characteristics such as parts of speech in conventional kana-kanji conversion devices, the present invention
The purpose of this method is to check the connection relationships between words based on the semantic information added to each word, and to improve the conversion efficiency of homophones.

[Structure of the Invention] An object of the present invention is to provide a kana-kanji conversion device having the structure described in the claims, that is, an input device that converts input pronunciation into a kana code, an input headword storage unit, and an output header. A conversion dictionary comprising a word storage section, a part-of-speech information storage section, and a meaning information storage section; and an input headword storage section of the conversion dictionary that stores a character string code containing kanji and kana corresponding to the kana code given from the input device. a dictionary retrieval unit that performs a search in the conversion dictionary, determines the priority of homophones based on the information stored in the semantic information storage unit of the conversion dictionary, and outputs the information stored in the corresponding output headword storage unit; This is achieved by a kana-kanji conversion device comprising: , an output control section that outputs a character string corresponding to the output of the dictionary search section to a display device.

[Effects of the Invention] According to the kana-kanji conversion device according to the present invention, a character string representing a reading input from an input device, such as a katakana character string, is compared with a character string registered in a conversion dictionary, and the character string is converted into a kanji-kana character string. A mixed notation is obtained, but in this case, the conversion is performed taking into account the semantic code. As a general rule, this semantic code is outputted with the closest meaning to the previously converted character string, and at the same time, the character string written in Kanji/kana and the meaning code is stored in a register in preparation for the next input. Therefore, it is possible to perform kana-kanji conversion that takes into account the meaning of the entire sentence.

[Examples] Hereinafter, the present invention will be disclosed with reference to the accompanying drawings showing examples.

FIG. 1 is a block diagram showing the basic configuration of a device implementing the kana-kanji conversion system according to the present invention.
Dictionary search unit 2, conversion dictionary 3, output control unit 4, display device 5
, register 6.

The input device 1, which can be a keyboard, a voice recognition device, a kana character reading device, etc., converts a first character string representing an input pronunciation into a kana character code, and transmits the converted character string to the dictionary search unit 2. It is. The first character string representing the reading is shown as hiragana, katakana, romaji, etc.

The dictionary search unit 2 converts the input character code string into a conversion dictionary 3.
Perform a matching search with the character strings registered in advance,
This is to search for the existence of a second character string consisting of display characters mixed with kanji and kana that corresponds to the input first character string.

The conversion dictionary 3 has a memory configuration as shown in FIG. 2, for example, and includes an input entry word storage section 3a, an output entry word storage section 3b, a part of speech code storage section 3C, and a meaning code storage section 3d. . A character string containing kanji and kana corresponding to the reading stored in the input word storage section 3a is stored in the output word entry storage section 3b. The part-of-speech code storage unit 3C stores information on the part of speech corresponding to the kanji-kana-mixed character string in the output headword storage unit 3b.

Further, the meaning code storage unit 3d stores classifications made according to the meaning of the corresponding word in the form of codes. This meaning code is, for example, desire, as shown in the right column of Figure 2.
They are classified into objects, machines, people's names, etc., and similar classifications with similar meanings are assigned similar numerical values.

If the dictionary search unit 2 is unable to search because the second character string that corresponds to the first character string as described above does not exist,
Move on to other processing. If only one corresponding second character string exists, the corresponding output entry word stored in the output entry word storage section 3b and the meaning code stored in the meaning code storage section 3d are stored in the register 6. Store.

If there are multiple corresponding second character strings, the corresponding output headword storage unit 3b of the '2nd character string whose meaning code is closest to the value of the meaning code already stored in the register 6 is stored. The stored output headword and the meaning code stored in the meaning code storage section 3d are stored in the register 6.

In this way, the output headword stored in the register is given to the output control section 4. The output control unit 4 outputs the output headword as a second character string corresponding to the first character string.

If more character strings are input, the above operation is repeated.

Such an operation is not only possible with word-for-word conversion, but also can be applied to batch conversion and batch conversion of sentences with multiple sentences by taking part-of-speech codes into consideration.

