JPH0969102A - Device and method for preparing document - Google Patents

Abstract

PROBLEM TO BE SOLVED: To shorten error detection time by previously registering a character string which tends to be used in error in a Japanese sentence and applying a molpheme analysis only to a sentence including the character string. SOLUTION: A character string retrieving part 105 reads out sentence data from a sentence storing part 102 and retrieves a character string to easily make an error which is previously prepared in a specific character string storing part 104. A sentence including the specific character string is extracted from the read data and an information sort necessary for evaluating the specific character string is referred from an information sort condition storing part 201. A necessary analysis method is selected based upon the referred information sort and a segmented document is analyzed based upon the analysis method. Then, a condition evaluating part 109 evaluates whether the specific character string to be an object is properly used in the sentence or not by the use of the analyzed result and a judgement condition stored in a condition storing part 106.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document preparation device and a document preparation method for detecting an error in a Japanese sentence.

[0002]

2. Description of the Related Art In the conventional error detection in a document preparation apparatus, when performing error detection using morphological analysis or syntactic analysis, the whole sentence is analyzed to detect the error, which is wasteful. It took a lot of processing time. In addition, the analysis method was fixed, and unnecessary information was taken out. This unnecessary information was not referenced and was inefficient.

[0003]

As described above, in the error detection in the conventional document creating apparatus, the analysis is performed only on the sentence in which the error is suspected to be hidden, or only the necessary information is obtained. It was necessary to select the text analysis method to retrieve.

Therefore, the present invention has been made in consideration of the above circumstances, and is a document which solves the above-mentioned problems, evaluates the minimum necessary sentences by the minimum necessary analysis, and shortens the error detection time. An object is to provide a creating device and a document creating method.

[0005]

In order to achieve the above object, the present invention provides a sentence storage unit for storing a sentence, a sentence display unit for displaying the sentence, and a morpheme analysis unit for dividing the sentence into morphemes. , A specific character string storage means for pre-storing a specific character string that is prone to misuse, a character string detection means for detecting the specific character string in a sentence, and each morpheme in a sentence including the specific character string A morpheme condition storage unit that stores a condition to be satisfied, and an error detection unit that evaluates whether or not a morpheme analysis result of a sentence including the specific character string satisfies the condition and detects an error in using the specific character string in the sentence. There is provided a document creation device characterized by performing morphological analysis only on a sentence including a character string that is prone to misuse and detects an error in the sentence.

Further, in order to achieve the above object, the present invention achieves the above-mentioned object, a sentence storage means for storing a sentence, a sentence display means for displaying the sentence, a morpheme analysis means for dividing the sentence into morphemes, and a morpheme analysis result. A syntactic analysis unit that obtains a sentence structure from a specific character string storage unit that stores in advance a specific character string that is prone to misuse; a character string detection unit that detects the specific character string from a sentence; Morphological condition storage means for storing a condition to be satisfied by each morpheme in a sentence including a string, and use of a specific character string in the sentence by evaluating whether or not a morphological analysis result of a sentence including the specific character string satisfies the condition A document comprising an error detecting means for detecting an error, performing morphological analysis and syntactic analysis only on a sentence including a character string that is likely to cause a usage error, and detecting an error in the sentence In the adult system.

Further, in order to achieve the above object, the present invention provides a plurality of sentence storage means for storing a sentence, a sentence display means for displaying the sentence, and a plurality of information types for obtaining each information type required as a morphological analysis result. Morphological analysis means, a plurality of syntactic analysis means for obtaining each required information type as a syntactic analysis result, a specific character string storage means for pre-storing a specific character string that is prone to misuse, and from the sentence A specific character string detection means for detecting the specific character string, a condition storage means for storing the specific character string and the condition to be satisfied by the sentence containing it,
An information type storage unit that stores an information type necessary for evaluating whether or not the condition is satisfied in association with a specific character string;
An error detecting unit that evaluates whether or not the condition is satisfied to detect a usage error of a specific character string, and targets only a sentence including a character string that is prone to a usage error, and a minimum necessary morphological analysis and A document creating apparatus characterized by performing syntax analysis and detecting an error in a sentence.

