JPH0561863A - Automatic correction processor for japanese sentence - Google Patents

Abstract

PURPOSE:To automatically correct an inflection word without using data on reading and a dictionary on a Japanese sentence automatic correction processor for substituting the inflection word detected from a Japanese sentence for an inflection word which is previously made to correspond to the word. CONSTITUTION:A corresponding table 12 storing corresponding information between a character string compared with a correction object and the type of the inflection, and a substituted character string and the type of the inflection are provided. When an input part 14 inputs the Japanese character string becoming the correction object, a detection part 16 retrieves the corresponding table 12 and detects the character string coincident with the compared character string from the inputted character string. When there is the compared character string in the inputted character string, a substitution part 17 substitutes the compared character string and inflection for the corresponding substitution character string and the inflection by referring to the corresponding table 12. An output part 18 outputs the result to an output device 20. Thus, the inflection word in the Japanese sentence is automatically corrected.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention replaces a detected inflection word with a previously associated inflection word by utilizing the fact that the inflection word in a Japanese sentence can be uniquely determined by specifying the stem and the type of inflection. The present invention relates to a Japanese sentence automatic correction processing device.

In recent years, with the spread of word processors and the like, many systems that can efficiently process Japanese documents have come to be used. Such a system has a main purpose of creating, saving, and editing a document, but there is also a system having an automatic sentence correction function as another function. The automatic correction of a sentence means that the system corrects an error in the sentence or rewrites the sentence in an easy-to-understand manner. It is desired to provide a more useful device by improving such a system.

[0003]

2. Description of the Related Art As a conventional device having an automatic correction function for a utilized word, there is, for example, a document revision processing device disclosed in Japanese Patent Laid-Open No. 62-262174.

This device detects a refined phrase from a sentence,
If the reading of the revised phrase matches the reading of the modified phrase when the revised phrase has a practical use, the phrase is replaced. That is, in this apparatus, the search means searches for a rubric phrase from the character information after the kanji conversion of the Japanese sentence stored in the storage means based on the bulge word information held by the dictionary means. Especially for inflectional words, the refined word is searched for based on the information of the word stem also possessed by the dictionary means.

Then, the correction means corrects the found refined word and its adjunct word on the basis of the stem of the amended word and phrase, its inflection ending and its change, and the adjunct word and its change information. Further, of the searched refined words and phrases, the words and phrases including the kanji are judged from the characters in the memory state before the kanji conversion based on the information of the kanji reading included in the dictionary means.

[0006] However, in this device, it is necessary to extract information about reading from the text, and whether to embed the reading in the text in advance or to determine the reading from the dictionary by performing morphological analysis. ， It will take time and effort.

[0007]

DISCLOSURE OF THE INVENTION The present invention solves the above problems, and even when a character string to be compared (hereinafter referred to as a comparison character string) or a replacement character string (hereinafter referred to as a replacement character string) includes utilization. , It is an object of the present invention to provide a device for automatic correction without using reading data or a dictionary.

[0008]

FIG. 1 illustrates the principle of the present invention. In FIG. 1, 10 is an input device such as a keyboard and an external storage device, 11 is a processing device including a CPU and a memory, 12 is a correspondence table of comparison character strings and replacement character strings, and 13 is an input character string to be corrected. , 1
4 is an input unit, 15 is a utilization table that stores information on the utilization form for each type of utilization, 16 is a detection unit, 17 is a replacement unit, 1
8 is an output unit for outputting the correction result, 19 is an output character string, 2
Reference numeral 0 represents an output device such as a display, a printer or an external storage device.

Since the Japanese sentence automatic correction processing apparatus of the present invention automatically corrects an erroneous conjugation word, the comparison character string to be compared with the correction target and the type of its utilization, the replacement character string to be replaced and the type of its utilization, The correspondence table 12 for storing the correspondence information of The input unit 14 is a processing unit that inputs the Japanese character string to be corrected, that is, the input character string 13 from the input device 10. The detection unit 16 is
It is a processing unit that searches the correspondence table 12 and detects a character string in the input character string 13 that matches the comparison character string.

