JPH0468466A - Kana / kanji converting device - Google Patents

Abstract

PURPOSE:To clearly distinguish a word by preventing a word, which is differentially described for plural words having various meanings in common, from being used in the case of conversion. CONSTITUTION:When the use limit of a word is instructed from a limit instruction part 7, a control part 8 stores the instruction in an internal memory. A conversion part 3 divides an input character string into several partial strings and instructs a dictionary retrieval part 5 so as to retrieve a word, which has each partial string as reading, from a word dictionary, and the dictionary retrieval part 5 retrieves the word having the designated reading from the word dictionary 4. Then, KANA (Japanese syllabary)/KANJI (Chinese character) conversion is executed without using the word which is differentially described for the plural words having various meanings. Thus, ambiguity can be decreased concerning the correctness of a converted result.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a kana-kanji conversion device that converts sentences input in kana-kanji or romaji into sentences containing kanji and kana.

[Conventional technology]

In addition to homophonic antonyms, different spellings of the same word may be registered in a kana-kanji conversion dictionary for the same reading. Examples of different spellings are ``do'' and ``do''.
There are words with different okurigana, such as ``,'' and words with different glyphs, such as ``tsutsuru'' and ``tsutomeru.''

Conventional kana-kanji conversion devices process all words registered in word dictionaries.

FIG. 2 is a block diagram of an example of a kana-kanji conversion device using the prior art. This kana-kanji conversion device includes an input section 1 for inputting characters, an output section 2 for outputting the conversion result, and a conversion section 3 for converting the input sentence into a sentence containing kanji and kana.
It has a word dictionary 24 that registers words used for conversion, a dictionary search section 25 that searches for words from the word dictionary 24 using pronunciation as a key, and a control section 28 that controls the operation of the entire device.

The input section 1 is a keyboard, the output section 2 is a CRT device or a printer, and the conversion section 3. Dictionary search section 25. The control unit 28 is a CP
By using U and the memory device, the word dictionary 24 is realized by a file device such as a hard disk or a memory device.

Table 1 shows "yomi" as an example of the structure of the word dictionary 24.

``The notation indicates the ``part of speech.''"Yomi" is the reading of the word, "notation" is the representation of the word, and "part of speech" is the part of speech of the word.

The first table conversion unit 3 receives words from the dictionary search unit 25 and combines the words to create a conversion result. Such a device is used, for example, in “Morphological Analysis of Solid Japanese Sentences Using the Minimum Clause Count Method” (Yoshimura et al., Information Processing Society of Japan Journal Vol. 1.24).
This can be easily achieved by the method described in No. 1).

FIG. 3 is a flowchart showing the operation of this conventional kana-kanji conversion device. The operation of this conventional example will be explained below with reference to FIG.

a. Input a kana character string from input section 1 (step 31
).

b. After the input is completed, the control unit 28 passes the input character string to the conversion unit 3 and instructs conversion.

The C3 conversion unit 3 divides the input character string into several substrings and instructs the dictionary search unit 25 to search the word dictionary 24 for words that have the pronunciation of each substring.

d. The dictionary search unit 25 searches the word dictionary 24 for a word with the specified pronunciation (step 32).

e. Pass the search results to the converter 3.

f. The conversion unit 3 combines the searched words to create a kana-kanji conversion result (step 33).

g. After the conversion is completed, the control unit 28 outputs the conversion result to the output unit 2 (step 34).

[Problem to be solved by the invention]

Even if some of the outputs of the conventional kana-kanji conversion device described above are generally correct conversion results, they may be inconvenient for specific purposes.

Assume now that the contents of the word dictionary 24 are as shown in Table 1 (descriptions of case particles and conjugated endings of verbs are omitted as they are not relevant to the following explanation). Then, “
The result of converting the sentence ``I will work to improve the equipment'' is either ``I will work to improve the equipment'' or ``I will work to improve the equipment.''

Of the three types of results above, “Working to improve equipment” is
This is a conversion error. Both of the results ``I will strive to improve the device'' and ``I will strive to improve the device'' are correct, and the choice is left to the person inputting the text.

However, when the conversion result is used as input to another language processing device, "working" may be inconvenient. For example, when used as input to a machine translation device, the translation results will differ depending on whether ``tsutomeru'' is interpreted as ``trying'' or ``working''. Therefore, it is necessary to take measures such as having the user of the machine translation device instruct the user to make the distinction or incorporating a mechanism for discrimination into the device.

In addition, when a device developer or the like verifies the conversion algorithm built into the device, especially the homophone determination algorithm, it is also possible to say ``I will work to improve the device'' even if the result is ``I will work to improve the device.'' It is necessary to determine whether it has been selected as a variant of ``to try'' or ``to work.'' This information also needs to be output, which increases the amount of output data and increases the burden on the people doing the evaluation work.

An object of the present invention is to provide a Kana-Kanji conversion device that can clearly distinguish between words by not using words that are different spellings common to a plurality of words with different meanings during conversion.

