JPH096782A - Document processing method and apparatus - Google Patents

Abstract

(57) [Abstract] [Purpose] To provide a document processing method and apparatus capable of efficiently performing proofreading processing of a sentence. An unknown word detecting step (S1, S2) of detecting an unknown word from a predetermined sentence based on a search of a word dictionary having at least word notation, a character string included in a predetermined document, and the unknown word Correlation calculation step (S) for calculating the correlation value of
4), and a selection step of selecting the character string having the calculated high correlation value as a correction word candidate for the unknown word (S5).
With.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document processing method and its apparatus, and more particularly to a document processing method and apparatus for detecting and pointing out a possible error in a sentence and proofreading the sentence. It is a thing.

[0002]

2. Description of the Related Art Generally, in a document processing device for proofreading a sentence, the sentence is analyzed, and in the analysis result, an unknown word portion obtained from a word that is not in a word dictionary or a portion that cannot be analyzed is erroneous. I had pointed out that there was a possibility of.

[0003]

However, in this type of conventional device, it is only pointed out that the unknown word portion may be erroneous because it uses a word that is not in the word dictionary or an unparseable portion as a clue. Then, there is a drawback that it is not possible to show what word should be corrected. The present invention has been made in view of the above-mentioned conventional example, and an object of the present invention is to provide a document processing method and an apparatus therefor capable of efficiently performing proofreading processing of a sentence.

[0004]

In order to achieve the above object, the document processing method and apparatus of the present invention have the following configurations. That is, an unknown word detecting step of detecting an unknown word based on a search of a word dictionary having at least word notation from a predetermined sentence, and calculating a correlation value between the character string included in a predetermined document and the unknown word. A correlation calculation step and a selection step of selecting the character string having a high calculated correlation value as a correction word candidate for the unknown word are provided.

Another aspect of the present invention is an unknown word detecting means for detecting an unknown word from a predetermined sentence based on a search of a word dictionary having at least word expressions, a character string included in a predetermined document, and A correlation calculation means for calculating a correlation value of an unknown word and a selection means for selecting a character string having a high calculated correlation value as a correction word candidate for the unknown word are provided.

[0006]

With the above structure, an unknown word is detected from a predetermined sentence based on a search of a word dictionary having at least word notation, and a correlation value between the character string included in the predetermined document and the unknown word is calculated. Then, the character string having a high calculated correlation value is selected as a correction word candidate for the unknown word.

In another invention, the unknown word detecting means detects an unknown word from a predetermined sentence based on a search of a word dictionary having at least word notation, and the selecting means is included in the predetermined document. The correlation value between the character string and the unknown word is calculated, and the selecting means selects a character string having a high calculated correlation value as a candidate for a correction word for the unknown word.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First, after summarizing one of the points of the structure of a document processing apparatus according to an embodiment of the present invention, a detailed description thereof will be given. The document processing apparatus of this embodiment is
A document file that holds text consisting of multiple sentences, and a similar part collection unit that collects, for the unknown word part generated as a result of checking the text, parts similar to the character string of the unknown word part from the document file. , The similar part collection result holding unit that stores the information collected by the similar part collection unit, and the correction of the unknown word part held in the document check result holding unit based on the information held in the similar part collection result holding unit By providing a correction word candidate estimation unit that estimates a candidate word and presenting the correction word candidate for an unknown word portion, the burden of proofreading of the sentence on the user side is reduced, and the effect that efficient editing processing can be performed can be achieved. Have.

A document processing method and apparatus according to an embodiment of the present invention will be described below in detail with reference to the drawings. FIG.
It is a block diagram which shows the outline of a structure of the process in the document processing apparatus which concerns on one Example of this invention. In the figure, 1 is a sentence holding unit that holds a sentence, 2 is a sentence that is held in the sentence holding unit 1 is analyzed, and unknown words such as words that are not in the word dictionary or parts that cannot be analyzed are used as clues. A sentence check processing unit that determines and detects a portion of a sentence that has a possibility of an error. Further, 3 is a word dictionary used in the sentence check processing unit 2, and 4 is a grammar rule used in the sentence check processing unit 2.

