JPH08115321A - Document preparation support device - Google Patents

Abstract

PURPOSE: To provide a document preparation support device which supports document preparation by retrieving a next cooccurrence word linkingly with a cooccurrence word retrieved in a cooccurrence word dictionary in order and displaying relative examples. CONSTITUTION: This device is equipped with a cooccurrence relation holding means 1 which holds the intensity of the cooccurrence relation of a 2nd word with a 1st word, an example data holding means 2 which holds example data, specific word input means 11 and 21 which specify and input optional words, cooccurrence word display means 12 and 22 which display lists of the 2nd words in cooccurrence relation with the 1st specific word and specify one of the words, and example data display means 13 and 23 which receive the specific cooccurrence word corresponding to the cooccurrence word display means 12 and 22, retrieve examples containing both the cooccurrence words and the specific words specified by the specific word input means, and display the retrieval results.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention sequentially searches for a next co-occurrence word from a co-occurrence word retrieved by a co-occurrence word dictionary in a chained manner and presents related examples to assist document creation. The present invention relates to a document creation support device.

[0002]

2. Description of the Related Art Conventionally, for example, the literature "Hiromi Saito, Hiroyasu Nogami," Natural Language Processing in Japanese Word Processors ", Information Processing, Vol. 34, no. 10, 12th
As described in “41 to 1248, 1994”, the use of co-occurrence words is an important technique in the kana-kanji conversion system and the document creation support system. The term “co-occurrence word” here means a word pair (group) that frequently appears at the same time in many document examples in sentences or paragraphs.

A general example of the use of such a co-occurrence word will be described. For example, as a first example of use, the co-occurrence word dictionary is registered in the co-occurrence word dictionary from among the homophone word candidates of the syllable series as a result of conversion in kana-kanji conversion. A word group that co-occurs with a certain word in the document creation support system, as a usage example in which the correct answer candidate is raised to the first rank by preferentially outputting a word that matches a given word pair. By showing the user to, there is an example of a usage example for supporting the idea of creating a document.

For example, the proposal of "Document Creation Device" disclosed in Japanese Patent Laid-Open No. 4-268945 is a co-occurrence function used together with example information in a document creation device having an example display function and a co-occurrence function. This is a proposal regarding a document creation apparatus that can display more examples by using information. According to such a document creation device, in order to assist the document creation, the user is shown an example including the word and both the word and the word co-occurring with the word. It can be used as a reference for creation.

[0005]

However, in the conventional technique as described above, as the example data stored in the example dictionary becomes larger, the example including both the designated word and the word that co-occurs with the specified word. There will be a large number of, and unnecessarily many examples will be output. Therefore, when supporting the document creation, for example, there is a problem that the limit of the amount of text that the user can refer to by operating the sub-window on the display screen is exceeded.

[0006] Further, in order to support the document creation related to the document currently being edited, many examples unrelated to the subject that the user wants to refer to are displayed. Furthermore, it is difficult to find a co-occurrence word that exists in the example because the word that has a co-occurrence relationship with the specified word is not explicitly shown to the user. It is not easy to use the word of origin.

Further, in addition to the document creation support, the co-occurrence word of the co-occurrence word can be obtained even if it is desired to obtain the co-occurrence word for the co-occurrence word existing in the example in order to support the idea generation therein. There is a problem that it is not possible to perform a sequential chain search. Even if it is desired to search multiple co-occurrence words in a chain and retrieve an arbitrary word set from them to obtain a word that co-occurs with them, it must be done by a complicated manual operation to support the idea generation. In addition, the co-occurrence word dictionary is not constructed so that it can be searched in a chained manner, and the co-occurrence word dictionary is not sufficiently used to support such an idea.

As described above, it is an important usage of co-occurrence words to support an idea generation in document creation, by sequentially searching the co-occurrence words in a chained manner with reference to examples. However, there has been a problem that no document creation support device has realized the use of this type of co-occurrence word.

The present invention has been made to solve these problems, and an object of the present invention is to successively search for a next co-occurrence word from a co-occurrence word retrieved from a co-occurrence word dictionary. An object of the present invention is to provide a document creation support device that can perform chained operations and present related examples to provide document creation support and also idea creation support.

[0010]

In order to achieve the above-mentioned object, the document creation support apparatus, which is the first feature of the present invention, has the strength of the co-occurrence relation between the first word and the second word. A co-occurrence relation holding means (1), an example data holding means (2) holding example data, and designated word input means (11, 21) for designating and inputting an arbitrary word, respectively. Corresponding to the designated word input means of, a list of second words having a co-occurrence relationship with the designated first word is displayed, and each of the co-occurrence word display means (one of the designated word is designated. 12,
22) and corresponding co-occurrence word display means, receive a designated co-occurrence word, search for an example including both the co-occurrence word and the designated word by the designated word input means, and display the search result. It is characterized by including each example data display means (13, 23).

The second aspect of the document creating apparatus of the present invention further obtains the co-occurrence word displayed by each co-occurrence word display means from the co-occurrence relation holding means, and follows a sequential display chain process. The present invention is characterized by comprising a co-occurrence word chain processing means (3) for storing the same and a co-occurrence word chain display means (4) for displaying the display chain process stored in the co-occurrence word chain processing means.

A third aspect of the present invention, a document creation support device, is an example data holding means (7) for holding example data.
2) and the co-occurrence relation between the words, the first co-occurrence degree of the strength of the co-occurrence relation of the second word with respect to the first word, and the co-occurrence relation of the first word with respect to the second word. A co-occurrence relationship holding means (71) for respectively holding the second co-occurrence degree of the strength of
Any of the first co-occurrence degree or the second co-occurrence degree held by the co-occurrence relationship holding means for the word input by the word designation input means (61, 81) for designating a word A co-occurrence word search means (62, 82) for searching one of the two and outputting a word whose co-occurrence relationship strength exceeds a predetermined threshold value as a co-occurrence word of the input word, and the co-occurrence word search means. Co-occurrence word display instruction means (63, 83) for displaying the output co-occurrence words in a list, selecting and instructing a designated word of the co-occurrence word to be searched next from the displayed co-occurrence words, and the co-occurrence word. Example data search means (64, 84) for searching the example data holding means for example data containing both the co-occurrence word designated by the word display designating means and the word input by the word designation input means, and the example data search The example data retrieved by the means Shimesuru example data display means (6
5, 85).

