JP2887327B2 - Japanese document proofreading device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a Japanese document proofreading device that points out and corrects inappropriate expressions and errors in a document.

2. Description of the Related Art FIG. 5 shows a configuration diagram of a conventional calibration device. This device includes a text input unit 1, a text storage unit 2, an analysis unit 3,
An error detection unit 4 and an error display unit 5 are provided. This device stores an input Japanese sentence character string, and when an error detection is instructed for the stored Japanese document character string, analyzes the Japanese document character string and uses the result. To detect potential errors and display them as locations to be calibrated. Here, the detected error includes “error of term”, “ununiform expression”, and the like.

Problems to be Solved by the Invention Expressions of natural languages including Japanese are various. One of the Japanese language phenomena is called functional verb expression. A functional verb is a verb in which the original meaning of the verb is weakened and only the auxiliary meaning is expressed, such as "do" in "investigate" or "show" in "show increase". Therefore, almost the same meaning can be described between an expression using a functional verb (referred to as a functional verb expression) and an expression not using the functional verb (referred to as a non-functional verb expression). For example, in the above-described example, "investigate" can be replaced with "investigate" and "show increase" can be replaced with "increase" without any significant semantic difference. In other words, due to the existence of such a functional verb, almost the same semantic content can be described by an expression using the functional verb or not. The details of functional verb expressions are described in detail in "Muraki": "On Japanese Functional Verb Expressions", Japanese Language Research Report 65, 2,1983.3 ". In view of this point, the present invention considers the functions included in sentences. By detecting verb expressions and non-functional verb expressions and pointing out if they are mixed in the document, or by converting functional verb expressions and non-functional verb expressions to each other, It is an object of the present invention to provide a Japanese document proofreading apparatus that can easily perform a work of unifying Japanese documents and that can easily perform the operation and designation.

Means for Solving the Problems The invention according to claim 1 of the present application provides a text storage unit that stores a sentence character string, a morphological analysis unit that recognizes a morpheme that constitutes the sentence character string in the text storage unit, and at least its part of speech. A Japanese dictionary unit that holds many Japanese words with information, a functional verb dictionary unit that holds many verbs that can be functional verbs, and each morpheme becomes a functional verb expression using the data of the functional verb dictionary unit. Check if it is a verb,
A functional verb determining unit that performs a syntactic analysis on the sentence containing the verb using the data of the Japanese dictionary unit to determine whether the verb is a functional verb, and that all verb expressions in the sentence are functional verb expressions It is equipped with a non-unification judging unit that checks whether or not there is, and a display unit that displays the corresponding expression part when there is inconsistency. It is a Japanese document construction device characterized by displaying it.

According to the second aspect of the present invention, there is provided a text storage unit that stores a sentence character string, a morphological analysis unit that recognizes a morpheme constituting a document character string in the text storage unit, and a large number of Japanese words together with at least the part of speech information. A Japanese dictionary part to hold, a functional verb dictionary part to hold many verbs that can be functional verbs, and using the data of the functional verb dictionary part, check whether each morpheme is a verb that can be a functional verb expression, A functional verb determining unit that performs a syntax analysis on a sentence including a verb using the data of the Japanese dictionary unit to determine whether the verb is a functional verb, and an expression converting unit that converts an expression related to the functional verb A Japanese document proofreading apparatus, comprising: converting an expression relating to a functional verb in a Japanese document into a non-functional verb expression or a functional verb expression as necessary.

Further, the invention according to claim 3 is characterized in that, in the invention according to claim 2, the expression conversion unit converts an expression using a functional verb into an expression not using a functional verb.

Further, the invention according to claim 4 is characterized in that, in the invention according to claim 2, the expression conversion unit converts an expression not using a functional verb into an expression using a functional verb.

The invention according to claim 1 of the present application analyzes a Japanese sentence with the above configuration, detects a functional verb expression and a non-functional verb expression included in the sentence, and determines whether or not they are mixed in the document. Is determined, and if mixed, the location in the document is displayed. Further, the invention according to claim 2 of the present application analyzes Japanese sentences by the above configuration,
When a functional verb expression and a non-functional verb expression included in a sentence are detected and a conversion of the sentence expression is instructed, the verb expression in the sentence is converted into a functional verb expression or a non-functional verb expression according to the instruction.