Reference numeral 5 is a text check result holding unit for holding the result of the text check processing unit 2. Reference numeral 6 is a document file that holds a text composed of a plurality of sentences. Also,
Reference numeral 7 denotes a similar portion collection processing unit that collects, for the unknown word portion held in the sentence check result holding unit 5, a portion of a character string similar to the character string of the unknown word portion in the document file 6. Reference numeral 8 denotes a similar portion collection result holding unit that stores information collected by the similar portion collection processing unit 7. Further, 9 is a correction candidate word estimation processing unit that estimates a correction candidate word for the unknown word portion held in the sentence check result holding unit 5 based on the information held in the similar portion collection result holding unit 7.

Reference numeral 10 denotes a sentence check result display unit for designating a position where there is a possibility of an error based on the content of the document check result held in the document check result holding unit 5 and further presenting the correction candidate word / phrase. Is. Next, in FIG.
3 is a flowchart showing a processing procedure in the processing configuration shown in FIG. 1. Referring to this figure, the processing procedure of the document processing method, which is an embodiment of the present invention, will be described by taking as an example the input sentence shown in 5a of FIG. 5: "I can solve the problem by the method." .

In FIG. 2, first, in step S1, it is checked whether a text is held in the text holding unit 1. if,
If the sentence holding unit 1 is empty, it waits until the sentence holding unit 1 holds the sentence. If the text is held in the text holding unit 1, the process proceeds to step S2. In step S2, sentence check processing of the sentence held in the sentence holding unit 1 is performed. This is performed in the sentence check processing unit 2 by using the word dictionary 3 and the grammar rule 4, and as a result, the unknown word portion (character string), the detection position, the length (the number of characters), the correction candidate phrase (at this time point) , No information is set), and various information about unknown word parts such as the sentence check result holding unit 5
To hold.

FIG. 5b shows an example in which the check result is held in the text check result holding unit 5. In this example,
There is nothing corresponding to the grammatical rule 4 and the word dictionary 3 in the "by", and it is eventually determined as an unknown word portion, and its detection position is the "7" th character from the beginning of the sentence, and its length. Sa (the number of characters) indicates that the information that the value is "3" is stored.

In step S3, it is checked whether or not an unknown word portion is detected in the sentence check result holding unit 5, and if it is detected, the process proceeds to the similar portion collecting process in step S4. If not detected, the process proceeds to step S6. In step S4, the similar portion collection processing unit 7 collects similar character strings from the document file 6 for the unknown word portion held in the sentence check result holding unit 5, and outputs the information (similar character string and its appearance). The number) is stored in the similar portion collection result holding unit 8.

FIG. 5c shows an example of information about the obtained similar character string. The "No." column stores a serial number for identifying each similar character string. For example, No. 15 has a similar character string of "ni" and has a similarity of "0.58" with "done" in 5b of FIG. The information that the number of appearances of "is" is "1" is stored. The correction candidate priority column is
It is processed by the process of step 5.

In step S5, the correction candidate word estimation processing unit 9 refers to the similar portion collection result holding unit 8 to find out the most unknown word portion in the sentence check result held in the sentence check result holding unit 5. The similar candidate is estimated and the correction candidate word / phrase of the corresponding unknown word portion of the sentence check result holding unit 5 is stored. Here, the similarity L is calculated as follows. L = (m / La + m / Lb) / 2 where m: number of matching characters La: length of character string of unknown word portion (character number) Lb: length of character string of similar portion (character number) At this time, similarity If they are the same, priority is given to the one having the largest number of appearances.

The calculated priority is stored in the correction candidate priority column of 5c in FIG. Next, as shown in 5d of FIG. 5, the one having a high correction candidate priority shown in 5c of FIG. 5 is selected, and it is displayed in the correction candidate word column of the text check result holding unit 5 in an unknown word. It is stored as a correction candidate word corresponding to the part. In step S6, the text check result display unit 10 displays the text check result stored in the text check result storage unit 5, and the process ends.