[0013]

In the document creation support device of the first aspect of the present invention, the co-occurrence relation holding means (1) holds the strength of the co-occurrence relation of the second word with respect to the first word. The example data holding means (2) holds the example data.
When the user designates and inputs an arbitrary word by one of the designated word input means (11, 21), the corresponding 1
One co-occurrence word display means (12, 22) displays a list of second words having a co-occurrence relationship with the first word designated by the designated word input means, and one word can be designated from the list. To do so. In addition, each example data display means (13,
23) A corresponding example data display means receives a co-occurrence word designated by each corresponding co-occurrence word display means, and includes both the co-occurrence word and the designated word by the designated word input means. To display the search results. Accordingly, the user can sequentially search the co-occurrence words in a chained manner by designating a word from the displayed list of co-occurrence words, and display the example.

In the second aspect of the document creating apparatus of the present invention, the co-occurrence word chain processing means (3) holds the co-occurrence relations of the co-occurrence words displayed by the respective co-occurrence word display means. It is obtained from the means, and the sequential display chain process is traced and stored. The co-occurrence word chain display means (4) displays the display chain process stored in the co-occurrence word chain processing means. For this reason, the user can display the result of the operation of sequentially searching the co-occurrence words in a chain as a co-occurrence word chain process. The process can be used to support the idea.

In the document creation support device of the third aspect of the present invention, the example data holding means (72) holds the example data, and the co-occurrence relation holding means (71) holds the data between words. As the co-occurrence relation, the first co-occurrence degree of the strength of the co-occurrence relation of the second word with respect to the first word and the second co-occurrence of the strength of the co-occurrence relation of the first word with respect to the second word Holds the degrees and.

When the word designation input means (61, 81) designates a word, the cooccurrence word search means (62, 82) performs the cooccurrence relationship holding means for the word input by the word designation input means. Either the first co-occurrence degree or the second co-occurrence degree held by is searched, and a word having a strength of the co-occurrence relationship exceeding a predetermined threshold is output as a co-occurrence word of the input word. The co-occurrence word display instructing means (63, 83) displays the co-occurrence words output by the co-occurrence word searching means in a list, and displays the designated word of the co-occurrence word to be searched next from the displayed co-occurrence words. Select and instruct. Then, the example data search means (64, 8
4) searches the example data holding unit for example data containing both the co-occurrence word designated by the co-occurrence word display designating unit and the word input by the word designation input unit. In response to this result, the example data display means (65, 85)
Displays the example data retrieved by the example data retrieval means.

As described above, in the document creation support apparatus of the present invention, a list of words that co-occur with a word is displayed when the word is specified, and the co-occurrence word group is selected from the co-occurrence word group. When one co-occurrence word is designated, a plurality of co-occurrence word display means and a plurality of example data display means for displaying an example including both the designated word and the designated co-occurrence word are prepared. While referring to the displayed example, the user identifies a co-occurrence word that is highly related to the document being created from the displayed co-occurrence word group, and further identifies the co-occurrence word related to that co-occurrence word. It is possible to search for an origin and search for its example. By repeating such a search operation, a co-occurrence word that is highly related to the document being created can be searched in a chained manner, and even document creation support and idea creation support can be performed. Further, since the chain process of the co-occurrence words searched in a chain can be referred to, it is possible to create the document while confirming the chain of the co-occurrence words.

Since the co-occurrence word dictionary used for supporting such document creation searches the co-occurrence words in a chained sequence, the co-occurrence relationship is particularly limited to words having a strong co-occurrence relationship. It is desirable that it is set. Regarding the co-occurrence relation, it is preferable that the degree of the co-occurrence relation and the directionality of the co-occurrence relation are set. For this reason, the co-occurrence word dictionary here holds the strength of the co-occurrence relationship between the first word and the second word, or the co-occurrence relationship between the first words. And a second co-occurrence degree of the strength of the co-occurrence relation of the second word with respect to and a strength of the co-occurrence relation of the first word with respect to the second word, respectively. .

[0019]

Embodiments of the present invention will be specifically described below with reference to the drawings. FIG. 1 is a block diagram showing the basic configuration of the document creation support device according to the first embodiment of the present invention.
In FIG. 1, 1 is a co-occurrence relation holding unit, 2 is an example data holding unit, 3 is a co-occurrence word chain processing unit, and 4 is a co-occurrence word chain display unit. Also, 10 is the first search window and 11 is the first
Designated word input unit, 12 is a first co-occurrence word display unit, 13 is a first example data display unit, 20 is a second search window,
Reference numeral 21 is a second designated word input unit, 22 is a second co-occurrence word display unit, and 23 is a first example data display unit.

In the document creation support device shown in FIG.
As will be described later, the co-occurrence relation holding unit 1 holds the strength of the co-occurrence relation of the second word with respect to the first word as a co-occurrence word dictionary, and the example data holding unit 2 holds the co-occurrence word. It also holds the example data.

In the first search window 10, the first
When the user designates an arbitrary word through the designated word input unit 11, the corresponding first co-occurrence word display unit 12 receives the word designated by the first designated word input unit 11, From the co-occurrence relation of the word held in the origin relation holding unit 1, the first
A list of second words having a co-occurrence relation with the word is displayed, and one word can be designated from the list. For example, the strengths of co-occurrence relationships are displayed in a list, and one word can be designated for the search of the next co-occurrence word. As a result, the corresponding first example data display unit 13 receives the co-occurrence word designated by the first co-occurrence word display unit 12, and the corresponding co-occurrence word and the preceding first designated word input unit 11 are received. The example including the specified word specified by is searched, and the search result is displayed to the user.

When one word is selected from the list of co-occurrence words displayed in the first co-occurrence word display section 12 of the first search window 10, the selected word is It is input to the second designated word input unit 21 of the second search window 20 as a designated word for the co-occurrence word search.
As a result, in the second search window 20 as well, similar to the operation in the first search window 10, the cooccurrence word search and display and the display of the example thereof are performed. In addition,
Although not shown, also in the third search window, the fourth search window, ..., By repeating such search display operation in sequence, the co-occurrence words are sequentially searched in a chained manner. Examples can be displayed to the user.