Embodiment (Embodiment 1) FIG. 1 shows a configuration diagram of a calibration apparatus according to a first embodiment of the present invention. In FIG. 1, reference numeral 10 denotes a text input unit for inputting a Japanese sentence character string. For this, a word processor, an optical character reader, or the like is used. Reference numeral 11 denotes a text storage unit that holds an input Japanese sentence character string. Reference numeral 12 denotes a morphological analysis unit that performs morphological analysis on Japanese sentences stored in the text storage unit 11 and identifies morphemes that constitute each sentence. Reference numeral 13 denotes a Japanese dictionary unit which stores a large number of Japanese word information, which is information for the morphological analysis unit 12 to analyze a Japanese document, together with at least the part of speech information. 14 is a functional verb determination unit,
For all morphemes recognized as verbs by the morphological analysis unit 12, it is determined whether or not they are functional verbs. Fifteen
Is a functional verb dictionary unit, which is a dictionary used when the functional verb determination unit 14 determines whether or not it is a functional verb, and records a lot of information on verbs that can be functional verbs. Reference numeral 16 denotes an inconsistency determination unit that determines whether all verb expressions included in the document are functional verb expressions or non-functional verb expressions, or whether they are mixed. A display unit 17 displays the verb expression on the screen together with contexts before and after when it is determined that the functional verb expression and the non-functional verb expression are mixed by the inconsistency determining unit 16. FIG. 6 is a block diagram showing an example of a hardware configuration for realizing the Japanese document proofreading apparatus shown in FIG. Text input section in FIG.
Reference numeral 10 denotes an input device 105.
Is the output device 104, the text storage unit 11 is the RAM 102, the Japanese dictionary unit 13 and the functional verb dictionary unit 15 are the external storage device 103, the morphological analysis unit 12, the functional verb determination unit 14, and the unification determination unit is the CPU 1.
01 and a RAM 102 for storing processing results.

The operation of the Japanese document proofreading apparatus of the present embodiment configured as described above will be described below.

(1) The operator inputs a Japanese sentence character string from the text input unit 10. The input Japanese character string is stored in the text storage unit 11. Then, in order to proofread the input sentence, it tries to look for inconsistency in the verb expression in the sentence and gives an instruction to that effect.

(2) The morphological analysis unit 12 analyzes the Japanese sentence character string held in the text storage unit 11 while referring to the Japanese dictionary unit 13, and recognizes the morpheme constituting the character string.

(3) The functional verb determination unit 14 determines whether or not all morphemes recognized as verbs by the morphological analysis unit 12 are functional verbs. The determination is made based on whether or not the verb in the sentence exists in the functional verb dictionary 15 and whether the main noun of the noun phrase which forms the object dominated by the verb is a sa noun. For example, for the expression "indicating a large increase", the verb is "indicate", and the main noun of the noun phrase which forms the dominant object is "increase". At this time, “show” is in the functional verb dictionary and “increase”
Is a paranoun, it is determined that this is a functional verb expression. Those that cannot be determined as functional verb expressions by the above algorithm are determined as non-functional verb expressions.

(4) For all verb expressions included in the text storage unit 11, the judgment shown in (3) is performed, and whether all are verb expressions, all are non-verb expressions, or are they mixed? Determine whether Further, when they are mixed, the frequency is calculated and output.

(5) Based on the information output from the inconsistency determination unit 16, the display unit 17 compares the frequencies of the functional verb expression and the non-functional verb expression when they are mixed, and compares the frequencies of the expressions. Display a list of expressions on a display device. At that time, the position information in the document and the contexts before and after are displayed so that the location can be easily identified. For example, the sentence shown in FIG. 2 is displayed in the format shown in FIG. 3, and if all verb expressions are unified into functional verb expressions or non-functional verb expressions, this fact is indicated. indicate.

(6) The operator performs a document proofreading operation while viewing the information displayed on the display unit 17.

As described above, in the present embodiment, it is possible to easily determine whether or not a functional verb expression and a non-functional verb expression are mixed in a sentence text. Further, it is possible to display, in an easy-to-understand format, a list of expressions having a lower frequency, which are effective information for correcting expressions in a sentence. This makes it easy to find non-uniform expressions in the text, and the operator can quickly and surely unify the verb expressions in the text. In this embodiment,
As a function verb detection algorithm, it was determined whether the verb in the sentence exists in the function verb dictionary, and whether the main noun of the noun phrase that forms the object dominated by the verb is a sa-variable noun. A determination algorithm may be used. In the present embodiment, when the functional verb expression and the non-functional verb expression are mixed, the part of the expression having the lower frequency is displayed, but the expression having the higher frequency is displayed, Needless to say, all expressions may be displayed. In the present embodiment, the corresponding verb expression is displayed together with the position information in the sentence and the context before and after the display, but may be displayed together with other information.