In the example of 5e in FIG. 5, the original input sentence, the detected phrase "by" detected as an unknown word therein, and the corresponding correction candidate word "by" are displayed at the same time. are doing. Also, in the original input sentence, an unknown word is underlined to indicate its position. Next, FIG. 3 shows the contents of the document file 6 which is an example used in the above processing.

Further, FIG. 4 shows a similar portion collection result holding unit 8
It shows an example of the contents of the above, and stores information on similar words and the number of appearances. You. FIG. 6 shows an example of a hardware configuration for executing the document processing shown in the above-mentioned flowchart.

Referring to FIG. 6, CPU 200 controls the entire document processing apparatus, and sequentially reads and executes programs corresponding to the above-mentioned flowcharts stored in memory 202 in advance. Further, the memory 202 stores the contents of the word dictionary 3, the grammar rule 4, the document file 6, the similar portion collection result holding unit 8, the sentence check result holding unit 5, and the like described above. The keyboard 203 and pointing device 204 input commands, text data, and the like. The display monitor 201 is an example of the above-mentioned sentence check result display unit, and displays the above-mentioned processing result and the like. (Other Embodiments) (01) In the above embodiment, the similar portion collection processing unit considers it as a temporary one, performs similar portion collection processing for each unknown word portion in the input sentence, and performs correction candidate word estimation processing. Although an example has been described, the similar part collection processing unit may be of a constant type, that is, the similar candidate information may be collected in advance from the document file and the correction candidate word estimation process may be performed.

(02) In the above embodiment, as the unknown word part,
Pay attention to what is composed of consecutive hiragana character strings,
In addition, in the similar portion collection processing and the correction candidate word estimation processing, the method of collecting similar words and estimating the correction candidate words from the text in the document file has been explained, paying attention to the continuous hiragana character strings. , But not limited to this, pay attention to an arbitrary character string as the unknown word portion, and in the similar portion collection processing and the correction candidate word estimation processing, the length of the character string of the unknown word portion Alternatively, the allowable range may be set, and similar portions may be collected within the allowable range to estimate the correction candidate word / phrase.

(03) In the above embodiment, the character string of the unknown word portion in the sentence check result held in the sentence check result holding means and the similar portion collection result holding unit at the time of the correction candidate word estimation process The method of estimating the correction candidate phrase for the corresponding unknown word part by calculating the similarity based on the number of matching characters and the number of appearances of the character string of the similar phrase registered in. As for the method of calculating the degree of similarity, any method can be used as long as the degree of similarity can be calculated and prioritization can be performed.

(04) In the above-described embodiment, the correction candidate word estimation process is described in which the correction candidate word estimation process is performed and one most similar correction candidate phrase is presented, but the present invention is not limited to this. The correction candidate word / phrase may be in a format in which a plurality of correction candidate word / phrase estimation processes are presented in order of increasing similarity. (05) In the above embodiment, the similar portion collection processing and the correction candidate word estimation processing are automatically performed, but the user can interactively select each of the similar portion collection processing and the correction candidate word estimation processing. The format may be such that each process can be skipped.

(06) In the above-mentioned embodiment, in the text check result display, the format of only displaying the part of the text check result by using the underline has been described, but the present invention is not limited to this and, for example, a color Any format that can be recognized by the user, such as displaying, changing the font of characters, or using voice, may be used.

(07) In the above-described embodiment, in the document check result display, the part of the text check result is displayed by using underlining, and the correction candidate word / phrase for each detection is presented in another display area. The present invention is not limited to this, and for example, a format in which information such as a correction candidate regarding a detected word is presented by clicking the location of the text check result with a mouse or the like may be used.

(08) In the above embodiment, Japanese has been described as an example, but the present invention is not limited to this and can be applied to any language such as English or German.
The present invention may be applied to a system including a plurality of devices or an apparatus including a single device.
Needless to say, the present invention can be applied to a case where the present invention is achieved by supplying a program to a system or an apparatus.

As described above, according to this embodiment,
Regarding the unknown word parts obtained by the sentence check process, similar parts can be collected using document check, and the most similar one can be estimated as the correction candidate word and, as a result, it is possible that the error is unknown. Not only the word portion is pointed out, but also correction candidate words and phrases relating to the unknown word portion can be presented, and the effect that the sentence proofreading process can be efficiently performed is obtained.