In this manner, when the co-occurrence words are sequentially searched in a chained manner, each co-occurrence word displayed by the co-occurrence word is subjected to a display chain process in the co-occurrence word chain processing unit 3. It is traced and stored. That is, the co-occurrence word chain processing unit 3
However, the co-occurrence word displayed in each co-occurrence word display unit (12, 23) is obtained from the co-occurrence relation data of the co-occurrence relation holding unit 1 and stored by following a sequential display chain process. Then, the co-occurrence word chain display unit 4 displays the display chain process stored in the co-occurrence word chain holding unit 3. As a result, the result of the operation in which the co-occurrence words are searched in a chain is displayed as a chain process. The user can use it for supporting the idea of the chain starting result of the operation so far.

The co-occurrence word dictionary held in the co-occurrence relationship holding unit 1 holds the strength of the co-occurrence relationship of the second word with respect to the first word. , Used as a co-occurrence dictionary to support ideas. Since the co-occurrence word dictionary is preferably structured in a data structure that enables sequential search of co-occurrence words, the strength of the co-occurrence relationship is set between words having a strong co-occurrence relationship. I shall. Therefore, as the co-occurrence word dictionary here, the strength of the co-occurrence relationship between the first word and the second word is retained, or the co-occurrence relationship between the words is determined as the first co-occurrence relationship.
And a second co-occurrence degree of the strength of the co-occurrence relation of the second word with respect to the word and a strength of the co-occurrence relation of the first word with respect to the second word, respectively. The directionality when searching the co-occurrence relation of one from the other can be determined.

FIG. 2 is a diagram showing an example of the data structure of the co-occurrence word dictionary held in the co-occurrence relation holding unit. Figure 2
As shown in, the data structure of the co-occurrence word dictionary 25 is given in a tabular format. Each cell (column) 26 in the table is provided corresponding to all word sets. In each cell 26, three pieces of data are stored as a co-occurrence relationship between two words, a word a and a word b, which form a set. This 3
If one data is c1, c2, c3 from the left, c
The strength Wab of the co-occurrence relationship stored as 1 is the word a
It is a numerical value indicating the strength of the co-occurrence relation of the word b (word b) with respect to (word a), and the strength Wba of the co-occurrence relation stored as c2 is conversely the word b (word b).
Is a numerical value indicating the strength of the co-occurrence relationship of the word a (word a) with respect to. However, there is no Wab for which a = b. Further, the data F stored as c3 is a variable indicating the search process of the co-occurrence word. The value of the data F of the variable indicating the search process indicates the direction of the co-occurrence process of the word set corresponding to each cell when the co-occurrence word search is performed in a chain. Using the value of this data F, a process for tracing the chain of co-occurrence word search processes is performed as will be described later.

That is, in the co-occurrence word dictionary 25, cell 2
When the co-occurrence relation strengths stored as c1 and c2 in 6 are Wab and Wba, the value of the data F stored as c3 indicates the process of searching for the co-occurrence word. If the value of the data F is “0”, it indicates that the word a and the word b have not been displayed (not searched) as co-occurrence words, and if the value of the data F is “1”, the designated word It indicates that the word b is displayed as the co-occurrence word in the co-occurrence word display unit 12 with respect to the word a. Also, the data F
A value of 2 indicates that the word a is displayed as the co-occurrence word on the co-occurrence word display unit 12 with respect to the word b of the designated word. Further, if the value of the data F is "3", it means that the display is made in these two cases. That is, it indicates that the word b is displayed as the co-occurrence word for the word a of the designated word, and the word a is already displayed as the co-occurrence word for the word b of the designated word.

Further, in general, the word b corresponds to the word a.
The strength (importance) Wab of the co-occurrence relation of Wab does not match the strength (importance) Wba of the co-occurrence relation of the word a with respect to the word b. Therefore, Wab and Wba often have different values. Since the co-occurrence relation is compared, the numerical value of Wab needs to be standardized in the co-occurrence word dictionary 25, so the numerical value of Wab is normalized and held here.
Next, the algorithm when the co-occurrence word dictionary 25 is constructed from the example data will be described.

The co-occurrence strength Wab between two words stored in each cell 26 in the co-occurrence word dictionary 25 is obtained by counting the number of co-occurrence times between two words from the reference example data. And standardize the data as the strength of the co-occurrence relationship. The strength Wab of the co-occurrence relation between the two words in this case is determined by the algorithm of the following steps (1) to (3). Step (1): Words included in the example data (number N)
Variables Cij (1 ≦ i ≦
N, 1 ≦ j ≦ N, except when i = j) is prepared. Note that Cij = Cji. Step (2): For each variable Cij, the number of times that the word i and the word j co-occur in the example data (the number of times they simultaneously appear in one sentence or paragraph) is stored. Step 3: The value of the variable Cij is standardized for each word based on each of the following equations, and the strength of the co-occurrence relation Wkl (≠
Wlk) is determined. Ck = (ΣCkm) / N (Σ is the sum of m) σk = (Σ (Ckm− Ck ) ² / N) ^1/2 (Σ is the sum of m) Wkl = 0.5 + 0.1 (Ckl− Ck / σk )

Based on the algorithm of steps (1) to (3), the strength of the co-occurrence relation Wa
The numerical data of b is determined and stored in the cell according to the data structure of the co-occurrence word dictionary of the co-occurrence relation holding unit. That is, the co-occurrence word dictionary here is an average value ( Ck ) of the number of times one word co-occurs with each of all the other words, and
The deviation (σk) is calculated, and the co-occurrence strength Wab is converted into numerical data so that the central value becomes “0.5”.
Has been standardized. FIG. 3 shows an example of numerical data of the co-occurrence word dictionary obtained as a result of calculation by such an algorithm.

