JPH07129577A - Document proofreading supporting device - Google Patents

Abstract

PURPOSE:To provide a document proofreading supporting device which investigates the possibility of an error to a compound word included in a sentence as a processing object and prompts a user to watch the possibility of error when it is high. CONSTITUTION:The feature of the device is providing a sentence storing means 1 storing an inputted sentence, a compound word storing means 3 storing compound word information detected from the sentence storing means 1 and a compound word detection means 2 detecting the compound word from the sentence stored in the sentence storing means 1 and registering the notation of the compound word and its position in the sentence storing means 1 into the compound word storing means 3 by setting the reading of the compound word to be a key. It is also the feature of the device to furthermore provide a compound word dictionary 5 registering the reading of the compound word and its notation.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document proofreading support apparatus, which investigates the possibility of an error in a compound word contained in a sentence to be processed. The present invention relates to a document proofreading support device capable of prompting caution.

[0002]

2. Description of the Related Art Recently, it has become common to create sentences using a sentence creating device such as a word processor. By the way, in this sentence creation device, when Japanese is created by kana-kanji conversion, for example, erroneous characters often occur due to the conversion error. That is, in Japanese, there are so-called homonyms having the same reading but different notations, and a word different from the desired word may be mistakenly selected at the time of conversion. Therefore, there is a demand for the realization of an error detection device that checks the validity of use of a word in a sentence created by such a sentence creation device and detects a typographical error, and efforts are being made to develop it. Conventionally, such an error detection device uses a method of using morphological analysis or syntactic analysis in a sentence as a creation target and detecting an unregistered word or a non-grammatical word based on the analysis result. It was However, according to this method, it is difficult to detect a morphological error in a compound word formed by combining a plurality of words. Therefore, it is possible to effectively detect the typographical error of the word included in the compound word. Could not be done. In order to improve such a problem, JP-A-2-3
A compound word error detecting device has been developed as exemplified in 09448 [Error detecting device]. In the error detection device illustrated here, when a part of the compound word contains a homonym, the compound word created by replacing this part with a homonym is searched for in the dictionary, When it exists, the above compound word is set as an error. However, this method has a problem that a compound word that is not registered in the search dictionary cannot be detected, and the number of searches of the compound word dictionary increases, which causes a difficulty in operation efficiency. Further, as a device for detecting a word that is not unified in notation appearing in a document, there is disclosed in Japanese Patent Laid-Open No. 5-143.
The apparatus disclosed in Japanese Patent Laid-Open No. 587 [Document processing method and apparatus] can be cited. According to this, a word dictionary that stores information in which word expressions and grammatical information are stored in association with each other and has information that connects words having the same readings is used to point out words that have the same readings and different expressions in a document. To be done.

[0003]

As described above, in recent years it has become common to create sentences using a sentence creation device such as a word processor. For example, Japanese is created by Kana-Kanji conversion. In such cases, typographical errors often occur due to errors in the conversion operation.Therefore, it is necessary to check the validity of use of the words in a sentence created by such a sentence creating device, and There is a demand for the realization of an error detection device for detecting the error, and the corresponding development has been made. By the way, the conventional error detection device of this type is
Although the method of detecting unregistered words and non-grammatical words from the results by using morphological analysis and syntactic analysis of the sentence as the processing target, when such a method is used, multiple words It is difficult to morphologically detect an error in a compound word formed by combining two words, and it is impossible to effectively detect a typographical error in a word included in the compound word. Therefore, in order to improve such a problem, a compound word error detecting device exemplified in JP-A-2-309448 [error detecting device] was developed. When this method is used, as described above. However, there is a problem that a compound word that is not registered in the search dictionary cannot be detected, and the number of times the compound word dictionary is searched increases, resulting in difficulty in operation efficiency. Further, another publication, Japanese Patent Laid-Open No. 143587/1993 [Document processing method and device], discloses a device for detecting a word that is not unified in notation appearing in a document. According to this, although the inconsistent part of the notation of the word registered in the dictionary can be detected, for example, "Yesterday" and "Function" of the sentence "Yesterday, about the functional specification" are also detected, and a false indication is made. There is a problem that it increases. Here, if the compound word is to be detected while being conscious of it, the compound word is registered in the dictionary, but the problem arises that the size of the dictionary to be used becomes huge. Further, there is a problem that extra information for connecting words having the same reading is required.

The present invention has been made in order to eliminate or improve the above-mentioned drawbacks of the prior art, and uses the unification of the notations of compound words which are detected in the compound word and not registered in the compound word dictionary as a clue. It is an object of the present invention to provide a document proofreading support device which points out the possibility of an error in a synonym and reduces the number of searches of a compound word dictionary to further improve processing efficiency.