[Means to solve the problem]

The kana-kanji conversion device of the present invention includes: an input section for inputting a sentence; a conversion section for converting the text input from the input section into kana-kanji; and a word dictionary including one pronunciation, one spelling, and one part of speech of a word.
The word dictionary is equipped with a dictionary search unit that searches for a word from the word dictionary using the pronunciation of the word as a key, and an output unit that outputs a kana-kanji conversion result. a restriction instruction section that further includes information indicating whether the words are written in a different way and instructs the device to perform kana-kanji conversion while restricting the words to be used; and the dictionary search section when instructed by the restriction instruction section. and a word restriction unit that selects words from the searched words according to the information.

[Effect]

The above-mentioned problem to be solved with the conventional kana-kanji conversion device is caused by the fact that a certain word has a common variant spelling for multiple words with different meanings. Therefore, the kana-kanji conversion device of the present invention has a function of converting multiple words with different meanings without using words that are different spellings. That is, information for distinguishing whether or not it is a different notation of a plurality of words is added to each word, and words to be used for conversion are selected based on the information. Thereby, ambiguity regarding whether the conversion result is correct or incorrect can be reduced.

[Example] Next, an example of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing one embodiment of the present invention. This kana-kanji conversion device has an input section 1 for inputting characters, an output section 2 for outputting the conversion result, a conversion section 3 for converting the input sentence into a sentence containing kanji and kana, and a register for registering the words used for conversion. A certain word dictionary 4, a dictionary search unit 5 that searches for words from the word dictionary 4 using pronunciation as a key, and a word restriction unit 6 that determines whether or not to use the words searched from the word dictionary 4 for conversion are used. It includes a restriction instruction section 7 that instructs to perform kana-kanji conversion by limiting words, and a control section 8 that controls the operation of the entire device.

Dictionary search section 5. Word restriction section 6. The control unit 8 can be realized by a CPU and a memory device, and the word dictionary 4 can be realized by a file device such as a hard disk or a memory device. The restriction instruction section 7 can be realized by incorporating it into the input section 1 as an instruction key. It should be noted that blocks with the same numbers in FIG. 1 and FIG. 2 are the same constituent elements, so explanations thereof will be omitted.

Table 2 shows an example of the structure of the word dictionary 4. In this example,
It shows "pronunciation", "notation", "part of speech", and "restrictions".

Table 2: ``Yomi'' indicates the word's reading, ``Spelling'' indicates the word's notation, ``Part of speech'' indicates the word's part of speech, and ``Restriction'' indicates whether this word is a different spelling of multiple words with different meanings. This information (referred to as a restriction flag in the following explanation) indicates that it is a different notation when it is ``1'', and that it is not a different notation when it is ``0''. ”, so the restriction flag is 1. On the other hand, "furuma" is a different notation common to "behavior" and "behuma," but since there is no difference in meaning between "behavior" and "behuma," the restriction flag is set to O.

FIG. 4 is a flowchart showing the operation of this embodiment.

The operation of this embodiment will be explained below with reference to FIG.

a. Input a kana character string from input section 1 (step 41
). When the restriction instruction unit 7 instructs to restrict the use of a word, the control unit 8 stores the instruction in its internal memory.

b. After the input is completed, the control unit 8 passes the input character string to the conversion unit 3 and instructs conversion.

The C9 conversion unit 3 divides the input character string into several substrings and instructs the dictionary search unit 5 to search the word dictionary 4 for words with readings for each substring.

d. The dictionary search unit 5 searches the word dictionary 4 for a word with the specified reading (step 42).

e. If the restriction instruction unit 7 has previously instructed to restrict the use of a word, the word restriction unit 6 deletes the searched words with restriction flags of °゛11 from the search results (
Step 43).

f. Pass the search results to the conversion unit 3.

g. The conversion unit 3 combines the searched words to create a kana-kanji conversion result (step 44).

h. After the conversion is completed, the control unit 8 outputs the conversion result to the output unit 2 (step 45).

By doing this, the result of converting the sentence ``I work to improve the equipment'' becomes either ``I work to improve the equipment.''J ``I strive to improve the equipment.''

〔Effect of the invention〕

As described above, the present invention clarifies the distinction between words by not using words that are different spellings common to multiple words with different meanings during conversion. the result,
This has the effect of making it easier to use the conversion results as input data for machine translation, and making it easier for device developers to evaluate conversion algorithms built into the device.

[Brief explanation of the drawing]

1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a block diagram showing an example of a conventional kana-kanji conversion device, FIG. 3 is a flowchart showing the operation of the kana-kanji conversion device shown in 2, and 4. The figure is a flowchart showing the operation of the embodiment of FIG. ...Input section...Output section...Conversion section...Word dictionary...Dictionary search section...Word restriction section, restriction instruction section, control section

Claims (1)

[Claims] (1) an input section for inputting sentences; a conversion section for converting the sentences inputted from the input section into kana-kanji; and readings of words;
In the kana-kanji conversion device, the kana-kanji conversion device includes a word dictionary including notation and parts of speech, a dictionary search unit that searches for a word from the word dictionary using the pronunciation of the word as a key, and an output unit that outputs the kana-kanji conversion result. a restriction instruction section that further includes information indicating whether a word in the word dictionary is a different spelling of multiple words with different meanings, and instructs the device to perform kana-kanji conversion while restricting the words to be used; A kana-kanji conversion device comprising: a word restriction section that selects words from the words searched by the dictionary search section according to the information when instructed by the restriction instruction section.