For example, when the word k for which a co-occurrence word is requested is specified by the specified word input unit 11, the co-occurrence word display unit 12 receives the specified word k and holds it in the co-occurrence relation holding unit 1. The co-occurrence word is searched based on the co-occurrence word dictionary (FIG. 2). An algorithm for performing the co-occurrence word search in that case will be described. In the co-occurrence word search process in this case, the co-occurrence word search process is executed by the following steps to steps. Step: Search for Wki exceeding the threshold value s from the values Wki (1 ≦ i ≦ N, i ≠ k) stored in the corresponding cell of the co-occurrence dictionary for the word k. Step: W obtained as a result of step search
From the cell storing ki, the word i of the corresponding set is displayed as a co-occurrence word. At this time, when the cell storing the value Wki stores the value Wki as c1 and the value of the data F is “1” or “3”, the value Wk is
If the cell storing i stores the value Wki as c2 and the value of the data F is “2” or “3”, a value indicating “displayed” is output. Step: Furthermore, the value of the data F is updated with respect to the cell storing Wki obtained as a result of the step search. That is, the cell was searched first,
If the value of the data F is “1” and the value Wki is stored as c2, the value of the data F is set to “3”. Also,
If the value of the data F is “2” and the value Wki is stored as c1, the value of the data F is set to “3”. Also,
If the cell here is not searched first, the value of the data F is “0”, and the value Wki is stored as c1, the value of the data F is set to “1”. Similarly, if the value of the data F is “0” and the value Wki is stored as c2, the value of the data F is set to “2”. Otherwise, do nothing.

In this way, the number of co-occurrences for each word is obtained from the example data, and the co-occurrence between each word is further calculated according to the above-mentioned algorithm "step (1) to step (3)". The strength Wab of the origin relationship is determined and a co-occurrence word dictionary is constructed. Further, using the constructed co-occurrence word dictionary, the co-occurrence word display unit 12 searches for and displays the co-occurrence word from the designated word according to the strength of the co-occurrence relationship between the words.

Next, the co-occurrence word search process in the co-occurrence word display unit 12 will be described according to a concrete example. In the process here, the co-occurrence word dictionary has a data structure having the above-described three data, and therefore, according to the search result,
This involves a process of determining the value of the data F indicating the search process.

FIG. 5 is a flowchart showing the processing flow of the co-occurrence word search processing. This co-occurrence word search processing has a processing flow in accordance with the above-mentioned algorithm “step to step”. The co-occurrence word search process will be described with reference to FIG. When the process starts, first, step 31
At, the designated word (word k) is received. Next, in step 32, the specified word (word
For k), the value Wki (1 ≦ i ≦ N, i ≠ k) stored in the corresponding cell of the co-occurrence dictionary is searched for Wki exceeding the threshold s.

Wki exceeding the threshold value s obtained by the search
The following steps 33 to 44 are performed for all of the above, and the process of outputting the co-occurrence word according to the search result and determining the value of the data F indicating the search process is involved.

In the process from step 33 for one Wki exceeding the threshold value s, first, in step 33, the obtained Wki is determined whether or not the cell storing the Wki stores the value Wki as c1. To determine. As a result of this determination, when the obtained Wki is stored in c1, it is further determined in step 34 whether or not the value of the data F of the cell is “F = 1 or F = 3”. When the value of F is “F = 1 or F = 3”, in step 35, the corresponding word of this cell is output as a co-occurrence word (word i) together with a value indicating “displayed”. In this case, the value of the data F is left as it is, and the process for this Wki ends.

If the value of the data F in the cell is not "F = 1 or F = 3" in the determination in step 34, the corresponding word in this cell is determined as the co-occurrence word (word i) in step 36. Output. Further, in the next step 37, the value of the data F of the cell is “F =
It is determined whether it is "0". If the value of F is "F = 0", the value of the data F in the cell containing this Wki is set to "1" in step 38, and the processing for this Wki is terminated. If the value of F is not “F = 0” in the determination of step 37, the value of F is “F = 2” in this case. Therefore, in step 39, the data in the cell containing Wki here is written. The value of F is set to "3", and the process for this Wki is completed.

On the other hand, in the determination in step 33,
If it is determined that the obtained Wki is not stored in c1, the obtained Wki is stored in c2 of the cell, and therefore the same process is performed corresponding to the case where the direction of the search process is reversed. I do. That is, in this case, step 40
And the value of the data F of the cell is “F = 2 or F =
3 ”is determined. When the value of F is “F = 2 or F = 3”, in step 41, the corresponding word of this cell is output as the co-occurrence word (word i) together with the value indicating that it has been displayed. In this case, the value of the data F is left as it is, and the process for this Wki ends.

If the value of the data F in the cell is not "F = 2 or F = 3" in the determination in step 40, the corresponding word in this cell is determined as the co-occurrence word (word i) in step 42. Output. Further, in the next step 43, the value of the data F of the cell is “F =
It is determined whether it is "0". When the value of F is “F = 0”, the value of the data F in the cell including this Wki is set to “2” in step 44, and the process for this Wki ends. When the value of F is not “F = 0” in the determination of step 43, the value of F is “F = 1” in this case, and therefore, in step 39, the data in the cell including Wki here is written. The value of F is set to "3", and the process for this Wki ends.

The processes of steps 33 to 44 are performed for all the retrieved Wkis,
Output the co-occurrence word. As a result of this processing, a group of co-occurring words is obtained for the designated word (word k).

In order to understand the co-occurrence word search processing here, for example, the case where the strength of the co-occurrence relationship is stored in the co-occurrence word dictionary 25 as concrete numerical data as shown in FIG. An example of the processing will be described. FIG. 3 shows an example of numerical data of the co-occurrence word dictionary before the co-occurrence word search process is performed, and FIG. 4 shows an example of numerical value data of the co-occurrence word dictionary after the co-occurrence word search process is performed. ing.

It is assumed that a designated word (word 4) has been designated in the designated word input section 11. When the co-occurrence word is searched, the threshold value s of the strength of the co-occurrence relation is “0.
5 ″. That is, in this case, as shown in FIG.
W41, W42, W43, W45, W of the co-occurrence dictionary 27
Two Ws exceeding the threshold value "0.5" out of 46 and W47 (numerical data shown by being shaded)
42 and W47 (numerical data underlined) are searched. According to this search result, the word in the corresponding cell of W42 and W47 is output as a co-occurrence word, and W
The value of the data F of the corresponding cell storing 42 and W47 is updated.

That is, in this case, according to the search result, W
From 42 and W47, the word (word 2) and the word (word 7) are output as co-occurrence words for the designated word (word 4). As shown in FIG. 3, W4
Since the values of the data F in the cell including 2 and W47 are both “0” before the co-occurrence word search process is performed, the value indicating “displayed” is not output. Further, since W42 is stored as c1 in the cell of the co-occurrence word dictionary and the value of the data F is “0”, the value of the next data F of this cell is updated to “1”. Since W47 is stored as c2 and the value of the data F is "0", the value of the next data F of this cell is updated to "2". As a result, as shown in FIG. 4, the numerical data in the co-occurrence word dictionary 28 is updated.