[0005]

The present invention has been made in order to achieve the above object, and a document proofreading support apparatus according to the present invention is a sentence storage means 1 for storing an inputted sentence. A compound word storage means for storing compound word information detected from the memory means, a compound word is detected from a sentence stored in the sentence storage means, and the compound word is written and the sentence storage means is used with the reading of the compound word as a key. And a storage start position of the same are registered in the compound word storage means 2, and a pointing means 4 for pointing out a compound word having a plurality of notations for one reading in the compound word storage means 2. , Is provided. Further, another document proofreading support apparatus according to the present invention further comprises a compound word dictionary 5 in which readings and notations of compound words are registered, and the pointing means is the compound word storage means and the compound word. It is characterized by collating with a dictionary and judging / pointing out the presence or absence of different compound words for the same reading.

[0006]

In the document proofreading support apparatus of the present invention configured as described above, the text storage means 1 stores the Japanese text to be processed. In the subsequent compound word detection means 2, the built-in word analysis means 21 and analysis dictionary 22 are provided.
Is used to decompose the sentence stored in the sentence storage means 1 into corresponding word strings, and the result is analyzed result storage means 23.
Memorized in. Then, in the decomposed word string, a predetermined one such as a part in which nouns are continuous is detected as a compound word, and a notation of the detected compound word; reading; a storage start position in the text storage means 1 and the like are required. Information is registered in the compound word storage means 3 using the reading as a key. And
The pointing means 4 has an action of pointing out a compound word having a plurality of compound word notations for one reading in the compound word storage means 3. Further, another document proofreading support apparatus of the present invention further includes a compound word dictionary 5 for storing at least readings of compound words and their notations. Then, according to the pointing means 4 included in this another document proofreading support device, it is judged whether or not there are a plurality of compound word notations for one reading in the compound word storage means 3. ,
Further, the compound word dictionary 5 is searched based on the reading of the compound word registered in the compound word storage means 3, and if there is a compound word registered according to the corresponding reading, the search result is searched. This has the advantage that the indications about the use of compound word errors are more effective. That is, according to the device of the present invention, a compound word included in the sentence to be processed is detected, and the unification of the notation of the compound word not registered in the compound word dictionary is used as a clue to detect the error of the homonym. The possibility is pointed out, and the advantage is that the efficiency of processing by the device is significantly improved by reducing the number of searches of the compound word dictionary.

[0007]

Embodiments of the document proofreading support apparatus according to the present invention will be described below with reference to FIGS.

[First Embodiment] FIG. 1 is a block diagram schematically showing the configuration of a first embodiment of a document proofreading support apparatus according to the present invention. In FIG. 1, 1 is a text storage means,
2 is a compound word detection means, 21 is a word analysis means, 22 is an analysis dictionary, 23 is an analysis result storage means, 3 is a compound word storage means,
And 4 is a pointing means. First, the sentence storage means 1
Stores a Japanese sentence to be processed. The compound word detecting means 2 decomposes the sentence stored in the sentence storing means 1 into a sequence of corresponding words using the word analyzing means 21 and the analysis dictionary 22, and stores the result in the analysis result storing means 23. Then, regarding the Japanese sentence as the processing target decomposed into a sequence of corresponding words, a part in which nouns are continuous or a word in which the compound word flag is ON is detected as a compound word, and the detected compound word notation; reading; The storage start position or the like in the text storage means 1 is registered in the compound word storage means 3 using the reading as a key. The compound word storage means 3 stores the compound word detected by the compound word detection means 2 together with the notation of the compound word and the storage start position in the sentence storage means 1, using the reading as a key. The pointing means 4 points out a compound word in the compound word storage means 3 in which a plurality of compound expressions are present for one reading.

2 and 3 are flow charts for explaining the flow of processing in the first embodiment. Less than,
The flow of processing in the first embodiment will be described mainly with reference to FIGS. 2 and 3.

Here, FIG. 4 exemplifies the text contents stored in the text storage means 1. In FIG. 4, only the parts necessary for explaining the first embodiment of the present invention are shown, and the corresponding parts of the sentences are the sentence numbers 101, 102, and 10.
3, 104, 105. Here, the sentence with the sentence number 101 is [... function use ...
・. ], The intermediate position between [...] and [] is position 1010, the intermediate position between [] and [function] is position 1011, and [function] and [use] An intermediate position between and is a position 1013, and an intermediate position between [use] and [ga] is a position 1015. The same applies to sentences with other sentence numbers. The operation of the first embodiment shown in FIG. 1 will be described below with reference to FIGS. 2 and 3 by using the text shown in FIG. 4 as a processing target.

First, referring to FIG. 2, the compound word detecting means 2
Initializes the contents of the analysis result storage means 23 and the compound word storage means 3 as shown in FIGS. 6 to 10 and FIGS. 11 to 15 respectively (S0). Next, the first sentence (sentence number 101) is retrieved from the sentence storage means 1 and set in the variable sent (S1). in this case,
The variable sent is set to the sentence number 101, “... use of function ...”. In addition, such a cut-out of a sentence is performed by using a punctuation mark “.” Or a period “.” As a clue, and using the one that appears most frequently in the sentence. Next, it is checked whether the content of the variable sent is empty, and since it is not "empty", the process proceeds to S3 (S2). Then, the content of the current variable sent is decomposed into word strings using the word analysis means 21 and the analysis dictionary 22, and the result is stored in the analysis result storage means 23 (S3).

