JPH0916595A - Text editing device and text editing method - Google Patents

Abstract

PURPOSE: To detect a sentence to be an original error by checking the possibility of other conjugations even when the sentence is grammatically right. CONSTITUTION: The sentence data inputted through an input part 11 is processed by a KANA (Japanese syllabary)/KANJI (Chinese character) converting part 14 and an editing part 13 and is stored in the sentence storage area 16a of a storage part 16. At this stage, when a 'sentence elaboration' function is requested by the input part 11, a control part 12 transmits the sentence data within a designated range to a sentence elaboration control part 15 and executes a sentence elaboration processing. By analyzing document data to be an elaboration object through an analysis part 20 in a sentence elaboration control part 15, the error of the conjugation is detected by a conjugation error detection part 22. The sentence data inputted at this stage is analyzed by a morpheme level and the error of the conjugation is detected based on a prescribed conjugation changing rule for each morpheme. The sentence elaboration control part 15 generates the sentence in which the conjugation is changed by a generation part 29 and displays this sentence on a display part, 18.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a text selection device and a text selection method for selecting text.

[0002]

2. Description of the Related Art Conventionally, some Japanese word processors have a sentence revision function for checking whether a created sentence is grammatically correct. Usually, this type of sentence revision function is configured to detect a grammatically incorrect portion by performing morphological analysis and syntactic analysis on a created sentence.

[0003]

[Problems to be Solved by the Invention] By the way, there are many sentences written by foreigners who are not used to Japanese, even if they are supposed to be erroneous even if they are grammatically correct. . For example, a sentence such as "I learned how to speak" is correct. This is because "kata" was judged as a noun. That is, the word "kata" can be used as both a suffix and a noun. Originally,
Although "kata" should be used as a suffix to be "speaking style", it should be "speaking style" when used as a noun, which is not a grammatical error.

The present invention has been made in view of the above points, and it is possible to detect a sentence that is supposed to be an error by checking the possibility of another conjugation even if the grammar is correct. The present invention relates to a text reviewing device and a text reviewing method.

[0005]

A text selection device of the present invention analyzes input data for inputting text data and text data input by the input means at a morpheme level, and determines a predetermined morpheme for each morpheme. The conjugation type error detection means for detecting an conjugation type error based on the conjugation type change rule, the alteration means for changing the conjugation type detected by the conjugation type error detection means to the correct one, and the alteration means It is characterized by comprising a display means for displaying the inflectional form.

[0006]

According to the above construction, the sentence data to be subject to sentence revision is subjected to morpheme analysis, and conjugation errors are detected for each morpheme based on a predetermined conjugation change rule. Then, when there is a corresponding morpheme, the inflectional form is changed to the correct form.

[0007]

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of a text selection device according to an embodiment of the present invention. In FIG. 1, the input unit 1
Reference numeral 1 denotes an input device such as a keyboard, for inputting data and giving instructions. The control unit 12
This is for controlling the entire apparatus, and in particular, here, the text revision control unit 15 is activated to execute the text revision processing as described with reference to FIGS. 2 and 3.

The editing unit 13 edits text data.
The kana-kanji conversion unit 14 performs kana-kanji conversion processing. Under the control of the control unit 12, the text revision control unit 15 controls the analysis unit 2
0, the utilization error detection unit 22 and the generation unit 24 perform sentence revision processing.

The storage unit 16 is composed of, for example, a RAM, and has a text storage area 16a for storing the created text data and a rule learning area 16b for storing the learning data of the inflectional form change rules.

The display control section 17 controls display. The display unit 18 is composed of a display device such as a CRT (CathodeRay Tube) and displays data. Document file part 1
A storage device 9 such as a hard disk device stores a plurality of document files.

The analysis unit 20 analyzes text data (grammatical analysis such as morphological analysis and syntactic analysis). Sentence analysis dictionary 21
Is a dictionary used for analyzing text data, and stores various information such as headings, parts of speech, connection information, and attribute information (properties) of words.

The in-use error detector 22 detects in-use errors. The error detection rule dictionary 23 stores error detection rule data. The generation unit 24 generates a sentence in which the inflectional form is changed.

Next, the operation of the embodiment will be described. The text data input through the input unit 11 is processed by the Kana-Kanji conversion unit 14 and the editing unit 13, and is stored in the text storage area 16 a of the storage unit 16. Here, when the "text editing" function is requested from the input unit 11, the control unit 12 sends the text data within the specified range to the text editing control unit 15 to execute the text editing process.

The text revision control unit 15 analyzes the text data to be revised through the analysis unit 20,
The practical error detection unit 22 detects a practical error.
Then, the sentence revision control unit 15 generates a sentence in which the inflectional form is changed by the generation unit 24, and displays the sentence on the display unit 18.

Specific processing operations will be described with reference to the flowcharts shown in FIGS. 2 and 3 and the example sentences shown in FIGS. 4 and 5. As shown in FIG. 4, in the case of sentence revision of sentence data such as “I learned how to speak”, the sentence revision control unit 15 first analyzes the sentence data at the morpheme level through the analysis unit 20 as follows. (Step S11).

“Story”… verb, syllable five-stage conjugation, adnominal form “kata”… noun “o”… particle “traditional”… verb, syllable five-stage conjugation, continuous form “ta”… particle “.”… Punctuation Next, the text revision control unit 15 sets the counter to "0" (step S12), and sends the morpheme "talk" indicated by the counter to the inflectional error detection unit 22, so that The change rule is applied (step S13).