When there is a comparison character string in the input character string 13, the replacing unit 17 refers to the corresponding table 12 for the comparison character string and the inflection ending, and thereby the corresponding replacement character string and the inflection ending thereof. Is a processing means for replacing with. The output unit 18 is a processing unit that outputs the character string obtained by replacing the input character string 13 by the replacing unit 17 as the output character string 19 to the output device 20.

[0011]

[Function] Japanese conjugation words are divided into verbs, adjectives, and adjective verbs. All of these can be created by specifying the stem and the type of usage. For example, the verb "move" has a stem "verb" and a type of conjugation "ka-line five-stage verb", and from this, an inflectional form can be created according to the inflection rule.

In the present invention, the user designates the stem and inflection of each of the comparison character string and the replacement character string in advance and registers them in the correspondence table 12. As a result, it is possible to accurately perform the search for the comparison character string in consideration of the usage form and the creation of the replacement character string for the comparison character string in consideration of the usage form by using the information on the usage.

The input unit 14 displays, for example, "The dog has moved."
When the input character string 13 is input, the detection unit 16 searches the correspondence table 12 and detects the comparison character string “moving”. This is notified to the replacement unit 17. Replacement unit 17
Creates an inflectional replacement character string of "in motion" consisting of a stem of "in motion" and an inflection end of "i" for the comparison character string of "in motion", and replaces "in motion" in the input character string Replace with "moving".

When the detection unit 16 detects the end of the character string, the output unit 18 outputs the result as an output character string 19 to the output device 20. By the above processing, the conjugation words including mistakes in the Japanese sentence are automatically corrected.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 2 is an explanatory view of the inflectional form creation according to an embodiment of the present invention. For example, the rule to use the five-stage verb is
As shown in (a) of (1), the first antecedent is the stem of the inflection word + the letter in the nth line, and the second is the inflection, the stem of the inflection word + the letter in the nth row, ... Has become. here,
Characters in the nth row and the nth row are:
It means the character "ka", and the character in the nth row is n.
= If "ka" (K line), indicate the character "ko".

The phonological stool of the character in the n-th step in the second continuous form can be uniquely determined by n. For example, n =
If it is “ta” (ta row), “tsu” is phonological. Since the utilization rules as shown in FIG. 2 (a) are determined for each type of utilization, based on this, the utilization forms can be created as shown in FIG. 2 (b). That is, when creating the first form of "moving", the inflectional ending of "ka" is added to the "moving" of the stem to create "moving". The same applies to other forms of utilization.

Next, an embodiment of the present invention will be specifically described with reference to FIGS. FIG. 3 is an explanatory diagram for searching a comparison character string according to an embodiment of the present invention, FIG. 4 is an explanatory diagram for searching a comparison character string according to an embodiment of the present invention when the word is a conjugation word, and FIG. FIG. 8 is an explanatory diagram of creating a row.

(1) The input unit 14 shown in FIG.
The input character string 13 is obtained from 0. The input character string 13 may be a document unit, a sentence unit, or a simple character string. In this example, it is assumed that the input character string 13 is as shown in FIG.
At the very end of the input character string 13, there is a code of the character "EOS" indicating the end of the sentence, and this character does not exist in the sentence.

(2) The detector 16 searches the input character string 13 for a comparison character string. Specifically, the following (a) to (i)
To execute. (a) First, set the input character string pointer so that it points to the beginning of the input character string.

(B) Next, the comparison character string pointer is set so as to point to the comparison character string at the head of the column of the comparison character string in the correspondence table 12. However, the code of the character "EOT" indicating the end of the table is assumed to be present in the last row of the correspondence table 12.