Further, in order to achieve the above-mentioned object, the present invention stores in advance a specific character string which is prone to misuse, and stores a condition of a morpheme that a sentence including the specific character string must satisfy. Detects the specific character string in the sentence, detects whether the specific character string in the sentence is used by evaluating whether or not the condition is satisfied from the result of morphological analysis of the sentence including the specific character string, and is a character that is prone to use error A document creating method is characterized by performing morphological analysis only on a sentence including a sequence and detecting an error in the sentence.

Further, in order to achieve the above object, the present invention stores in advance a specific character string which is likely to be misused, and stores a condition of a morpheme that a sentence including the specific character string must satisfy. Detects the specific character string in the sentence, detects whether the specific character string in the sentence is used by evaluating whether or not the condition is satisfied from the result of morphological analysis of the sentence including the specific character string, and is a character that is prone to use error A document creating method is characterized by performing morphological analysis and syntactic analysis only on a sentence including a string to detect an error in the sentence.

Further, in order to achieve the above object, the present invention prepares a plurality of morpheme analysis methods for obtaining each information type required as a morpheme analysis result, and obtains each information type required as a syntactic analysis result. Prepare a plurality of parsing methods for the purpose, pre-store a specific character string that is prone to misuse, detect the specific character string in the sentence, the condition that the specific character string and the sentence containing it should be satisfied Memorize
Select the syntax analysis method necessary to evaluate whether the condition is satisfied and apply it to the sentence, evaluate whether the condition is satisfied and perform morphological analysis and syntactic analysis of the specific character string to check the error in the sentence. A document creation method characterized by detecting.

[0011]

With the above structure, a specific character string that is likely to be misused in a Japanese sentence is registered in advance, and morphological analysis is performed by subjecting only the sentence containing the specific character string to morphological analysis. You can save time. Further, by selecting the analysis method according to the information type necessary for determining whether or not the character string is correctly used, it is possible to detect an error without performing unnecessary analysis.

[0012]

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 one embodiment of the present invention. In FIG. 1, an input unit 101 is a part where a user inputs characters or commands, and specifically, is a keyboard, a pen, a mouse, or the like.

The sentence storage unit 102 is a unit for storing a sentence to be processed, and is specifically a floppy disk device, a hard disk device, a memory or the like. Output unit 10
Reference numeral 3 denotes a target sentence or a specific character string portion in the document, and a portion for displaying / outputting a result or a message after the evaluation, specifically, a display, a printer or the like.

The specific character string storage unit 104 is a unit for storing a character string containing an expression that is likely to make an error. The character string search unit 105 is a unit that searches for a character string stored in the specific character string storage unit 104 from the text stored in the text storage unit 102 and extracts one sentence including the character string.

The condition storage unit 106 is a specific character string storage unit 1.
When the character string stored in 04 is correctly used, the condition to be satisfied by the sentence is stored for each character string. An information type condition storage unit 201 to be referred to before analysis is further added here.

The morphological analysis unit 107 includes a character string search unit 10
This is a part that analyzes the sentence extracted in 5 in units of morphemes and outputs various grammatical information. Here, it is further subdivided by function and divided into a part-of-speech information analysis unit 301 and a fine part-of-speech analysis unit 302. For the subdivided functions,
Normally, all are used, and if there is a special designation from the information type condition storage unit 201, that function is used.

The syntax analysis unit 108 includes a character string search unit 105.
It is a part that analyzes the sentence extracted in and outputs the syntactic structure. Here, it is further subdivided by function and divided into a basic analysis unit 401 and a dependency analysis unit 402. Normally, all the subdivided functions are used, and if there is a special designation from the information seed condition storage unit 201, the function is used.

The condition evaluation unit 109 is a morpheme analysis unit 107.
It is a part for evaluating whether or not the target sentence is correct according to the analysis result obtained from the syntax analysis part 108 and the condition requested from the condition storage part 106. The evaluation result is sent to the output unit 103. The information type condition storage unit 201
Information types necessary for evaluation by the condition evaluation unit 109 are stored, and what kind of analysis is performed on a target sentence is specified to the morpheme analysis unit 107 and the syntax analysis unit 108.

