JPH01321553A - Japanese word processing method with literary style converting function - Google Patents

Abstract

PURPOSE:To thoroughly detect the lack of coordination in literary styles of sentences under preparation and to automatically convert these styles into optional ones by preparing a style converting function to recognize the uncoordinated styles and to convert them into a coordinated styles designated by a user. CONSTITUTION:When a style converting function 3 is started, the mixture of plural styles is detected from the head through the end of a character string text stored in a character string text store part 4 by reference to a single sentence head detecting table 51, a table 52 for previous/continuous/end-form/ participial adjective/assumptive forms of adjectives as well as for previous/ continuous/assumptive forms of verbs and adjective verbs, etc., stored in a table store part 5. Then the warning is produced. Thus the conversion of styles is carried out based on the designation of a user. In such a way, the uncoordinated styles can be automatically coordinated with no manual check. Furthermore the sentences can be automatically converted into the optional styles.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a Japanese language processing method, and is suitable for improving the quality of created Japanese sentences and preprocessing input Japanese sentences in machine translation systems. This invention relates to a Japanese language processing method having a writing style conversion function.

[Conventional technology]

In a word processor that creates Japanese documents,
As a method of converting the writing style of constituent sentences in a document, for example, as described in Japanese Patent Application Laid-Open No. 180472/1982, inconsistencies in writing style are identified, the tone of each clause is checked, The number of phrases in the tone of ``desu・mazu'' and the number of phrases in the tone of ``da・deru'' are calculated, and the tone with the larger number is determined to be the tone of the entire sentence, and the result is inconsistent. If so, there is a way to output the discrepancies.

However, the above prior art does not mention the automatic conversion function. Furthermore, there is no mention of a function that allows the user to select an appropriate writing style from multiple writing styles.6 [Problem to be Solved by the Invention] In the above conventional method, the function of automatically converting writing style is not mentioned. This was not taken into account, and there was a problem that it was not possible to automatically convert stylistic changes sequentially or all at once.

The purpose of the present invention is to automatically unify the writing style without the need for the creator to manually check for inconsistencies in writing style when creating a sentence, and furthermore, to change the writing style to any writing style as necessary. The object of the present invention is to provide a Japanese processing method with a writing style conversion function that can automatically convert the text into Japanese.

[Means to solve the problem]

In order to achieve the above object, the present invention identifies constituent simple sentences for each sentence in a single sentence, and uses a writing style in which the user specifies the last constituent word for a compound term appearing in each simple sentence. It is characterized by converting other constituent words into the writing style of ordinary expressions, and also unifying the tenses based on the tense at the end of the sentence in the writing style of polite expressions.

[Effect]

The writing style conversion function according to the present invention automatically identifies inconsistencies in writing style and automatically converts the writing style into a unified writing style specified by the user.
Furthermore, if the user specifies a specific writing style from the beginning, the writing style is automatically unified and converted according to the instruction.

As a result, it is possible to automatically detect any inconsistencies in the writing style of the text being created, automatically convert it to any writing style, and to revise and proofread the text to be of higher quality.

〔Example〕

An embodiment of the present invention will be described below with reference to FIGS. 1 to 26.

FIG. 1 shows the functional configuration of the Japanese word processing with writing style conversion function of the present invention, in which 1 is a word processing (hereinafter abbreviated as word processor) processing control function block, 2 is a word processing function block group, and 3 is a word processing function block group. 1 is a writing style conversion function block, 4 is a character string text storage unit, 5 is a table 51 for detecting the beginning of a single sentence, and a table 51 for detecting the beginning of a single sentence, as well as verbs and adjectives for occurrence, conjunction, termination, adjunction, imperative, hypothetical, and adjectives for occurrence, conjunction, and termination.・Adnominal/imperative/hypothetical form table 52; verb/adjective verb final/adjunctive/imperative form table 53; auxiliary verb auxiliary/adjunctive/final/adjunctive/imperative/hypothetical form and particle table 54; colloquial expressions・The writing style conversion function 3 is a table storage unit that stores the honorific expression table 55 and the phonetic table 56.