(Embodiment 2) FIG. 4 shows a configuration diagram of a calibration device according to a second embodiment of the present invention. In FIG. 4, reference numeral 20 denotes a text input unit for inputting a Japanese sentence character string. For this, a word processor, an optical character reader, or the like is used. 21 is a text editing unit for editing a Japanese sentence character string. For this, a word processor device or the like is used. Reference numeral 22 denotes a text storage unit which holds an input Japanese sentence character string. 23 is a morphological analysis unit, a text storage unit
A morphological analysis is performed on the Japanese sentences stored in 22, and morphemes constituting each sentence are recognized. Reference numeral 24 denotes a Japanese dictionary unit which stores a large number of Japanese word information, which is information for the morphological analysis unit 23 to analyze Japanese sentences, together with at least the part of speech information. Reference numeral 25 denotes a functional verb determination unit, which determines whether or not all morphemes recognized as verbs by the morphological analysis unit 23 are functional verbs. Reference numeral 26 denotes a functional verb dictionary unit, which is a dictionary used when the functional verb determination unit 25 determines whether or not it is a functional verb, and records a large number of information on verbs that can be functional verbs. Reference numeral 27 denotes a sentence expression designating unit for inputting conditions to be satisfied by the converted sentence when the current sentence expression is converted into another sentence expression. Here, a lower limit value and an upper limit value, which are limit values of the total number of characters of the document, are specified. Reference numeral 28 denotes an expression conversion instruction unit, which determines whether to convert a functional verb expression to a non-functional verb expression or vice versa, based on the content specified by the sentence expression specifying unit 27 and the current text information. . A non-functional verb expression unit 29 converts a functional verb expression in a sentence into a non-functional verb expression and changes information in a text storage unit. 30
Is a functional verb expression unit that converts a non-functional verb expression in a sentence into a functional verb expression and changes information in a text storage unit. FIG. 6 is a block diagram showing an example of a hardware configuration for realizing the Japanese sentence proofreading device shown in FIG. The text input unit 20, the text editing unit 21, and the sentence expression designating unit 27 in FIG. 4 are constituted by an input device 105. Similarly, the text storage unit 22 is a RAM 102, the Japanese dictionary unit 24 and the functional verb dictionary unit 26 are external storages. Device 103, morphological analysis unit 23,
The functional verb determination unit 25, the expression conversion instructing unit 28, the non-functional verb expression unit 29, and the functional verb expression unit 30 are each composed of a CPU 101 and a RAM 102 for storing processing results.

The operation of the calibration device of the present embodiment configured as described above will be described below.

(1) The operator inputs a Japanese sentence character string from the text input unit 20. The input Japanese character string is stored in text storage unit 22.

(2) Instruct to correct the input sentence. Here, it is assumed that a document, such as a manuscript, having a limit on the length of the document has been created in advance, and when the document has been completely created, the limit value of the length to be satisfied by the document is specified in the text expression designation section.
Specify with 27.

(3) The morphological analysis unit 23 analyzes the Japanese sentence character string held in the text storage unit 22 while referring to the Japanese dictionary unit 24, and recognizes the morpheme constituting the character string.

(4) The functional verb determination unit 25 determines whether or not all morphemes recognized as verbs by the morphological analysis unit 23 are functional verbs. The determination is made based on whether or not the verb in the sentence exists in the functional verb dictionary 26, and whether the main noun of the noun phrase forming the object dominated by the verb is a sa noun. For example, for the expression “showing a large increase”, the verb is “show”, and the main noun of the noun phrase that forms the object dominated by the verb is “increase”. At this time, “show” is in the functional verb dictionary and “increase”
Is a paranoun, it is determined that this is a functional verb expression. Those that cannot be determined as functional verb expressions by the above algorithm are determined as non-functional verb expressions.

(5) The expression conversion instructing unit 28 compares the limit value of the length designated by the sentence expression designating unit 27 with the length of the actual sentence stored in the text storage unit 22, and determines the length of the actual sentence. If the length is longer than the upper limit, the function verb expression is determined to be converted to the non-function verb expression, and the control is passed to the non-function verb expression unit 29. Conversely, if the actual sentence length is shorter than the lower limit, it is determined that the non-functional verb expression should be converted to the functional verb expression, and the control is passed to the functional verb expression unit 30.

(6) The non-functional verb expression unit 29 generates an expression that does not use a functional verb for the expression determined to be a functional verb expression by the functional verb determination unit 25. For example, the expression “increase” is generated for the expression “show increase”. Also, at this time, if the context before and after the context needs to be changed, an expression including the context is generated. For example, for the expression "indicating a large increase", the unified modifier for "increase" of "large" is transformed into a combined modifier for "increase".

(7) The functional verb expression unit 30 generates an expression using a functional verb for the expression determined by the functional verb determination unit 25 to be a non-functional verb expression. For example, for the expression "increase", the expression "show increase" is generated. Also, at this time, if the context before and after the context needs to be changed, an expression including the context is generated.

(8) Then, based on the contents generated by the non-functional verb expression unit 29 and the functional verb expression unit 30, the text storage unit
Change the actual text in 22.