In the above-mentioned configuration, it is configured to operate as follows. FIG. 2 is a flowchart showing the operation of this embodiment. In FIG. 2, the step is to read the text data from the text storage unit 102 and display the text.

Step S2 is a step of searching the text data read in step S1 for a character string prepared in advance in the specific character string storage unit 104 and prone to make an error. Step S3 is a step of extracting from the data a sentence that is found to contain a specific character string as a result of the search in step S2.

Step S4 is a step of determining whether or not the user selects the analysis unit when the analysis unit can realize a plurality of processing methods. If the user desires to reduce the analysis time, the process proceeds to step S6, and if not, the process proceeds to step S5.

When the analysis unit supports only one type of operation, the process skips from step S3 to step S5. Step S5 is a method of processing the one sentence cut out in step S3 by a usual analysis method. Here, if the analysis unit has a plurality of processing methods, it is necessary to define whether to advance what kind of analysis. Certain character strings include those that use multiple pieces of information to make conditional judgments, so when judging all character strings, it should be set so that as much information as necessary can be obtained. is there.

Step S6 is a step of referring from the information type condition storage unit the information type necessary for evaluating the specific character string in one sentence cut out in step S3. Step S7 is a step of selecting a necessary analysis method based on the information type referred to in step S6.

Step S8 is a step of analyzing the sentence cut out in step S3 based on the analysis method selected in step S7. In step S9, the analysis result obtained in step S5 or S8 and the condition storage unit 106 are stored.
Is a step of evaluating whether or not the target specific character string is appropriately used in one sentence by using the determination condition of.

Step S10 is a step of presenting the determination result of step S9 to the user. The specific operation of this embodiment will be described below.

The text to be processed is read and displayed on the screen (step S1). Next, a specific character string that is easy to make an error prepared in advance is searched from the sentence (step S2). If found in the sentence, one sentence including the character string is cut out from the sentence data (step S3). Only this one sentence is the target of the subsequent processing, and the other sentences are not analyzed. By this method, an error can be detected in a sentence without wasting processing time. The specific character string has a determination condition and an information type in the condition storage unit correspondingly. The judgment condition is a condition used for judging whether or not the character string is correctly used, and the information type is analysis information of a sentence necessary for the judgment. An example of a specific character string, its determination condition, and information type is shown below. (Example) 1) Other than specific character string Judgment condition: "The position is not the beginning of a sentence": "There is no particle in front" Information amount: Part-of-speech information 2) Specific character string: Even in dream Judgment condition: "Dependency relationship" The verb in is a continuation form. "It is accompanied by the negative adjunct" not "." Information amount: part-of-speech information, dependency information If the analysis unit consists of one processing block, the cut-out sentence is The analysis is performed as it is (step S5). if,
When there are a plurality of analysis units and the processing for obtaining the data required by the information type can be additionally selected, the processing necessary for improving the processing speed can be selected (step S
4). For example, if only part-of-speech information is required, syntactic analysis is not required, so by performing only morphological analysis,
Saves unnecessary processing time. In addition, if dependency information and part-of-speech information are needed, refined part-of-speech (part-of-speech information in which more detailed information such as "animal noun" or "plant noun" is added in addition to "noun") is not needed, and part-of-speech information is obtained. It is sufficient to perform morphological analysis and syntactic analysis including dependency.
When the user desires to shorten the processing time, the information type stored in the condition storage unit is referred to (step S6), the necessary analysis method is selected (step S7), and the analysis is executed (step S8). Then, the correctness of the sentence is evaluated from the analysis result and the judgment condition (step S9). Example 1 above)
An example of analysis and evaluation in the case of an incorrect sentence containing the character string of is shown below. (Example sentence) It was a blind spot.

Morphological analysis result: [that] (pronoun) "ha" (particle) "other" (noun) → specific character string "na" (unknown word) "blind point" (noun) "da" (particle) "ta ”(Particle) Condition determination result: A particle exists before“ other than ”.

Therefore, this sentence is a misuse. For the sentence judged to be erroneous as described above, the user is informed accordingly (step S10).

[0029]

As described above in detail, according to the present invention, by reducing the processing amount of sentence analysis, the time required for error detection can be shortened.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention.