In this way, the co-occurrence word is searched a plurality of times, the value of the data F in the co-occurrence word dictionary is updated, and the co-occurrence word is then updated based on the updated value of the data F. The process of following the chain of is performed. This processing is performed by the co-occurrence word chain processing unit 3
Performs processing by referring to the co-occurrence dictionary of the co-occurrence relation holding unit 1, and performs processing for tracing the chain process of display co-occurrence words. FIG.
It is a flowchart which shows the search processing flow of a cooccurrence word chain.

The co-occurrence word chain search processing will be described with reference to FIG. When the process is started, in step 51, the first designated word (word k) for tracing the co-occurrence word chain.
To receive. Next, at step 52, the designated word number k is substituted for the work variable x. Next, step 53
In the process of step 56, the cell related to the designated word (word k) in the co-occurrence word dictionary is searched, the value of the data F is discriminated, and the co-occurrence word chain is sequentially traced. A cell having a value of data F of "0" is excluded because it is not displayed and output as a co-occurrence word.

That is, in step 53, Wxi
Regarding (1 ≦ i ≦ N, i ≠ x), “Wxj in which the value of the data F is“ 1 ”or“ 3 ”when Wxi is included in the cell c1 is determined according to the direction in which the cooccurrence chain is sequentially traced. , Or “Wxj in which the value of the data F is“ 2 ”or“ 3 ”when Wxi is included in the cell c2” is searched. next,
In step 54, the word (word j) obtained from the search result Wxj is output as the cooccurrence word of the result of the cooccurrence word search from the word (word x), with its direction clearly indicated. With this output, in the past co-occurrence word search, when the word (word x) is the designated word, the word (word j) is displayed as the co-occurrence word of the search result. By storing it in, the chain of co-occurrence words can be traced. Next, step 55
In, it is determined whether or not the processing has been recursively performed for all Wxj, and if the processing has not been completed for all Wxj, the process proceeds to step 56, and the number j of the co-occurrence word is changed to the next designated number. It is substituted as the next number of the work variable x, the process returns to step 53, and the processing from step 53 is repeated. As a result, the co-occurrence words of the specified word are sequentially searched in sequence based on the value of the data F in the co-occurrence word dictionary. If it is determined in step 55 that the processing has been completed for all Wxj, the processing here is ended.

By using the value of the data F of such a co-occurrence word dictionary to sequentially trace the co-occurrence words from the designated word, it becomes possible to display the chain of past co-occurrence word searches. . By displaying such a chain of co-occurrence words, co-occurrence word search can be used not only for document creation support but also for idea creation support.

FIG. 7 is a block diagram showing the configuration of the document creation support apparatus of the second embodiment of the present invention as a combination of program processing modules by a software processing mechanism. In FIG. 7, 60 is the first search window,
61 is a word designation section, 62 is a co-occurrence word search section, 63 is a co-occurrence word table / designation section, 64 is an example data search section, 65 is an example data display section, 70 is a document editing window, 71 is a co-occurrence dictionary. A holding unit, 72 is an example data holding unit, 73 is a co-occurrence word chain processing unit, 74 is a co-occurrence word chain display unit, 75 is a word designation unit,
76 is a co-occurrence word chain search window, 80 is a second search window, 81 is a word designation section, 82 is a co-occurrence word search section, 8
3 is a co-occurrence word table / designating unit, 84 is an example data search unit, 85
Is an example data display section.

The document edit window 70 includes an input / output processing module of a text input processing section 70a and an edit screen display processing section 70b. The edit screen is displayed and the user operates the document creation. In the document creation here, if the document creation support and the idea creation support are desired, the first search window 60, the second search window 80,
The co-occurrence word chain search window 76 and the like are sequentially opened, and in each window, a co-occurrence word sequential search operation or a co-occurrence word chain search operation is performed for the document being created.

The co-occurrence word dictionary holding unit 71 stores a set of words,
It is a module that provides a set of words having a co-occurrence relationship with each word and a storage area for storing the search process of the co-occurrence words. For example, it holds the co-occurrence word dictionary as shown in FIG. The example data storage unit 72 is a module that provides a storage area for storing example data for creating a document. Corresponding to the document to be created, a set of documents that serves as a reference example collection is stored. That is, the example data holding unit 72 holds the example data, and the co-occurrence relationship holding unit 71 defines the co-occurrence relationship between the words created from the example data as the second word with respect to the first word. The first co-occurrence degree of the strength of the co-occurrence relation and the second co-occurrence degree of the strength of the co-occurrence relation of the first word with respect to the second word are held respectively.

Before describing the operation of the document creation support device here, the contents of the program processing module operating in each window will be described. Note that the program processing module (6
1 to 65) correspond to the processing modules (81 to 85) in the second search window 80 and have the same contents, the contents of the program processing modules (61 to 65) in the first search window 60 are the same. Make it representative by explanation.

The word designating sections 61 and 75 are program processing modules having a user interface that allows the user to designate a word.
Is a module having a user interface for designating a word for a search for a co-occurrence word in the first search window 60, and the word designating unit 75 searches for a co-occurrence chain in the co-occurrence chain search window 76. It is a module having a user interface for specifying a word for.

The co-occurrence word search unit 62 receives the word designated by the word designation unit 61, and searches the information stored in the co-occurrence word dictionary holding unit 71 to find a word group having a co-occurrence relationship with the designated word. It is a program processing module to be obtained. When a word that has already been searched is obtained, a value indicating that the word has been searched is notified to the co-occurrence word display / designation unit 63 together with the word.

The co-occurrence word display / designation unit 63 displays the co-occurrence word group based on the search result from the co-occurrence word search unit 62, and the user interface through which the user can specify the words in the co-occurrence word group. Is a program processing module having.

The example data retrieving unit 64 includes a co-occurrence word and word designating unit 61 designated by the co-occurrence word displaying / designating unit 63.
Is a program processing module which receives a word designated by and retrieves an example containing both of them by searching the example data holding unit 72. The example data display unit 65 is a program module having a user interface for displaying the example data group based on the search result from the example data search unit 64.