When activated, the table 51 for detecting the beginning of a single sentence stored in the table storage unit 5 and the verb and
A table 52 of adjective verbs in the precipitous, conjunctive, and hypothetical forms and adjectives in the precipitous, conjunctive, final, adjunctive, and presuppositional forms; a table 53 for verbs and adjective verbs in the final, adjunctive, and imperative forms; Imperative table 54 and auxiliary verbs
While referring to the conjunctive, final, adnominal, hypothetical, and particle tables 55, a mixture of multiple writing styles is detected and a warning is issued. Further, according to the user's specifications, character string text is input, proofread, edited, displayed, and outputted by activating word processing functions 2 other than the 0 writing style conversion function that converts writing styles. From FIG. 2 onwards, the control and processing of the writing style conversion function in FIG. 1 will be explained in detail.

FIG. 2 shows the hardware configuration of an embodiment of the present invention, in which 6 is a processor, 7 is a keyboard input device for inputting characters and editing instructions, and 8 is a keyboard input device for inputting characters and editing instructions.
is a display, and 9 is a storage device. In the storage device 9, 10 represents a program storage section, 11 a work area, 12 a character string text storage section, and 13 a table storage section.

FIG. 3 shows the board of the keyboard input device 7. On the board, a group of keys 72 for inputting characters,
A writing style conversion instruction key 711 is one of the keys 71 for instructing word processing.

FIG. 4 shows the processing contents of the word processing control function block 1 in FIG.
lysisDiagra+a). When the word processing control function is activated, a key for instructing word processing is read. In process 101, the input key is read, and in process 102, while the input key is not the end key, process 103 and process 101 of reading the input key are performed.
Repeat each. Processing 103 is.

This is to judge whether the input key is a writing style conversion instruction key or not. When the input key is a writing style conversion instruction key, the writing style conversion function 3 shown in Fig. 1 is activated, and when it is not, the word processing function group is activated. 2, start the word processing corresponding to the instruction key.

After the word processing process corresponding to the instruction key is completed, the state returns to the input key reading state and the input key is read (process 10).
1). If the end key is input, all word processing functions are completed (process 102).

FIG. 5 shows the content of the processing 106 of the 1 writing style unification function in PAD. 0 When the writing style unification function processing 106 is activated, the writing style unification range is read (process 601). If there is no range specification, the entire text is set as the stylistic uniformity range (process 602,
603).

Next, the target level of the writing style to be unified is read (process 604). The text data is skipped (processing 606) up to just before the beginning of the specified stylistic uniformity range (processing 605). Next, while within the specified range (process 6
07), read the text data one sentence at a time (processing 6
08), processes 609, 610, and 615 are performed sequentially. In the processing of module 609, the tense of the target sentence is determined from the tense of the predicate at the end of one sentence. Processes 611 and 612 are executed until the process is completed for the read sentence (process 610). In the processing of module 611, the read sentence is cut into simple sentences from the beginning.

In process 612, when a simple sentence is extracted by the process of module 611, it is sequentially determined whether the sentence is in the specified writing style, and if a writing style other than the specified writing style or normal expression is included, a warning is issued. Also, if the specified writing style is not included at all, a warning will be issued.

At this time, if there is a writing style unification instruction (process 613), process 6
In step 14, writing style conversion is performed, and in step 615, colloquial expressions and broken expressions are detected and a warning is issued.

Furthermore, it detects honorific expressions and points out the existence of honorific expressions, if any. When the writing style unification process is completed for all sentences in the specified range in the text data, the writing style unification function is ended (process 106).

In FIG. 6, when the processing 107 of the 0 writing style conversion function, which indicates the processing contents of the writing style conversion function in PAD, is activated, the 1 writing style conversion range is read (process 6o1). If there is no range specified, the entire text is set as the uniform stylistic range (processing 602, 603)
.

Next, read the target level of the writing style to determine which writing style to convert (process 604), read up to just before the beginning of the specified stylistic uniformity range (process 605), and read the text data (skipping).
Process 606). Next, while within the specified range (process 607), text data is read one sentence at a time (process 608), and processes 609, 610, and 615 are sequentially performed.

In the processing of module 609, from the tense of the sentence-final predicate,
Recognize the tense of the target sentence. Processes 611 and 614 are executed until the process is completed for the read sentence (process 610). In the processing of module 611, the read sentence is cut into simple sentences from the beginning.