FIG. 2 is a flowchart for explaining the operation of the embodiment.

[Explanation of symbols]

101 ... Input unit, 102 ... Sentence storage unit, 103 ... Output unit, 104 ... Specific character string storage unit, 105 ... Character string search unit, 106 ... Condition storage unit, 107 ... Morphological analysis unit, 10
8 ... Syntax analysis unit, 109 ... Condition evaluation unit, 201 ... Information type condition storage unit.

Claims (6)

[Claims] 1. A sentence storage unit that stores a sentence, a sentence display unit that displays the sentence, a morphological analysis unit that divides the sentence into morphemes, and a specific character string that is prone to misuse is stored in advance. Specific character string storage means to be placed, character string detection means to detect the specific character string from the sentence, morpheme condition storage means to store conditions to be satisfied by each morpheme in the sentence including the specific character string, the specific character Equipped with an error detection means that evaluates whether the morphological analysis result of a sentence including a string satisfies the condition and detects an error in using a specific character string in the sentence, and only a sentence including a character string that is prone to use error A document creation device characterized by performing morphological analysis on a target and detecting an error in a sentence. 2. A sentence storage unit that stores a sentence, a sentence display unit that displays the sentence, a morpheme analysis unit that divides the sentence into morphemes, and a syntactic analysis unit that obtains a sentence structure from the morpheme analysis result. Specific character string storage means for storing in advance a specific character string that is prone to misuse, character string detection means for detecting the specific character string in the sentence, and each morpheme in the sentence including the specific character string is satisfied A morpheme condition storage unit that stores a power condition; and an error detection unit that evaluates whether or not a morpheme analysis result of a sentence including the specific character string satisfies the condition and detects a usage error of the specific character string in the sentence. A document creation apparatus, which performs morphological analysis and syntactic analysis only on a sentence containing a character string that is prone to misuse, and detects an error in the sentence. 3. A sentence storage means for storing a sentence, a sentence display means for displaying the sentence, a plurality of morphological analysis means for obtaining each information type required as a morphological analysis result, and a syntactic analysis result. A plurality of parsing means for obtaining each information type, a specific character string storage means for pre-storing a specific character string that is likely to be misused, and a specific character string for detecting the specific character string in the sentence Detecting means, condition storing means for storing a specific character string and a condition to be satisfied by a sentence including the specific character string, and an information type for storing an information type necessary for evaluating whether or not the condition is satisfied in association with the specific character string A storage unit and an error detection unit that evaluates whether or not the above condition is satisfied to detect a usage error of a specific character string are provided. A document creation apparatus characterized by performing a small amount of morphological analysis and syntactic analysis to detect an error in a sentence. 4. A specific character string that is prone to misuse is stored in advance, a morpheme condition to be satisfied by a sentence including the specific character string is stored, and the specific character string is detected from the sentence, From the morphological analysis result of the sentence containing the specific character string, it is evaluated whether or not the condition is satisfied to detect the use error of the specific character string in the sentence, and the morpheme is applied only to the sentence containing the character string that is prone to the use error. A method of creating a document, characterized by performing an analysis and detecting an error in a sentence. 5. A specific character string which is prone to misuse is stored in advance, a morpheme condition to be satisfied by a sentence including the specific character string is stored, and the specific character string is detected from the sentence, From the morphological analysis result of the sentence including the specific character string, it is evaluated whether or not the condition is satisfied to detect the use error of the specific character string in the sentence, and the morpheme is applied only to the sentence including the character string that is prone to the use error. A document creation method characterized by performing an analysis and a syntax analysis to detect an error in a sentence. 6. A plurality of morpheme analysis methods for obtaining each information type required as a morpheme analysis result, and a plurality of syntax analysis methods for obtaining each information type required as a syntactic analysis result are prepared. , A specific character string that is prone to misuse is stored in advance, the specific character string is detected from the sentence, and the condition that the specific character string and the sentence including it are to be satisfied is stored, and the condition is satisfied. Select the syntax analysis method necessary to evaluate whether or not it is applied to the sentence, evaluate whether the condition is satisfied, perform morphological analysis and syntax analysis of the specific character string, and detect the error in the sentence. Characteristic document creation method.