The co-occurrence word chain processing unit 73 has a word designating unit 75.
This is a program processing module that performs a process of following the search process from the information of the search process of the co-occurrence word held by the co-occurrence word dictionary holding unit 74, as described with reference to FIG. In addition, the co-occurrence word chain display unit 74
Is a program module having a user interface for receiving the processing result of the co-occurrence word chain processing unit 73 and displaying the co-occurrence word search process.

Next, in the document creation support apparatus configured as described above, each program processing module is activated, a co-occurrence word search process is performed during document creation, and a co-occurrence word chain search is performed. I do. An operation example in this case will be described with reference to FIGS. FIG. 8 is a diagram showing a screen example of the document edit window, and FIG. 9 is a diagram showing a screen in a state where the first search window is opened from the document edit window. FIG. 10 is a diagram showing a screen in a state where a co-occurrence word search is subsequently performed in the first search window. Further, FIG. 11 is a diagram showing a screen in a state where an example is searched from the co-occurrence word in the first search window.

FIG. 12 is a diagram showing a screen in which a second search window for searching a co-occurrence word is opened following the first search window. FIG. 13 is followed by the second
FIG. 8 is a diagram showing a screen in a state where a co-occurrence word search is performed in the search window of and a further example search is performed. In addition, FIG.
FIG. 16 is a diagram showing a screen in a state where the third search window is subsequently opened from the second search window, and FIG.
It is a figure which shows the screen in the state which opened the 3rd search window for a co-occurrence word search which continues from the search window of FIG. FIG.
Similarly, FIG. 6 is a diagram showing a screen in a state where the co-occurrence word search is performed in the third search window subsequent to FIG.
FIG. 17 is a diagram showing a screen in a state where the example is subsequently searched from the co-occurrence word in the third search window. Further, FIG. 18 is a diagram showing a screen in a state in which a co-occurrence word chain search window is searched for a co-occurrence word chain.

First, referring to FIG. 8, for example, a document editor is activated in the document edit window 91,
When creating a document titled "1. Introduction" and wanting document creation support and idea creation support, a specific word is designated and a co-occurrence word is searched for. In this case, for example, the character string “autonomous distribution” of the word in the document is selected by operating the mouse cursor 90, and the pull-down menu opened by clicking the “co-occurrence” button on the menu bar at the top of the document editing window 91. Select the "Co-occurrence word search" menu of. As a result, the co-occurrence word search window 92 is opened as shown in FIG. When the co-occurrence word search window 92 is opened, the specified word “autonomous decentralized” when searching for a co-occurrence word is entered in the specified word input field 92a at the top of the co-occurrence word search window 92. ing.

In the operation of the program processing module shown in FIG. 7, such a series of operations is performed by the document editing window 70 while the first search window 60 is opened while a document is being created, The word is designated in the designation input section 61. Next, the co-occurrence word search unit 62 selects either the first co-occurrence degree or the second co-occurrence degree held by the co-occurrence relation holding unit 71 for the word input from the word designation input unit 61. Will be output as a co-occurrence word of the input word.

In the operation of the co-occurrence word search window 92 shown in FIG. 9, such co-occurrence word search processing is performed by clicking the "co-occurrence" button provided on the menu bar at the top of the co-occurrence word search window 92. By clicking with the operation of the mouse cursor 90, the co-occurrence word search process is started. As shown in FIG. 10, the search result of the co-occurrence word search is displayed in the list item field 92b provided on the left side of the co-occurrence word search window 92. Each co-occurrence word displayed in the list item field 92b becomes an instruction of a word for searching for an example when the user selects the one word by operating the mouse cursor 90, and the co-occurrence word is displayed. Contains examples are displayed. Further, as will be described later, it becomes an instruction of a designated word for the next co-occurrence word search when the next co-occurrence word search window is opened, and an input operation of the designated word.

That is, as shown in FIG. 11, by operating the mouse cursor 90, one word "tacit knowledge" is designated from the co-occurrence words displayed in the list item field 92b of the co-occurrence word search window 92. When the operation is performed, the example is searched, and the example data is displayed in the example list display field 92c on the right side of the cooccurrence word search window 92. In this case, the example data displayed here includes “implicit knowledge” of the designated word in the co-occurrence word displayed in the list item field 92b and the designated word “autonomous” input in the designated word input field 92a. An example including "dispersion" as a co-occurrence word is searched for, and the example of the search result is displayed in the example list display field 92c.

In the operation of the program processing module shown in FIG. 7, such a series of operations causes the co-occurrence word display / instruction unit 63 to display the co-occurrence words output by the co-occurrence word search unit 62 in a list. , It is an operation to select and instruct the specified word of the co-occurrence word to be searched next from the displayed co-occurrence word,
Furthermore, the example data search unit 64 and the co-occurrence word designated by the co-occurrence word display / instruction unit 63 and the word designation input unit 61
The example data holding unit 72 searches for example data containing both the words input by, and the example data display unit 65 receives the search result and displays the searched example data.

Further, subsequently, when the second search window 80 is opened to search for the co-occurrence word with the previously searched co-occurrence word as the designated word, the second search window 8
By the operation of the program processing modules 81 to 85 in 0, the same processing as in the case of the first search window 60 described above is performed. That is, to explain an operation example in this case, as shown in FIG. 11, when the first stage co-occurrence word search is performed and the example is displayed, the co-occurrence word search window 92 is provided in the upper menu bar. The "other window" button is clicked by operating the mouse cursor 90. As a result, as shown in FIG.
The next co-occurrence word search window 93 is opened.

When the co-occurrence word search window 93 is opened,
As in the case described above, the designated word “tacit knowledge” for searching the next co-occurrence word is input in the designated word input field 93a at the top of the co-occurrence word search window 93. . Next, in the co-occurrence word search window 93, as in the case described above, the "search" button provided on the menu bar at the upper part of the co-occurrence word search window 93 is clicked by operating the mouse cursor 90. The origin search processing is started. As a result, a word having a co-occurrence relationship with the word “implicit knowledge” specified here is searched, and the search result of the co-occurrence word search is the same as the co-occurrence word search window shown in FIG. It is displayed in the list item field 93b provided on the left side of 93. The "autonomous decentralized" and "synergetics" of the co-occurrence words displayed in the list item field 93b are underlined. This indicates that the word is displayed as a co-occurrence word in the word search window, and is displayed in response to the value indicating that the word has been searched.