In process 614, when the simple sentences are extracted in process 611, they are sequentially converted into the specified writing style. In process 615, a colloquial expression or a broken expression is detected and a warning is issued. Furthermore, it detects honorific expressions and points out the existence of honorific expressions, if any. When the writing style conversion process is completed for all sentences in the specified range in the text data, the writing style conversion function is ended (process 107).

FIG. 7 shows the processing contents of the simple sentence style recognition module 611 in PAD. When the simple sentence style recognition processing function is activated, in process 6111, words other than nominal words or plain name characters that can be the beginning of a simple sentence are recognized. In process 6112, when a nominal or hiragana character that can be the beginning of a simple sentence is recognized in process 6111, it is determined whether the word immediately before the character of the recognized word can be connected to the recognized word, or if a word other than hiragana is recognized. If so, it is determined whether the predicate is in the final/adjunctive form, and if so, processes 6113 and 6114 are executed. Process 6113 stores the character immediately before the recognized character other than the nominal or plain name character that can be the beginning of a simple sentence as the end of a simple sentence. In process 6114, compound predicates are recognized.

FIG. 8 shows the processing contents of the compound predicate recognition module 6114 in PAD. When the compound predicate recognition function is activated, processes 61141 and 61142 are performed sequentially.
Process 61141 performs processing to separate the compound predicate stored in process 6113 into component predicates. Processing 61142 identifies the style of each of the divided constituent predicates.

FIG. 9 shows the processing contents of the warning module 612 in PAD. When the warning function is activated, processing 612
Do 1. Processing 6121 is performed when the constituent predicates constituting the compound predicate include a writing style other than the specified writing style and the writing style of ordinary expression, or when the child writing style is not included.

Execute. Processing 6122 includes writing styles other than the writing style specified as 0 and the writing style of normal expression.

Alternatively, a warning is issued by clearly indicating the end of a single sentence that does not contain the child style.

FIG. 10 shows the processing contents of the writing style conversion module 3705 in PAD. When the writing style conversion function is activated, processes 6141 and 6142 are sequentially performed. In process 6141, all constituent predicates forming a compound predicate, except for the terminal predicate at the end of a simple sentence, are converted into ordinary forms. In process 6142, if the specified writing style is polite form, process 61
43. If it is not a polite form, process 6144 is executed. Processing 6143 converts the final predicate at the end of a simple sentence among the constituent predicates that make up the compound predicate into a polite expression at the specified level by matching the tense of the sentence final predicate that is memorized. do. Processing 6144 converts the final predicate at the end of a simple sentence among the constituent predicates forming the compound predicate into the specified writing style by making it match the tense before conversion.

FIG. 11 shows a colloquial expression/honorific expression detection module 615.
The processing contents are shown in PAD.

When this colloquial expression/honorific expression detection function is activated.

In the process 6151, colloquial expressions, broken expressions, and honorific expressions in the sentence to be processed are detected, and if any, the detected locations are clearly indicated and a warning is issued.

FIG. 12 shows the contents of the simple sentence detection table 51. The simple sentence detection table 51 includes pronouns,
Demonstrative pronouns, prefixes, conjunctions and conjunctive particles are set. In the process of the simple sentence style recognition module 611, in process 6111, the simple sentence class detection table 51 is referred to to identify a word that can be the beginning of a simple sentence.