Rule 1 ... "*" is a verb (adnominal form) "* +"
When "1" is "kata (noun)", the use of "*" is in the adjunct form. (* Indicates the target word, and * + 1 indicates the word next to the target word.) In this case, the target word "story" is a verb or adnominal form,
The following word applies to Rule 1 above because it is a noun and is a noun. In this way, when it is applied to the inflectional form change rule (step S14), the text revision control unit 15 changes the inflectional form to the correct form according to the rule (step S1).
5). That is, in the above example, the inflectional form of "talk" is changed from the continuous form to the continuous form.

In this way, the text revision control unit 15 updates the counter (step S16) and performs the same processing as above for each morpheme (step S17). As a result, in the above example, only "talk" is detected. Since "speaking" is changed to the continuous form, "learning how to speak" is obtained as a recommended candidate for correcting an error, as shown in FIG. 4 (b).

Here, when the inflectional form is changed (step S18), it is possible to judge whether the changed sentence data is grammatically correct or not before pointing it out to the user.

That is, the sentence revision control unit 15 sends the changed sentence data to the analysis unit 20 to grammatically analyze the entire sentence data (step S19). As a result, in the sentence data of FIG. 4 (b), "story" ... verb, Sa line five-stage conjugation, continuous form "kata" ... suffix "o" ... particle "Study" ... verb, wa line five-stage utilization, continuous form ""..." Particles "." ... Phrases are obtained, and the grammatically correct results are obtained.

Also, in this case, "speaking" is changed to "speaking"
Since it has been corrected to, when pointing out to the user the error in its usage, search for similar errors that exist after this sentence,
It is possible to point out at once including that part.

That is, for example, as shown in FIG. 5 (a), if there is sentence data such as "I learned how to speak. However, it was more difficult to learn how to speak than to learn how to write.""Speaking" in "Speaking"
When corrected to, the sentence revision control unit 15 retrieves the same "speaking person" from the subsequent sentences (step S2).
1).

As a result, if the "speaker" also exists in the subsequent sentences, the sentence revision control unit 15 displays the correction candidates as shown in FIG. The error of is collectively pointed out to the user (step S24). If the "speaking person" does not exist in the subsequent sentence, the sentence revision control unit 15 points out only that (step S23).

Further, after the text is refined, if the result is that the inflectional change rule applied to the text data is only "rule 1", this is used as rule learning data in the storage unit 16
By storing it in the rule learning area 16b, the error can be detected from the next time using only "rule 1". That is, in step S13 described above, error detection can be performed using only "rule 1" without searching all the rules registered in the error detection rule dictionary 23. As a result, unnecessary processing can be suppressed, and error detection suitable for the user can be efficiently performed.

[0025]

As described above, according to the present invention, morphological analysis is performed on sentence data that is a target of sentence revision, and conjugation errors are detected based on a predetermined conjugation change rule for each morpheme. Since the inflectional form is changed to the correct form, even if it is grammatically correct, it is possible to detect a sentence that is originally an error due to the possibility of another inflectional form.

[Brief description of the drawings]

FIG. 1 is a block diagram showing the configuration of a text selection device according to an embodiment of the present invention.

FIG. 2 is a flowchart showing the operation of a text selection process in the same embodiment.

FIG. 3 is a flowchart showing the operation of a text selection process in the same embodiment.

FIG. 4 is a diagram showing an example of a sentence which is a target of sentence revision in the embodiment.

FIG. 5 is a diagram showing an example of a sentence which is a target of sentence revision in the embodiment.

[Explanation of symbols]

 11 ... Input part, 15 ... Sentence revision control part, 16 ... Storage part, 16a ... Sentence storage area, 16b ... Rule learning area, 18 ... Display part, 20 ... Analysis part, 21 ... Sentence analysis dictionary, 22 ... Inflectional error Detection unit, 23 ... Error detection rule dictionary, 24 ... Generation unit.

Claims (6)

[Claims] 1. An input means for inputting text data, and the text data input by this input means is analyzed at a morpheme level, and morphological errors are detected based on a predetermined morphological change rule for each morpheme. The present invention further comprises: an inflectional type error detecting means for detecting; a changing means for changing the inflectional type detected by the inflectional type error detecting means to a correct form; and a display means for displaying the inflectional type changed by the changing means. A text reviewing device. 2. The grammatical analysis means for grammatically analyzing the sentence data of which the inflectional form is changed by the changing means, and for judging whether or not the inflectional form change is grammatically correct, further comprising: the display means. 2. The sentence reviewing device according to claim 1, wherein when the grammar determining means determines that the grammar is correct, the modified inflectional form is displayed. 3. When the conjugation-type error detection unit detects an conjugation-type error, the retrieval unit further includes a retrieval unit for retrieving a similar error from the text data, and the display unit includes the retrieval unit. 2. The text reviewing device according to claim 1, wherein the modified usage forms including the similar errors retrieved by are displayed together. 4. The morpheme of the morpheme is correct by analyzing the sentence data to be sentence-reviewed at the morpheme level and detecting an error of the morpheme based on a predetermined morpheme change rule for each morpheme. A text revision method characterized by being adapted to a shape. 5. The grammatical analysis of the sentence data in which the inflectional form is changed to determine whether or not the change in the inflectional form is grammatically correct.
How to improve the written text. 6. The sentence revision method according to claim 1, wherein when an error in the inflectional form is detected, a similar error is searched from the sentence data.