Further, when referring to the example of each co-occurrence word displayed in the list item field 93b, as in the case described above, as shown in FIG. By selecting one of the words "synergetics" by an operation, this operation becomes an instruction input of a word for searching an example, and an example including the co-occurrence word is displayed. Further, subsequently, it becomes an instruction of a designated word for the next co-occurrence word search when the next co-occurrence word search window is opened, and the input operation of the designated word is performed.

That is, as shown in FIG. 13, by operating the mouse cursor 90, one word "synergetics" is selected from the co-occurrence words displayed in the list item field 93b of the co-occurrence word search window 93. When the designated operation is performed, in this case, the designated word input field 93
It is possible to search for example data that includes both the specified word “tacit knowledge” input in a and the specified word “synergetics” among the co-occurrence words displayed in the list item field 93b as co-occurrence words. Done, co-occurrence word search window 9
In the example list display field 93c on the right side of 3, the example of the search result is displayed.

Such a series of operations are as follows in the operation of the program processing module shown in FIG. That is, from the designation of the co-occurrence word searched in the first search window 60, the second search window 80 is opened, and the word designation input unit 81 is used to designate and input the word to be searched next. Thus, the word input from the word designation input unit 81 is searched for either the first co-occurrence degree or the second co-occurrence degree held by the co-occurrence relation holding unit 71, and the co-occurrence relation A word whose strength exceeds a predetermined threshold is output as a co-occurrence word of the input word. Then, the co-occurrence word display / instruction unit 83 displays the co-occurrence words searched by the co-occurrence word search unit 82 in a list, and selects a designated word of the co-occurrence word for which an example is searched from the displayed co-occurrence words. And instruct. Subsequently, the example data search unit 84 causes the co-occurrence word display / instruction unit 83.
For the co-occurrence word designated by
The operation is such that the example data storage unit 72 retrieves the example data including both the words input by 1, and the example data display unit 85 displays the example data of the search result.

Further, when the next co-occurrence word search window is opened and the co-occurrence word search is continuously performed, the co-occurrence word search window 9 is displayed as shown in FIG.
The "other window" button provided on the menu bar at the top of 3 is clicked by operating the mouse cursor 90. As a result, the next co-occurrence word search window 94 is opened as shown in FIG.

Also in this case, when the co-occurrence word search window 94 is opened, the designated word input field 94a at the top of the co-occurrence word search window 94 is opened as in the case described above.
In, the designated word “synergetics” when searching for the next co-occurrence word is input. When starting the search, in the co-occurrence word search window 94, as in the case described above, the "search" button provided on the menu bar above the co-occurrence word search window 94 is clicked by operating the mouse cursor 90. As a result, the process of the third stage co-occurrence word search is started. As a result, a word having a co-occurrence relationship with the word “synergetics” specified here is searched, and the search result of the co-occurrence word search is the same as in FIG.
6, the list item field 94b is provided on the left side of the next co-occurrence word search window 94.

Further, when referring to the example of each co-occurrence word displayed in the list item field 94b, as in the case described above, as shown in FIG. By selecting the one word "autopoiesis" by the operation, this operation becomes an instruction input of the word for the example search, and the example including both the co-occurrence words is searched and displayed. In addition, here, the designated word “autonomous decentralized” is further input in the designated word input field 94a, and these two designated words are used as co-occurrence words to search for an example including the selected word “autopoiesis”. I have to.

That is, as shown in FIG. 17, by operating the mouse cursor 90, one word "autopoiesis" is designated from the co-occurrence words displayed in the list item field 94b of the co-occurrence word search window 94. In this case, first, the designated word input field 94a
The specified word “synergetics” and the specified word “autonomous decentralized” entered in “A” and the specified word “autopoiesis” among the co-occurrence words displayed in the list item field 94b are co-occurrence words together. The example data including “” is searched, and the example of the search result is displayed in the example list display field 94c on the right side of the co-occurrence word search window 94.

When a plurality of sequential co-occurrence word searches are performed using the plurality of co-occurrence word search windows, the search process using the co-occurrence word dictionary is as described with reference to FIGS. , As a process of updating the value of the data F of the co-occurrence dictionary,
It is registered in the co-occurrence word dictionary of the co-occurrence word dictionary holding unit 71. As a result, based on the value of the data F, in the co-occurrence word chain search window 76, the co-occurrence word chain processing unit 73 performs processing to sequentially follow the search process starting from the word designated by the word designation unit 75. , In the co-occurrence word chain display portion 74,
The "co-occurrence word group sequentially displayed in each co-occurrence word search window" as the processing result is displayed as a display chain process.

An operation example when displaying such a co-occurrence word chain will be described. Referring to FIG. 18, there is shown an example of a screen in a state where a cooccurrence word chain search window 95 is searched for a cooccurrence word chain. When the co-occurrence word chain is displayed, after performing the co-occurrence word search operation a plurality of times,
As shown in, for example, the character string “autonomous distribution” of the word in the document is selected by operating the mouse cursor 90, and is opened by clicking the “co-occurrence” button on the menu bar at the top of the document editing window 91. Select the "Cooccurrence chain" menu from the pull-down menu. As a result, the process of sequentially searching the search process starting from the designated word “autonomous distribution” is performed, and the cooccurrence chain search window 95 is displayed.
Is opened, and the chain process of the co-occurrence words is displayed in the co-occurrence word chain search window 95. The user can see the chain process of the co-occurrence words displayed in the co-occurrence word chain search window 95 and use it for idea support.

As described above, according to the document creation support apparatus of this embodiment, the co-occurrence words are explicitly displayed during the document preparation, and they are searched in a chained manner so that the co-occurrence words are commonly used. Relevant knowledge about a document can be retrieved by searching for examples that include. Therefore, in the document being created,
Even if the knowledge about the description content is ambiguous, efficient document creation can be performed from the viewpoint of idea support.