Here, referring to FIG. 5, the analysis dictionary 2
The contents of 2 will be described as an example. The analysis dictionary 22 here includes at least a headline (notation); reading; part of speech; a compound word flag. In this FIG.
The part-of-speech [sahen noun] indicates that it is both a noun and also a verb followed by the inflectional ending “suru”. The word analysis unit 21 searches for a sentence to be processed from the beginning of the sentence to the end of the sentence on a word-by-word basis, and checks the connection relation between words and the clause end condition. Then, the sentence analysis is performed by the method of the minimum number of clauses, which outputs the one with the smallest number of clauses as a result among the combinations in which the connection relation between words is correct. As a method for selecting the most appropriate one from the combinations of mutually connectable word strings, any desired one such as the longest matching method or the minimum phrase number method can be selected and used.

FIG. 6 shows the contents of the analysis result storage means 23 in which the analysis result of the variable sent is stored. The analysis result storage means 23 includes a headline (notation); reading; part of speech; compound word flag; start position; word length (notation length); Now, it is assumed that the sentence of sentence number 101 is stored, and the start position of the word [no] of word length 1 is 101.
It is represented by the symbol 0; the start position of the word “function” having a word length of 2 is represented by the symbol 1011. In addition, the start position of the word "use" having a word length of 2 following this is 1013.
The starting position of the word [ga] having a word length of 1 is represented by the symbol 1015.

The compound word detecting means 2 checks the contents of the built-in analysis result storing means 23 to see if there is a part where nouns are continuous or a word whose compound word flag is ON (S4). As shown in FIG. 6, since the word "function" and the word "use" are consecutive nouns, the process proceeds to the next step S5 to detect "function use" as a compound word, Using the detected reading of the compound word as a key, the notation of the compound word and the storage start position in the sentence storage means 1 are registered in the compound word storage means 3 (S5).

Referring now to FIG. 11, "function usage".
The contents of the compound word storage means 3 for the compound word are shown. The compound word storage means 3 includes a heading section 31 and a writing section 32. Then, in the heading section 31, the address 3
11; Yomi 312; Number of notations (of detected compound word) 31
3; Includes address 314 to the (corresponding) notation. Further, the notation unit 32 has an address 321; a notation 322;
Address (address) 32 to the next notation (having the same reading)
3; (start of storage in the text storage means 1) position 324; address for reading 325; (dictionary) type 326 are included. Looking at the address 311 of the heading section 31,
This is represented by the symbol a1i (i is a natural number).
Further, regarding the address 321 of the notation part 32, this is represented by a symbol a2i (i is a natural number). In the (dictionary) type 326, if the detected notation of the compound word is registered in the word analysis dictionary 22 or the compound word dictionary 5 described later, the type is stored separately. That is, if the notation of the detected compound word is registered in the word analysis dictionary 22,
Store dic1 in this (dictionary) type 326, and
If it is registered in the compound word dictionary 5, dic2 is stored. If it is not registered in any dictionary, nothing is stored.

In the case of the compound word "function use" here,
This is the first detection and the heading 3
At first, nothing is registered in the beginning. Therefore, the required information regarding "Kinosuke" is stored in an appropriate place of the first address a11 in the address 311 of the heading section 31. Specifically, the reading 312 stores the kana kana kana kana kana kana, the number of notations 313 is set to 1 (since only one is currently detected), and the address 314 to the notation is stored. Stores the first address a21 of the address 321 in the notation unit 32. Also, "let's do yesterday" here
Nothing is initially registered in the notation part 32 for. Therefore, the first of the address 321 in the notation unit 32
Information regarding "function use" is stored in an appropriate place of the address a21. Specifically, regarding the corresponding part of the first address a21 of the address 321 in the notation unit 32, "function use" is stored in the notation 322, and the address (address) 323 to the next notation is stored. Stores own address a21 (since only one is currently detected), position 32
In 4, the storage start position 1011 in the text storage means 1 is stored. Then, in the address 325 for reading, the heading portion 31
The first address a11 in the address 311 is stored, and nothing is stored in the (dictionary) type 326.

Next, the compound word detecting means 2 advances the processing from S5 to S6, extracts one sentence following the contents of the variable sent from the text storage means 1, sets it in the variable sent, and returns the processing to S2. By repeating the processing from S2 to S6 several times, as shown in FIG.
It is assumed that the sentence with the sentence number 102 of “... Required use ...” is set in the variable sent. Variable se here
Since the content of nt is not "empty", the process proceeds from S2 to S3.