FIG. 13 shows the contents of the table 52 for verbs and adjectives in the pre-emptive, conjunctive, final, adjunctive, hypothetical, and imperative forms, and adjectives in the pre-emptive, conjunctive, final, adjunctive, and presupposition forms. Verbs and adjectives such as verbs, conjunctive, hypothetical forms, and adjectives such as
The conjunctive, continuation, adjunctive, and hypothetical forms table 52 includes ordinary expressions, polite expressions, etc. of the conjunctive, conjunctive, conjunctive, and hypothetical forms of verbs and adjectives, and the conjunctive, conjunctive, conclusive, adjunctive, and hypothetical endings of verbs and adjectives, etc. The writing style is set. In the process of the simple sentence style recognition module 611, in the process 6112,
Comparing and referring to the table 52 for the first sentence, verbs, adjectives in the pre-existing, conjunctive, final, adjunctive, and imperative forms, and adjectives in the pre-existing, conjunctive, final, adjunctive, and presupposition forms, if both can be connected, Recognized as a short collection. Furthermore, in the processing of the 1 writing style conversion module 614, processing 6143 and 6
At 144, when converting the writing style to the specified target writing style,
If the following particle is converted into a particle that does not connect to anything other than the conjunctive form during stylistic conversion, the following particles may be changed into a particle that does not connect to anything other than the adjunctive form: verbs and adjectives such as verbs, adjectives, continuations, cessation, adjunctions, presuppositions, imperative forms, and adjectives such as verbs, conjunctives, cessation, adjunctions, and presuppositions. With reference to the form table 52, the conjugated forms of verbs and adjectives are changed to conjunctive forms.

FIG. 14 shows the contents of the verb/adjective verb final/adjunctive/imperative form table 53. In the verb/adjective verb final/adjunctive/imperative form table 53, the writing styles provided by the system, such as regular expressions and polite expressions of verb/adjective verb final, adjunctive, and imperative forms, are set separately into present tense and past tense. It is. In the processing of the writing style conversion module 614, processing 6
In steps 143 and 6144, the verb/adjective verb final/adjunctive/imperative form table 53 is referred to to convert the verb/adjective verb final/adjunctive/imperative form into the specified target writing style.

Figure 15 shows auxiliary verbs and particles such as pre-emptive, conjunctive, final, adnominal,
The contents of the hypothetical form table 54 are shown. Auxiliary verbs/particles in the unforeseen, conjunctive, final, adnominal, and hypothetical forms table 54
There are a set of writing styles provided by the system, such as auxiliary verbs, particles, pre-emptive, conjunctive, final, adjunctive, hypothetical forms, 9-pointed expressions, polite expressions, etc.

In the processing of the writing style conversion module 614, processing 614
3 and 6144, refer to the verb/adjective verb termination/adjunction/imperative form table 53 to determine the verb/adjective verb termination/imperative form table 53.
Convert adjunctive/imperative forms to the specified target writing style.

FIG. 16 shows the contents of the colloquial expression/keigo expression table 55. The colloquial expression/keigo expression table 55 is set with colloquial expressions, broken expressions, and honorific expressions in the sentence to be processed. In the processing of the colloquial expression/honorific expression detection module 615, the colloquial expression/keigo expression table 55 is referred to in step 6151 to detect colloquial expressions, broken expressions, and honorific expressions in the processing target sentence.

FIG. 17 shows a table 52 of verbs and adjectives in the pre-emptive, conjunctive, final, adjunctive, imperative, and hypothetical forms, as well as adjectives in the pre-emptive, conjunctive, final, adjunctive, and presuppositional forms, and verbs and adjectives in the final, adjunctive, and imperative forms. When referring to the form table 53 and the preliminary, conjunctive, final, adjunctive, and hypothetical forms of auxiliary verbs and the particle table 54, if the flag for referring to the tone table is set, refer to the tone table 56 to find an example of stylistic conversion. performs the processing.

FIG. 18 shows an example of the original character string text in memory, and FIG. 19 shows an example of the character string text after stylistic conversion according to the present invention. that's all,
Examples have been described in detail.

According to this embodiment, there are the following effects.

Changes, additions, deletions, and corrections of words after writing style conversion are possible by entrusting them to the built-in functions of Japanese word processing, making it easy to improve the quality of Japanese sentences. .

There is no need to always activate the style conversion function and style unification function from the beginning to the end of the string text, and you can specify the beginning and end of a paragraph, the beginning and end of a page,
It is also easily possible to freely change the processing range.

In addition, when document creation is completed, the 0 writing style conversion function automatically detects inconsistencies in writing style. It is also easy to think of a way to use it by asking for confirmation. Furthermore, the stylistic conversion target is not only applied to the final constituent predicate of the compound predicate at the end of a simple sentence, but also proposes the final constituent predicate to the user as the first candidate for conversion. If there is no confirmation from the user, the immediately preceding constituent predicate is presented to the user as a conversion candidate.