[0075]

As described above, according to the document creation support apparatus of the present invention, when a word is designated, a list of words that co-occur with the word is displayed separately from the document editing apparatus. When one co-occurrence word is designated from the co-occurrence word group, a plurality of co-occurrence word display means and a plurality of example data display means for displaying an example including both the designated word and the designated co-occurrence word are prepared. The user is
While referring to the displayed examples, identify the co-occurrence words that are highly relevant to the document being created from the displayed co-occurrence word group,
From the co-occurrence word, a co-occurrence word related to the co-occurrence word can be further searched, and the example can be searched. By repeating such a search operation, a co-occurrence word that is highly related to the document being created can be searched in a chained manner, and even document creation support and idea creation support can be performed. Further, since the chain process of the co-occurrence words searched in a chain can be referred to, it is possible to create the document while confirming the chain of the co-occurrence words.

[Brief description of drawings]

FIG. 1 is a block diagram showing a basic configuration of a document creation support device according to a first embodiment of the present invention,

FIG. 2 is a diagram showing an example of a data structure of a co-occurrence word dictionary held in a co-occurrence relation holding unit.

FIG. 3 is a diagram showing an example of numerical data of a co-occurrence word dictionary before a co-occurrence word search process is performed,

FIG. 4 is a diagram showing an example of numerical data of a co-occurrence word dictionary after a co-occurrence word search process is performed,

FIG. 5 is a flowchart showing a processing flow of co-occurrence word search processing,

FIG. 6 is a flowchart showing a search processing flow of a co-occurrence word chain,

FIG. 7 is a block diagram showing a configuration of a document creation support apparatus according to a second embodiment of the present invention as a combination of respective program processing modules by a software processing mechanism,

FIG. 8 is a diagram showing a screen example of a document edit window,

FIG. 9 is a diagram showing a screen in which a first search window is opened from a document edit window,

FIG. 10 is a diagram showing a screen in a state where a co-occurrence word search is performed in the first search window,

FIG. 11 is a diagram showing a screen in a state in which an example is searched from a co-occurrence word in the first search window,

FIG. 12 is a diagram showing a screen in a state where a second search window for searching a co-occurrence word is opened following the first search window;

FIG. 13 is a diagram showing a screen in a state in which a co-occurrence word search is subsequently performed in the second search window and further an example search is performed;

FIG. 14 is a diagram showing a screen in which a third search window is opened from the second search window,

FIG. 15 is a diagram showing a screen in a state in which a third search window for a co-occurrence word search continuing from the second search window is opened,

FIG. 16 is a diagram showing a screen in a state where the co-occurrence word search is performed in the third search window similarly to FIG. 15,

FIG. 17 is a diagram showing a screen in a state in which an example is searched for from a co-occurrence word in the third search window,

FIG. 18 is a diagram showing a screen in a state in which a co-occurrence word chain search window is searched for a co-occurrence word chain.

[Explanation of symbols]

1 ... Co-occurrence relation holding unit, 2 ... Example data holding unit, 3 ... Co-occurrence word chain processing unit, 4 ... Co-occurrence word chain display unit. Also, 1
0 ... First search window, 11 ... First designated word input section, 12 ... First co-occurrence word display section, 13 ... First example data display section, 20 ... Second search window, 21 ... Second Second designated word input section, 22 ... second co-occurrence word display section, 23 ... first
Example data display unit, 25 ... Co-occurrence dictionary, 26 ... Cell,
27 ... Co-occurrence word dictionary (before search), 28 ... Co-occurrence word dictionary (after search), 60 ... First search window, 61 ... Word designation section, 62 ... Co-occurrence word search section, 63 ... Co-occurrence word table / Designation part, 6
4 ... Example data search section, 65 ... Example data display section, 70
... document editing window, 71 ... co-occurrence dictionary holding unit, 72
... Example data holding unit, 73 ... Co-occurrence word chain processing unit, 74 ...
Co-occurrence word chain display section, 75 ... Word designation section, 76 ... Co-occurrence word chain search window, 80 ... Second search window, 81
... word designation part, 82 ... co-occurrence word search part, 83 ... co-occurrence word table /
Designation section, 84 ... Example data search section, 85 ... Example data display section, 90 ... Mouse cursor, 91 ... Document edit window, 92 ... Co-occurrence word search window, 92a ... Specified word input field, 92b ... List item field, 9
2c ... Example list display field, 93 ... Co-occurrence word search window, 93a ... Specified word input field, 93b ...
List item field, 93c ... Example list display field, 94 ... Co-occurrence word search window, 94a ... Specified word input field, 94b ... List item field, 94c ... Example list display field, 95 ... Co-occurrence word chain search window.

Claims (3)

[Claims] 1. A co-occurrence relation holding means for holding the strength of a co-occurrence relation of a second word with respect to a first word, an example data holding means for holding example data, and an arbitrary word is designated and input. And a list of second words that have a co-occurrence relationship with the designated first word, corresponding to each designated word input means. An example in which each designated co-occurrence word display means and a designated co-occurrence word corresponding to each co-occurrence word display means are received and both the designated co-occurrence word and the designated word by the designated word input means are included. A document creation support device, comprising: each example data display means for searching and displaying a search result. 2. The document creation apparatus according to claim 1, further comprising obtaining the co-occurrence word displayed by each co-occurrence word display means from the co-occurrence relationship holding means, and following a sequential display chain process. A document creation support device comprising: a co-occurrence word chain processing means for storing; and a co-occurrence word chain display means for displaying the display chain process stored in the co-occurrence word chain processing means. 3. An example data holding unit for holding example data, and as a co-occurrence relation between words, a first co-occurrence degree of a strength of a co-occurrence relation of a second word with respect to a first word, A co-occurrence relation holding means for respectively holding a second co-occurrence degree of the strength of the co-occurrence relation of the two words, a word designation input means for designating a word, and the word designation input means. The first co-occurrence degree or the second co-occurrence degree held by the co-occurrence relation holding unit for the input word
A co-occurrence word search unit that searches for one of the co-occurrence degrees and outputs a word whose strength of co-occurrence relationship exceeds a predetermined threshold value as a co-occurrence word of the input word; A co-occurrence word display instructing unit that displays the output co-occurrence words in a list and selects and instruct a designated word of the co-occurrence word to be searched next from the displayed co-occurrence words, and the co-occurrence word display instructing unit. Example data search means for searching the example data holding means for example data containing both the instructed co-occurrence word and the word input by the word designation input means, and the example data searched by the example data search means are displayed. A document creation support apparatus comprising: an example data display unit.