In response to the sentence "... request use ..." With the sentence number 102, the compound word detecting means 2
The content of the variable sent is decomposed into a word string using the word analysis means 21 and the analysis dictionary 22, and the result is stored in the analysis result storage means 23 (S3). An example of the contents of this analysis result storage means 23 is shown in FIG. Next, the contents of the analysis result storage means 23 are checked to see if there is a part where nouns continue or there is a word with the compound word flag being ON (S4). As shown in FIG. 7, since the word "request" + word "use" part is a continuous noun part, the process proceeds to S5 to detect "request use" as a compound word, Using the detected reading of the compound word as a key, the notation of the compound word and the storage start position in the sentence storage means 1 are registered in the compound word storage means 3 (S5).

FIG. 12 shows the contents of the compound word storage means 3 to which the information on "request use" is added. Note that, in FIG. 11 described above, nothing is registered for “youkyu”. Here again Figure 1
With reference to FIG. 2, information about “youkyu” is stored in the second address a12 of the address 311 in the index section 31. And the address 32 in the notation 32
Information regarding "requested use" is stored in the second address a22 of 1.

Next, the compound word detecting means 2 executes the processing from S5 to S.
6, the sentence following the content of the variable sent is extracted from the sentence storage means 1 and set in the variable sent, and the process returns to S2. Now, it is assumed that the process “S2 to S6” is repeated several times and the sentence “... Induction use ...” Of the sentence number 103 in FIG. 4 is set in the variable sent. Since the content of the variable sent here is not "empty", the process proceeds from S2 to S3.

The sentence of the sentence number 103 "... inductive use ...
・ ・. In contrast, in the compound word detection means 2, the content of the variable sent is changed to the word analysis means 21 and the analysis dictionary 22.
Is used to decompose into a word string, and the result is stored in the analysis result storage means 23 (S3). An example of the contents of the analysis result storage means 23 here is shown in FIG. Then, the contents of the analysis result storage means 23 are checked to see if there is a part where nouns are continuous or a word whose compound word flag is ON (S4). As shown in FIG. 8, since the part of the word "induction" + the word "use" is a continuous part of the noun, the process proceeds to S5 to detect "induction use" as a compound word, and With the reading of the detected compound word as the key,
The notation of the compound word and the storage start position in the text storage means 1 are registered in the compound word storage means 3 (S5).

FIG. 13 shows the contents of the compound word storage means 3 to which the information regarding "induction use" is added. Note that, in FIG. 12 described above, the reading and the notation of "let's do yesterday" are registered (apparently already). Therefore, when tracing the notation connected by the address (address) 323 to the next notation (having the same reading) of the notation part 32 by the number (= 1) of the notation number 313 in the heading part 31 in FIG. , Arrives at the notation part indicated by the first address a21 in the address 321 of the notation part 32. The notation information regarding "function use" is stored in the proper place corresponding to the first address a21. Therefore, as shown in FIG. 13, the number of notations 313 in the index section 31 is 313.
Is increased by 1 (= 2), and the corresponding notation information “induction use” is added to the third address a2 in the address 321 of the notation unit 32.
3 in accordance with this, address 3 of the notation unit 32 in accordance with this
The address (address) 323 to the next notation (having the same reading) corresponding to the first address a21 in FIG.
A) in a) to a23. Then, the address (address) 3 to the next notation (having the same reading) corresponding to the third address a23 in the address 321 of the notation unit 32
The above a21 is stored in 23.

Next, the compound word detection means 2 advances the processing from S5 to S6, extracts one sentence following the content of the variable sent from the text storage means 1, sets it in the variable sent, and returns the processing to S2. Now, the processes of S2 to S6 are repeated several times, and the sentence "... required specifications ..." of the sentence number 104 in FIG.
Is set to the variable sent. Here, since the content of the variable sent is not "empty", S2 to S3
Proceed to.

The sentence of this sentence number 104 ...
・ ・. In contrast, in the compound word detection means 2, the content of the variable sent is changed to the word analysis means 21 and the analysis dictionary 22.
Is used to decompose into a word string, and the result is stored in the analysis result storage means 23 (S3). This analysis result storage means 23
An example of the contents of is shown in FIG. Analysis result storage means 23
Is checked to see if there is a part where nouns are continuous or there is a word whose compound word flag is ON (S
4). As shown in FIG. 9, since the compound word flag of "requirement specification" is ON, the process proceeds to S5, "requirement specification" is detected as a compound word, and reading of the detected compound word is performed. With the key as a key, the notation of the compound word and the storage start position in the sentence storage means 1 are registered in the compound word storage means 3 (S5).