FIG. 3 is a flow diagram showing the operation of the kana-kanji conversion system according to the present invention.

A key scan is performed in step S1, and if there is manual key power, it is determined in step S2 whether or not kana-kanji conversion is to be performed. If this determination is NO, another process S5 is performed.

If this determination is YES, the input character code string is read as in step S3.

Next, a search for the input entry word is performed as in step S4, and it is determined whether or not there are any candidates based on the search results as in step S5. If this determination is NO, other processing is performed. If this determination is YES, it is determined in the next step S6 whether there is only one candidate.

If the determination in step S6 is YES, the next step S
7, candidates are read out, and then in step S
At step 8, the presence or absence of a semantic code is determined.

On the other hand, if the determination in step S6 is NO, step S
9, the candidates are read out, and the next step S10
It is determined whether or not there is a semantic code. If the determination here is YES, it is determined in the next step 311 whether or not a numerical value is stored in the register.

If this determination is YES, it is determined in the next step S12 whether or not the numerical value of the meaning code is closest to the numerical value of the register among the candidates.

If this determination is YES, the input/output headword and meaning code are stored in a register as in step S13. Further, the output headword is displayed as in step S14, and then it is determined as in step S15 whether or not the input character string still remains. If the determination in step S15 is No, other processing is performed. Also, this judgment is YES
In this case, the process returns to before step S4 and the above-described flow is repeated.

If the determination in step 316 is YES, step S9
Go back to before. Further, if this determination is NO, the process skips to after step 313.

If the determination in step 311 is NO, step S12
is skipped and step S13 is performed.

If the determination in step S12 is No, the process returns to before step S9.

If the determination in step S8 is YES, step 813
, and in the case of No, before step S14.

Here, using the conversion dictionary 3 shown in Figure 2, convert the human character string ``Kuukiha Kitita'' to the target character string ``Air is a gas.''
A specific example of trying to obtain the following will be explained in conjunction with the above-mentioned flow diagram.

33 Human-powered character string “Kuukihakitita” S4
Search Candidate Only 1 item S7 Read candidate "Kuuki 1081 / Air" * There is a meaning code, but the register is empty.

S13 Store “(Uki1081/Air)” in the register S
14 Output [Display] “Air” * There are still input characters.

"Hakitai" S4 Search Candidate Only 1 S7 Candidate reading "Ha/□/Ha"* Since it is a particle, there is no meaning code.

S14 output [display] “Air is” * There are still characters left to input.

"Kitai" S4 Search 4 candidates S9 Candidate reading * The closest one to the meaning code (081) of the register is "Kitai 1081/Gas". "Kitai 1081/Gas" is stored in the S13 register.

S14 output [display] “Air is a gas” * There are still input characters. "Da" S4 Search Candidate Only 1 S7 Read Candidate "Da/□/Da"
* Since it is an auxiliary verb, there is no semantic code.

514 Output [Display] “Air is a gas” * There are no more input characters left.

S. Go to other processing.

Note that in the above steps, * indicates the determination result.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of the basic configuration of a kana-kanji conversion device according to the present invention. FIG. 2 is an example of a memory configuration diagram of a conversion dictionary used in the kana-kanji conversion device according to the present invention. FIG. 3 is a flow diagram showing the operation. The correspondence of reference numerals in the figure is as follows. 1: Input device 2: Dictionary search unit 3: Conversion dictionary 4: Output control unit 5: Display device 6: Register

Claims (1)

[Claims] an input device that converts the input pronunciation into a kana code; a conversion dictionary that includes an input headword storage unit, an output headword storage unit, a part-of-speech information storage unit, and a meaning information storage unit; and a kana code given from the input device. A character string code containing kanji and kana corresponding to the code is searched in the input headword storage unit of the conversion dictionary, and at the same time, the priority of the homophone is determined based on the information stored in the semantic information storage unit of the conversion dictionary. A dictionary search unit that determines and outputs information stored in a corresponding output headword storage unit; and an output control unit that outputs a character string corresponding to the output of the dictionary search unit to a display device. Features of Kana-Kanji conversion device.