[0028]

As described above, according to the present invention, the sentence proofreading process can be efficiently performed.

[Brief description of drawings]

FIG. 1 is a functional block diagram according to an embodiment of the present invention.

FIG. 2 is a flowchart showing a processing procedure according to the embodiment of the present invention.

FIG. 3 is an example of contents of a document file for explaining an embodiment of the present invention.

FIG. 4 is an example of contents of a similar portion collection result holding unit for explaining an embodiment of the present invention.

FIG. 5 is an example showing a process of specific processing for explaining the embodiment of the present invention.

FIG. 6 is a diagram showing a configuration of a document processing apparatus according to an exemplary embodiment of the present invention.

[Explanation of symbols]

 1 sentence holding unit 2 sentence check processing unit 3 word dictionary 4 grammar rule 5 sentence check result holding unit 6 document file 7 similar portion collection processing unit 8 similar portion collection result holding unit 9 correction candidate word estimation processing unit 10 sentence check result display unit

Claims (14)

[Claims] 1. An unknown word detecting step of detecting an unknown word from a predetermined sentence based on a search of a word dictionary having at least word notations, and a correlation value between a character string included in a predetermined document and the unknown word. And a selecting step of selecting a character string having a high calculated correlation value as a correction word candidate for the unknown word. 2. The document processing method according to claim 1, wherein the correlation value is calculated based on the number of matching characters between a character string included in a predetermined document and the unknown word. 3. The correlation value is calculated based on the number of matching characters between a character string included in a predetermined document and the unknown word, and further, based on the number of characters of the unknown word and the number of characters of the character string. The document processing method according to claim 2. 4. The correlation value (L) is the number of matching characters (m) between a character string included in a predetermined document and the unknown word, and further, the number of characters of the unknown word (La) and the number of characters of the character string. The document processing method according to claim 3, wherein the document processing method is obtained by the following formula: L = (m / La + m / Lb) / 2 based on (Lb). 5. The selecting step further includes: when a plurality of character strings having a high calculated correlation value have the same correlation value, the one having a higher appearance frequency of the character string in the predetermined document is the unknown. 2. The document processing method according to claim 1, wherein the word is selected as a correction word candidate for the word. 6. The document processing method according to claim 1, further comprising a first display step of displaying the unknown word and a correction word candidate selected corresponding to the unknown word. 7. The document processing method according to claim 1, further comprising a second display step of displaying the predetermined sentence and a position of the unknown word of the predetermined sentence with a mark. . 8. An unknown word detection means for detecting an unknown word from a predetermined sentence based on a search of a word dictionary having at least word notation, and a correlation value between a character string included in a predetermined document and the unknown word. A document processing apparatus, comprising: a correlation calculation unit that calculates a correlation value; and a selection unit that selects a character string having a high correlation value calculated as a correction word candidate for the unknown word. 9. The document processing apparatus according to claim 8, wherein the correlation value is calculated based on the number of matching characters between a character string included in a predetermined document and the unknown word. 10. The correlation value is calculated based on the number of matching characters between a character string included in a predetermined document and the unknown word, and further, based on the number of characters of the unknown word and the number of characters of the character string. The document processing device according to claim 9. 11. The correlation value (L) is the number of matching characters (m) between a character string included in a predetermined document and the unknown word, and further, the number of characters of the unknown word (La) and the number of characters of the character string. The document processing apparatus according to claim 10, wherein the document processing apparatus is obtained by the following formula: L = (m / La + m / Lb) / 2 based on (Lb). 12. The selecting unit further determines, when a plurality of character strings having a high calculated correlation value have the same correlation value, the unknown one having a higher appearance frequency of the character string in the predetermined document. 9. The document processing device according to claim 8, wherein the word processing device is selected as a correction word candidate for a word. 13. The document processing apparatus according to claim 8, further comprising a first display unit that displays the unknown word and a correction word candidate selected corresponding to the unknown word. 14. The document processing apparatus according to claim 8, further comprising second display means for displaying the predetermined sentence and a position of the unknown word of the predetermined sentence with a mark. .