FIG. 14 shows the contents of the compound word storage means 3 to which the information on the compound word "requirement specification" is added. Note that in FIG. 13 described above, the reading and the notation “youkyu” are (apparently already) registered. Therefore, the address (address) 32 to the next notation (having the same reading) of the notation part 32 by the number (= 1) of the notation number 313 in the heading part 31 in FIG. 13 described above.
By tracing the notations connected by 3, the user arrives at the notation part indicated by the second address a22 in the address 321 of the notation part 32. In the proper place corresponding to this second address a22,
Stores notational information related to "required use". Therefore, as shown in FIG. 14, the number of notations 31 corresponding to the second address a12 in the address 311 of the header section 31 is 31.
The number of 3 is increased by 1 (= 2), and the corresponding notation information “requirement specification” is added to the fourth address a in the address 321 of the notation unit 32.
It stores in 24. Then, in accordance with this, the address (address) 32 to the next notation (having the same reading) corresponding to the second address a22 in the address 321 of the notation unit 32.
Change 3 to a24. The compound word "requirement specification" here
Was already registered in the analysis dictionary 22, so (of the dictionary)
Storing dic1 in the type 326 enables the determination.

Next, in the compound word detecting means 2, the processing proceeds from S5 to S6, one sentence following the content of the variable sent is extracted from the text storage means 1 and set in the variable sent, and the processing is executed in S2. return. Now, by repeating the processing of S2 to S6 several times, the sentence "...
I built a dictionary. Is set to the variable sent. Since the content of the variable sent here is not "empty", the process proceeds from S2 to S3.

In response to the sentence of the sentence number 105, "... Dictionaries were constructed.", The compound word detecting means 2 uses the variable s.
The contents of ent are decomposed into word strings using the analysis means 21 and the analysis dictionary 22, and the result is stored in the analysis result storage means 23 (S3). An example of contents of the analysis result storage means 23 is shown in FIG.
It shows in 0. Check the contents of the analysis result storage means 23,
It is checked whether there is a part where nouns are continuous or a word whose compound word flag is ON (S4). As shown in FIG. 10, since the parts of "dictionary" and "construction policy" are consecutive nouns, the process proceeds to S5, "dictionary construction policy" is detected as a compound word, and the detected compound is detected. Using the word reading as a key, the notation of the compound word and the position in the text storage means 1 are registered in the compound word storage means 3 (S5).

FIG. 15 shows the contents of the compound word storage means 3 to which the information about "dictionary construction" is added. In FIG. 14 above, no heading section 31 for the reading of “dictionary construction policy” is registered. Therefore, the information regarding "jishokensaku" is stored in the third address a13 of the address 311 of the heading section 31. Also, the notation part 3
For 2, the information regarding the “dictionary construction policy” is stored in the fifth address a25 of the address 321.

Next, the compound word detecting means 2 advances the processing from S5 to S6, extracts one sentence following the content of the variable sent from the text storing means 1, sets it in the variable sent, and returns the processing to S2. Here, it is assumed that the sentence cannot be extracted from the sentence storage means 1 and “empty” is set in the variable sent. Since the content of sent is "empty", the processing is S
Go to S7 instead of 3.

The pointing means 4 checks the contents of the compound word storage means 3 and points out compound words having the same reading but different notations. FIG. 3 is a flow chart showing an operation related to the pointing means 4 (which is a detailed description of S7 in FIG. 2). The operation relating to the pointing means 4 will be described below with reference to FIGS. 1, 3, and 15.

The pointing means 4 first sets the head address a11 of the heading portion 31 of the compound word storage means 3 in the variable adr1 (S201). Next, whether or not the variable adr1 is 0 is checked (S202), but since it is not 0, S203
Proceed to. In S203, the variable adr1 (= a
The number of notations 313 in the heading section 31 specified in 11)
Check if is greater than 1. As shown in FIG. 15, since the number of notations 313 in the heading section 31 specified by the variable adr1 (= a11) is 2,
The process proceeds to S204.

In S204, the contents of "address 314 for notation" in the header section 31 designated by the variable adr1 (= a11) are set in the variables adr20 and adr2, and the process proceeds to S205. As a result of the above setting, a21 is set to the variable addr20 and the variable adr2.

In S205, adr2 (= a2
The notation 322 is pointed out via the position 324 of the notation 32 designated in 1). An example of displaying this indication is shown in FIG.
The sentence with the sentence number 101 of "... the function is used ..." is shown. Here, if nothing is stored in the dictionary type,
It only shades the detected compound words. if,
If a value is stored in the dictionary type, the detected compound word is further underlined. In this case, since nothing is stored in the dictionary type, the compound word “function use” is only shaded.

Next, the process proceeds to S206 and the variable adr
2 is set to "address 323 to next notation" (= a23) of the notation unit 32 designated by adr2, and the process proceeds to S207. In S207, adr20 (= a2
It is checked whether or not the contents of 1) and adr2 (= a23) are equal. Since they are not equal, the process returns to S205.

In step S205, the notation 322 is pointed out via the position 324 of the notation portion designated by adr2 (= a23). An example of displaying this indication is shown at 103 in FIG. In this case, since nothing is stored in the dictionary type, the compound word “inductive use” is simply shaded.