As described above, in the present embodiment, when a condition to be satisfied by a created sentence is specified, when a functional verb expression and a non-functional verb expression are mixed in the sentence, a portion to be rewritten among them is determined. Since it can be detected, there is no need to visually search for a corresponding portion. Further, since the rewritten expression is generated, it is possible to quickly determine how to change the expression. Further, desired sentences can be easily created by changing them to appropriate expressions. Also, based on the relatively easy-to-understand condition of the document length limit, how to change the expression in the text, and apply it,
You can easily specify the conditions that the document should satisfy. As a result, the operator can quickly and surely change the verb expression in the sentence. In this embodiment,
The algorithm for detecting functional verbs was to determine whether the verb in the sentence exists in the functional verb dictionary, and whether the main noun of the noun phrase that forms the objective case dominated by the verb is a sa-variant noun. A determination algorithm may be used. In the present embodiment, the conversion to the functional verb expression or the non-functional verb expression is performed semi-automatically. However, the conversion of the text is not performed automatically, and the position in the text that can be changed and its conversion candidate May be displayed, and the conversion operation may be actively performed by the operator. In the present embodiment, the condition that the sentence specified by the sentence expression specifying unit should satisfy is the length of the sentence. However, the specification such as “use a functional verb expression” or “use a non-functional verb expression” may be used. Is also good.

Effect of the Invention In the proofreading device of the present invention, it is possible to easily determine whether a functional verb expression and a non-functional verb expression are mixed in a sentence text. As a result, the operator can quickly, reliably, and easily find an inconsistent verb expression in a sentence. In addition, when a functional verb expression and a non-functional verb expression are mixed in a sentence, it is possible to detect an expression in the sentence to be rewritten among them, so that it is not necessary to visually search for a corresponding part. .
Further, since the rewritten expression is generated, it is possible to quickly determine how to change the expression. Furthermore, by converting them, a desired sentence can be created. As a result, the operator can quickly and surely change the verb expression in the sentence.
Further, it is determined whether the expression in the sentence should be changed to a functional verb expression or a non-functional verb expression from the conditions to be satisfied by the document, and the condition is easily applied.
These have a great practical effect.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a calibration device according to a first embodiment of the present invention,
FIG. 2 is a diagram showing an example of a document to be processed in the proofing device of the first embodiment of the present invention. FIG. 3 is a diagram showing a display example of the proofing device of the first embodiment of the present invention. FIG. 5 is a configuration diagram of a calibration device according to a second embodiment of the present invention, FIG. 5 is a configuration diagram of a conventional calibration device, and FIG. 6 is a configuration diagram of hardware that implements each embodiment of the present invention. It is a block diagram showing an example. 11: text storage unit, 12: morphological analysis unit, 14: functional verb determination unit, 16: inconsistency determination unit, 17: display unit, 22
… Text storage unit, 23… Morphological analysis unit, 25… Functional verb determination unit, 27… Sentence expression designation unit, 28… Expression conversion instruction unit, 29… Non-functional verb expression unit, 30… Functional verb expression unit.

Claims (4)

(57) [Claims] A text storage unit for storing a sentence character string;
A morphological analyzer that recognizes the morphemes that make up the sentence character string in the text storage unit, a Japanese dictionary that holds at least many Japanese words along with their part of speech information, and a functional verb that holds many verbs that can be functional verbs The dictionary unit and the data of the functional verb dictionary unit are used to check whether each morpheme is a verb that can be a functional verb expression, and a sentence including the verb is subjected to a syntax analysis using the data of the Japanese dictionary unit. And a functional verb determining unit that determines whether the verb is a functional verb, a non-uniformity determining unit that checks whether all verb expressions in a sentence are functional verb expressions, and applicable if there is inconsistency A display unit for displaying an expression part
An apparatus for constructing a Japanese document, wherein when a functional verb expression and a non-functional verb expression are mixed in a document, they are displayed. 2. A text storage unit for storing a sentence character string,
A morphological analyzer that recognizes the morphemes that make up the document character strings in the text storage unit, a Japanese dictionary that holds at least many Japanese words along with their part of speech information, and a functional verb that holds many verbs that can be functional verbs A dictionary unit, using the data of the functional verb dictionary unit, checks whether each morpheme is a verb that can be a functional verb expression, and performs a syntax analysis on a sentence containing the verb using the data of the Japanese dictionary unit. A functional verb determining unit for determining whether the verb is a functional verb, and an expression converting unit for converting an expression relating to the functional verb. A Japanese document proofreading device characterized by converting a verb expression or a functional verb expression. 3. The Japanese document proofreading apparatus according to claim 2, wherein the expression conversion unit converts an expression using a functional verb into an expression not using a functional verb. 4. The Japanese document proofreading apparatus according to claim 2, wherein the expression conversion unit converts an expression not using a functional verb into an expression using a functional verb.