(C) As shown in FIG. 3, the comparison character string pointed to by the comparison character string pointer is set as the current comparison character string. (d) Length of the character string pointed to by the input character string pointer (however,
If "EOS" is not included) is shorter than the length of the current comparison character string, proceed to process (g). If not, compare with the character string pointed to by the input character string pointer, as shown in in Fig. 3. Compares character strings character by character for the length of the current comparison character string. If they match, go to process (e). If they do not match, go to process (g).

(E) It is checked whether or not the current comparison character string is a conjugation word. If it is an inflection word, proceed to processing (f), and if not, proceed to processing (3). (f) As shown in FIG. 4, all inflection endings of the current comparison character string are created by referring to the utilization type of the current comparison character string. This is because the comparison character string “gogo” is currently “ka-go-godan” in the correspondence table 12, so as shown in FIG. Can be processed by. And FIG.
As shown in, check whether the character string pointed to by the input character string pointer matches any of their inflection endings. If there is a match, the utilization form is stored and the process (3) is performed. If there is no match, process
Proceed to (g).

(G) The comparison character string pointer is set so that it points to the comparison character string next to the current comparison character string currently pointed to. (h) Add one character pointed to by the input character string pointer to the output character string and proceed to processing (i).

(I) The input character string pointer is set so that the input character string pointer points to the next character. If the input character string pointer points to "EOS", the process proceeds to the process (4), and if not, the process returns to the process (c).

(3) The replacement unit 17 adds the utilization replacement character string corresponding to the current comparison character string to the output character string. Specifically, the following (j) to (m) are executed. (j) As shown in FIG. 5, by referring to the correspondence table 12, the replacement character string corresponding to the current comparison character string is extracted.

(K) Check whether the replacement character string is a conjugation word. If it is an inflection word, proceed to processing (l), and if not,
Go to process (m). (l) As shown in (5) of FIG. 5, it is checked which conjugation form the current comparison character string corresponding to the replacement character string is, and the conjugation form character string of the replacement character string is extracted. Then, as shown in in FIG.
The replacement character string is the replacement character string with its inflection ending added to it.

(M) The utilization replacement character string is added to the output character string. Set the comparison character string pointer so that the comparison character string pointer points to the beginning of the correspondence table 12, and execute the process (2).
Return to (i) and repeat the process.

(4) The output unit 18 shown in FIG.
9 to an output device 20 such as a terminal device or an external storage device.

[0029]

As described above, according to the present invention,
Even for words that are in use, you can search and replace by simply specifying the type of inflection without using reading data or dictionaries, and efficiently perform automatic correction of Japanese sentences. Will be able to.

[Brief description of drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is an explanatory diagram of inflectional form creation according to an embodiment of the present invention.

FIG. 3 is an explanatory diagram for searching a comparison character string according to an embodiment of the present invention.

FIG. 4 is an explanatory diagram of a search when a comparison character string is an inflection word according to an embodiment of the present invention.

FIG. 5 is an explanatory diagram of creating a replacement character string according to an embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Input device 11 Processing device 12 Correspondence table 13 Input character string 14 Input part 15 Utilization table 16 Detection part 17 Substitution part 18 Output part 19 Output character string 20 Output device

Claims (1)

[Claims] 1. A Japanese sentence automatic correction processing device for inputting and correcting a Japanese document, sentence or character string by a computer, wherein a comparison character string to be compared with a correction target and a type of its utilization, and replacement replacement A correspondence table (12) for storing correspondence information between character strings and types of utilization thereof, an input unit (14) for inputting a Japanese character string to be corrected, and the correspondence table (12) are searched, A detection unit (16) that detects a character string that matches the comparison character string from the input character string and the comparison character string and the inflection ending when the comparison character string is present in the input character string By referring to (12), a replacement unit (17) that replaces the corresponding replacement character string and its inflection ending, and an output unit (18) that outputs a character string in which the input character string is replaced by the replacement unit (17). It has and and automatically corrects inflection words in Japanese sentences. Japanese sentence automatic correction processing apparatus according to claim that it has to.