Next, the process proceeds to S206 and the variable adr
2 is set to "address 323 to next notation" (= a21) of the notation unit 32 designated by adr2 (= a23), and the process proceeds to S207. In this S207,
It is checked whether the contents of adr20 (= a21) and adr2 (= a21) are equal.
Proceed to 08. In this S208, the variable adr1
The address (= a12) next to the content (a11) is set in, and the process returns to S202.

In S202, it is checked whether or not the variable adr1 is 0. However, since it is not 0, the process proceeds to S203. The variable adr1 (= a12) in S203
It is checked whether or not the number of notations 313 of the heading section 31 designated by is larger than one. As shown in FIG. 15, since the number of notations 313 in the heading section 31 specified by the variable adr1 (= a12) is 2, the process proceeds to S204. In S204, the variable adr20 and the variable adr2
In response, the content (a22) of the "address 314 to the notation" of the heading section 31 designated by adr1 (= a12) is set, and the process proceeds to S205. As a result, a22 is set in the variables addr20 and adr2.

In S205, adr2 (= a2
The notation 322 is pointed out via the position 324 of the notation part 32 designated in 2). An example of displaying this indication is shown in FIG.
102. In this case, since nothing is stored in the dictionary type 326, the compound word "requested use" is only shaded.

Next, the process proceeds to S206 and the variable adr
For 2, set “address 323 to next notation” (= a24) of the notation unit 32 specified by adr2, and set S
The process proceeds to 207. In this S207, adr
20 (= a22) and adr2 (= a24) are checked to see if they are equal.
Return to 5.

At S205, adr2 (= a24)
The notation 32 via the position 324 of the notation part 32 designated by
Point out 2. An example of displaying this indication is 1 in FIG.
Shown in 04. In this case, the dictionary type 326 is dic
Since 1 is stored, the compound word “requirement specification” is shaded and underlined.

Next, the process proceeds to S206 and the variable adr
For 2, the “address 323 to the next notation” (= a22) of the notation portion specified by adr2 (= a24) is set, and the process proceeds to S207. In S207, a
It is checked whether the contents of dr20 (= a22) and adr2 (= a22) are equal.
Proceed to 8. In S208, the last address (= a13) of the content (a11) is set in the variable adr1,
The process returns to S202.

In S202, it is checked whether or not the variable adr1 is 0. However, since it is not 0, the process proceeds to S203. In S203, it is checked whether or not the number of notations 313 of the heading section 31 specified by the variable adr1 (= a13) is greater than 1. As shown in FIG. 15, the variable adr
Since the number of notations 313 of the heading section specified by 1 (= a13) is 1, the process proceeds to S208. In S208, the variable adr1 is set to the next address of the contents (a11) (0 in this case because there is no subsequent one), and the process returns to S202. In this S202, it is checked whether or not the variable adr1 is 0, but since it is 0, the processing is ended.

For a compound word having a plurality of notations for one reading, compound words of other homophones can be easily grasped by checking the compound word storage means 3. For example, the other notation of “function use” in the sentence “...
It is included in the address a21 of the notation part 32 from the position 1011. This address a21 is temporarily stored as the starting address, and the address to the next notation is
Follow until you reach the start address you remembered temporarily. Every time this tracing operation is performed, the corresponding notation 3 in the notation unit 32
If 22 is extracted, a plurality of different notations can be listed for one reading. Statement number 1 in FIG. 16
In the case of "function use" for the sentence 01, it is recognized that "induction use" is another notation in the document. Regarding the other notation of “inductive use”, the initial “functional use” can be obtained by tracing the address to the next different notation. By using such information as a clue, it is possible for the user of the apparatus to judge whether or not a certain predetermined compound word is used.

[Second Embodiment] FIG. 17 is a block diagram showing a second embodiment of the document proofreading support apparatus according to the present invention. In FIG. 17, 1 is a text storage means, 2 is a compound word detection means, 21 is an analysis means, 22 is an analysis dictionary, 23 is an analysis result storage means, 3 is a compound word storage means, 4 is a pointing means, and 5 is a compound. It is a word dictionary. Here, the sentence storage means 1,
The compound word detecting means 2, the compound word storing means 3, and the indicating means 4 are
This is the same as that of the first embodiment described above.

The pointing means 4 judges whether or not there are a plurality of compound word expressions for one reading in the compound word storage means 3. Further, the compound word dictionary 5 is searched by reading the compound word registered in the compound word storage means 3, and if there is a compound word registered by this reading, the search result is reflected in the compound word storage means 3. The compound word dictionary 5 here stores at least the reading and notation of compound words.

FIG. 18 is a flow chart for explaining the flow of processing in the present invention. The flow of processing will be described below with reference to FIG. Here, in FIG. 4, as described in the first embodiment, the sentence storage means 1 is used.
The text content stored in is shown. In FIG. 4, only the portions necessary for explaining the present invention are shown, and these exemplary text portions are sentence numbers 101 and 10.
2, 103, 104, 105. The operation of the second embodiment of FIG. 17 will be described below with reference to the text shown in FIG. 4 and also with reference to FIGS. 18 and 3.