If there is no further confirmation, it is easy to consider a method of use in which conversion candidates are sequentially moved to the immediately preceding constituent term. On this occasion.

By providing the user with a non-conversion key, it is possible to stop sequentially moving conversion candidates to the immediately preceding constituent word and easily move the process to the next one. Furthermore, by using it in a machine translation system, etc., it is possible to automatically unify and convert the writing style of Japanese sentences input to the system, and it is also possible to automatically convert the writing style of Japanese sentences output from the system.

Keys that instruct writing style conversion and writing style unification functions can be used with other keys in word processing, such as keys that indicate the end of a paragraph, keys that indicate a page break, or keys that instruct kana character strings in Japanese input using kana-kanji conversion type. It is also easily possible to activate writing style conversion and writing style unification functions in synchronization with each process by pressing a key that instructs conversion from to mixed kanji and kana characters.

〔Effect of the invention〕

According to the present invention, when creating or proofreading a Japanese text, it is possible to activate the writing style conversion function at any time, automatically detect inconsistent writing styles in the created Japanese text, and furthermore, automatically convert the writing style to an arbitrary writing style. Because it can be done, it is effective in improving the efficiency of creating high-quality Japanese sentences. In addition, when generating Japanese input sentences for machine translation, the writing style of the input Japanese sentences can be automatically unified into any writing style, and when generating Japanese output sentences, the writing style of the output Japanese sentences can be automatically changed to any writing style. It also has the effect of being able to be unified and converted. Furthermore, the present invention has the effect of making it possible to incorporate new functions such as detecting ununiformities in writing style and converting writing style, which were not provided in conventional Japanese word processors.

[Brief explanation of the drawing]

Fig. 1 is a functional block diagram of Japanese word processing with a writing style conversion function, which is the main part of the present invention; Fig. 2 is a hardware block diagram of an embodiment of the present invention; Fig. 3 is a board diagram of a keyboard input device; Figure 4 is a word processing control PAD diagram of the present invention, and Figure 5 is a processing control P of the stylistic unification function in Figure 4.
AD diagram, Figure 6 is the stylistic conversion processing control PAD diagram of Figure 4,
Figure 7 is the simple sentence style recognition processing control PAD of Figures 5 and 6.
Figure 8 is a PAD diagram of recognition processing control for compound predicates in Figure 7, Figure 9 is a PAD diagram of warning processing control in Figure 5, and Figure 10 is a PAD diagram of stylistic conversion processing control in Figures 5 and 6. PAD diagram, Figure 11 is the colloquial expression/honorific expression detection processing control P of Figures 5 and 6.
Figure AD, Figure 12 shows an example of the contents of the simple sentence detection table of the present invention, and Figure 13 shows the contents of the present invention's verbs and adjectives such as pre-emptive, conjunctive, final, adjunctive, hypothetical, imperative, and adjectives. FIG. 14 is a diagram showing an example of the content of the verb/adjective verb final/adjunctive/imperative table of the present invention; FIG.
Figure 16 shows an example of the contents of the preliminary, adjunctive, final, adjunctive, hypothetical, and particle tables of auxiliary verbs of the present invention; Figure 16 shows an example of the contents of the colloquial expression/honorific expression table of the present invention; Fig. 17 is a diagram showing an example of the contents of the sound table of the present invention, Fig. 18 is an explanatory diagram showing an example of the character string text in memory according to the present invention, and Fig. 19 is a diagram showing an example of the contents of the text list of the present invention that has been converted. FIG. 2 is an explanatory diagram showing an example of a character string text in a memory. 1 is a word processing control function block; 2 is a word processing control function block;
is a word processing function block group; 3 is a word processing function block group;
A writing style conversion function block group 4 is a character string text storage unit; 5 is a table for detecting the beginning of a sentence and a table for detecting verbs and adjectives;
Conjunction, hypothetical form, and adjectives, continuation, termination, adjunction,
a table storage unit that stores a hypothetical form table, a verb/adjective verb final/adjunctive/imperative form table, an auxiliary verb auxiliary/conjunction/final/adjunctive form/hypothetical form, a particle table, a colloquial expression/honorific expression table, and a tone table; 51 is a table for detecting the beginning of a sentence; 52 is a table of verbs and adjectives in the preliminary, conjunctive, and hypothetical forms; and adjectives in the preliminary, conjunctive, final, adjunctive, and hypothetical forms; and 53 is a table for the final, adjunctive, and imperative forms of verbs and adjectives. A form table 54 is a table of pre-emptive, conjunctive, final, adjunctive, and hypothetical forms of auxiliary verbs, a particle table, 55 is a colloquial expression/honorific expression table, and 56 is a phonetic table. Representative Patent Attorney Katsuo Ogawa ＄2 Illustrator qmanerO1J Ne! Figure 1, figure 12, figure 13, $ttt'$ Hayabusa! Figure 16, Figure 17