In FIG. 18, S0 of the compound word detecting means 2
The operations from S6 to S6 are the same as in the case of the first embodiment, so the description thereof is omitted. In the following, sentence number 10 in FIG.
The operation from the time when the sentence 5 is processed, "empty" is set in the variable sent, and the process proceeds to S7A instead of S3 in S2 will be described as an example.

In this S7A, the pointing means 4 is shown in FIG.
For each of the readings 312 in the index section 31 of the compound word storage means 3 shown in FIG. 5, the compound word dictionary 5 is searched using the corresponding reading as a key. If the search result is not "empty", the result is reflected in the compound word storage means 3. In FIG. 15, the second part of the address 311 in the index part 31
Taking the reading of the address a12 of "Yuukyushou" as an example, since the corresponding notation is not registered in the compound word dictionary 5 shown in FIG. 19, the search result is "empty". On the other hand, with respect to the readings of "Kinoyushou" and "jishokensaku", the expressions "functional specifications" and "dictionary search" are registered in the compound word dictionary 5 of FIG. 19 described above. This search result is reflected in the compound word storage means 3.

FIG. 20 exemplifies the contents of the compound word storage means 3 in which the search results of the compound word dictionary 5 are reflected. Hereinafter, how to reflect this will be specifically described. The contents of the compound word storage means 3 before the above reflection is shown in FIG.
As shown in. Here, the number of notations 313 of the heading section 31 including the reading "Kinoyu Shou" (= 2: FIG.
Only the value shown in 5) follows the notation linked by the address (address) 323 (to the next notation) of the notation part 32, and arrives at the “induction use” of the notation part 32 indicated by the address a23. . Then, the notation information regarding the “functional specification” is stored as the notation next to the notation unit 32 indicated by the address a23. That is, the address 311 of the header section 31
The number of notations 313 corresponding to the first address a11 in
And the notation information (the notation information regarding the “functional specification”) is increased to 3, and the next available address a26 of the notation unit 32 is given.
And the address (address) 323 to the next notation corresponding to the address a23 is changed from a21 to a26. Then, a21 is stored in the address (address) 323 to the next notation corresponding to this address a26. Since the compound word "functional specification" here is registered in the compound word dictionary 5, the dictionary type 326 includes d
The information ic2 is stored. Next, "jishokensaku"
The number of the notation number 313 corresponding to the third address a13 in the address 311 of the heading section 31 including the reading (= 1: the value shown in FIG. 15) in the notation section 32 (to the next notation) The notation linked by the address (address) 323 is traced to arrive at the notation “dictionary construction policy” indicated by the address a25 in the notation section 32. Here, the notation information related to the other notation “dictionary search” is stored as the notation next to the notation part indicated by the address a25. That is, the third address a1 in the address 311 of the heading section 31.
The number of notations 313 corresponding to 3 is increased by 1 to 2 and the notation information (notation information regarding “dictionary search”) is stored in the address a27 of the next available notation unit 32, and in accordance with this,
Address (address) 3 to the next notation corresponding to address a25
23 is changed from a25 (in FIG. 15) to a27. Then, a25 is stored in the address (address) 323 to the next notation corresponding to this address a27. The compound word "dictionary search" here is also a compound word dictionary 5.
Since it has been registered in, the information dic2 is similarly stored as the dictionary type 326.

The pointing means 4 advances the processing to S7. The operation in S7 is similar to that of the first embodiment, and the operation is as shown in FIG.
FIG. 21 shows an example in which the position of a compound word having a plurality of notations for the same reading is displayed. What is recognized in comparison with FIG. 16 relating to the first embodiment is that in the display of the second embodiment in which the compound word dictionary 5 is used,
The compound word “dictionary construction” included in the sentence number 105 is underlined.

For a compound word having a plurality of notations for one reading, compound words of other homophones can be easily grasped by checking the compound word storage means 3.
For example, the other notation of “function use” of sentence number 101 in FIG. 21 is included from the position 1011 to the address a21 of the notation part 32. This address a21 is temporarily stored as a start address, and the address to the next notation is traced until the temporarily stored start address. When making this trace,
If the corresponding notations are sequentially extracted in the notation section, a plurality of notations can be listed for one reading.
Now, regarding the case of “function use” of sentence number 101 in FIG. 21, there are other expressions such as “induction use” in the target document, and “function specification” in the compound word dictionary 5. Is recognized. In this case, it is possible to determine that compound words other than "functional specification", "functional use" and "inductive use" are incorrect, and the "functional specification" is correct for the "functional use" of the sentence number 101. It can be presented as a compound notation. Regarding the other notation of "inductive use" of sentence number 103, "function use" and "functional specification" are similarly obtained by tracing the address to the next notation.
Can be asked. Even in this case, it is possible to present that the "induction use" is incorrect and the "functional specification" is correct.

[0052]

As described above in detail, the document proofreading support apparatus according to the present invention uses, as a key, the reading of the corresponding compound word in the compound word information detected from the sentence input / stored from the outside. It is configured so as to include means for temporarily storing (compound word storage means 3) so that a compound word having the same reading but different notation can be accurately detected. Further, another document proofreading support apparatus according to the present invention is configured to further include a compound word dictionary 5. According to the pointing means 4, the compound word storage means 3 and the compound word dictionary 5 are collated. Thus, it also has a function of judging / pointing out whether or not different compound words exist for a certain predetermined same reading. Then, the device according to the present invention checks the possibility of an error with respect to the compound word included in the sentence to be processed, and if there is a high possibility of the error, it is possible to call attention to that effect. When a predetermined compound word in the target sentence is detected, it is possible to point out the possibility of a homonym error due to the inconsistency of the notation of the compound word not registered in the compound word dictionary. . Further, the notation groups of homonyms in the document can be easily listed. Furthermore, when a compound word dictionary is used, the compound word dictionary is not searched each time a compound word is detected, but the related compound word is detected from the document, and then the reading is used as a key. Since the search is performed, the number of searches can be steadily reduced, and the desired efficient processing can be achieved.

[Brief description of drawings]

FIG. 1 is a block diagram illustrating a configuration of a document proofreading support apparatus according to a first embodiment of the present invention.

FIG. 2 is a flow chart for explaining the operation of the first embodiment device.

FIG. 3 is a flowchart for partially partially explaining the operation of the first embodiment device (and a second embodiment device described later) in detail.

FIG. 4 is an illustrative view of a sentence used for explaining the operation of the device of the first embodiment (and the device of the second embodiment described later).

FIG. 5 is a diagram showing an example of contents of an analysis dictionary used for explaining a configuration and an operation of the first embodiment device (and a second embodiment device described later).

FIG. 6 is a diagram showing an example of contents of analysis result storage means used for explaining the configuration and operation of the apparatus of the first embodiment.

FIG. 7 is a diagram showing an example of contents of analysis result storage means used for explaining the configuration and operation of the apparatus of the first embodiment.

FIG. 8 is a diagram showing an example of contents of analysis result storage means used for explaining the configuration and operation of the apparatus of the first embodiment.

FIG. 9 is a diagram showing an example of contents of analysis result storage means used for explaining the configuration and operation of the apparatus of the first embodiment.

FIG. 10 is a diagram showing an example of contents of analysis result storage means used for explaining the configuration and operation of the apparatus of the first embodiment.

FIG. 11 is a diagram showing an example of the contents of a compound word storage means used for explaining the configuration and operation of the apparatus of the first embodiment.

FIG. 12 is a diagram showing an example of contents of a compound word storage means used for explaining a configuration and an operation of the apparatus of the first embodiment.

FIG. 13 is a diagram showing an example of contents of a compound word storage means used for explaining a configuration and an operation of the apparatus of the first embodiment.

FIG. 14 is a diagram showing an example of contents of a compound word storage means used for explaining a configuration and an operation of the apparatus of the first embodiment.

FIG. 15 is a diagram showing an example of contents of compound word storage means used for explaining the configuration and operation of the apparatus of the first embodiment.

FIG. 16 is an exemplary display diagram used for explaining the operation of the apparatus of the first embodiment.

FIG. 17 is a block diagram for explaining the configuration of a document proofreading support apparatus according to the second embodiment of the present invention.

FIG. 18 is a flow chart for explaining the operation of the second embodiment device.

FIG. 19 is a diagram showing an example of contents of a compound word dictionary used for explaining a configuration / operation of the apparatus of the second embodiment.

FIG. 20 is a diagram showing an example of the contents of a compound word storage means used for explaining the operation of the apparatus of the second embodiment.

FIG. 21 is an exemplary display diagram used for explaining the operation of the second embodiment apparatus.

[Explanation of symbols]

1-Sentence storage means; 2--Compound word detection means; 21 ---
22 --- analysis dictionary; 23--analysis result storage means; 3--compound word storage means; 4--pointing means; 5-
A compound word dictionary.

Claims (2)

[Claims] 1. A sentence storage means for storing an inputted sentence,
A compound word storage unit that stores compound word information detected from the sentence storage unit, a compound word is detected from a sentence stored in the sentence storage unit, and a compound word notation and the sentence storage unit using the reading of the compound word as a key A compound word detecting means for registering the storage start position in the compound word storing means in the compound word storing means, an indicating means for pointing out a compound word having a plurality of notations for a certain reading in the compound word storing means, A document proofreading support device characterized by comprising: 2. The system further comprises a compound word dictionary in which the reading of the compound word and its notation are registered, and the pointing means collates the compound word storage means with the compound word dictionary to obtain a certain same reading. The document proofreading support apparatus according to claim 1, wherein the presence / absence of a different compound word is also determined and pointed out.