Claims (1)

[Claims] 1. Japanese language with writing style conversion function, characterized by comprising a step of inputting and processing Japanese sentences, and a step of unifying sentences constituting the created Japanese character string text into an arbitrary writing style. Word processing methods. 2. In the method according to claim 1, the writing style conversion step includes: detecting colloquial expressions and broken expressions; determining the writing style of sentences forming the sentence;
A writing style conversion function characterized by providing a step of unifying the writing style to the writing style specified by the user when the determined writing style is inconsistent, and a step of automatically unifying the writing to the writing style specified by the user from the beginning. Japanese language processing method. 3. A Japanese language processing method with a writing style conversion function according to claim 1, characterized in that the writing style conversion step includes a step of detecting colloquial expressions and broken expressions. 4. A Japanese language processing method with a writing style conversion function according to claim 1, characterized in that the writing style conversion step includes a step of determining the writing style of sentences forming the sentence. 5. In the method described in claim 1, when the determined writing style is inconsistent in the writing style conversion step,
A Japanese language processing method with a writing style conversion function, characterized by providing a step of unifying the writing style to a writing style specified by a user. 6. In the method described in claim 1, as the writing style conversion step, the writing style is changed from the beginning to the writing style specified by the user.
A Japanese language processing method with a writing style conversion function, characterized by providing a step for automatically unifying sentences. 7. In the method described in claim 2, the step of detecting colloquial expressions and broken expressions includes the step of setting simple sentences constituting a sentence, and the step of detecting each of the complex sentences identified without focusing only on the end of the sentence. A Japanese language processing method with a writing style conversion function, comprising the step of determining the writing style of a simple sentence. 8. In the method described in claim 3, the step of determining the writing style of the sentences that make up the sentence includes the step of identifying simple sentences that make up the sentence, and the step of identifying complex sentences and complex sentences without focusing only on the end of the sentence. 1. A Japanese language processing method with a writing style conversion function, comprising the step of determining the writing style of each simple sentence. 9. In the method set forth in claim 4, the step of identifying a simple sentence constituting a sentence includes the step of identifying a simple sentence based on the relative relationship between the conjugated ending of the predicate and the attached word that can be connected to it. A Japanese language processing method with a writing style conversion function. 10. The method according to claim 5, characterized in that the step of determining the writing style of the identified simple sentence comprises a step of comprehensively determining the writing style of compound predicates that appear consecutively in the simple sentence. A Japanese processing method with a writing style conversion function. 11. In the method described in claim 5, if the determined writing style is inconsistent, the step of unifying the writing style to the writing style specified by the user, and the step of automatically unifying the writing to the writing style specified by the user from the beginning For compound predicates that appear consecutively in a simple sentence, the step does not convert all the words that make up the compound predicate into the writing style specified by the user.
1. A Japanese language processing method with a stylistic conversion function, characterized in that only the last word of the words constituting a compound word is converted into a stylistic style, and the remaining constituent words are converted into ordinary expressions. 12. In the method described in claim 5, if the determined writing style is inconsistent, the step of unifying the writing style to the writing style specified by the user and automatically unifying the writing to the writing style specified by the user from the beginning A Japanese language processing method with a writing style conversion function characterized in that the step is to match tenses in a sentence based on the final tense of a sentence in a polite writing style. 13. In the method described in claim 6, there is provided a step of automatically unifying the text into a writing style specified by the user from the beginning, and in this step, the user can select any one of the multiple writing styles provided by the system. , a Japanese language processing method with a writing style conversion function, which is characterized by selecting a writing style that unifies sentences.