JPH09212491A - Document processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To predict how much a user understand terms and expressions included in an existent document. SOLUTION: This processor is provided with knowledge state managing means 1a, 1b, and 1c by individual users. The knowledge state managing means 1a, 1b, and 1c manages how much the users understand arbitrary terms. A term extracting means 3 extracts a specific term 3a from an inputted document 2. An understanding degree analyzing means 4 gathers information regarding who much the users understand the terminal 3a extracted by the term extracting means 3 from the respective knowledge state managing means 1a, 1b, and 1c and calculates the statistics of the gathered information to generate prediction data. The prediction data are stored in a storage part 4a in the understanding degree analyzing means 4. The prediction data stored in the storage part 4a can be displayed on a display device at a request that a person who creates a document to be distributed to the users makes.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document processing apparatus that assists in creating a document to be provided to a user, and more particularly to a document processing apparatus that individually manages a user's degree of understanding of a predetermined term.

[0002]

2. Description of the Related Art Documents play a large role in human information transmission. It is the most popular means to preserve the results of human intellectual activity and to keep it in the form of documents for later use by other humans. Also, in daily communication, electronic documents such as electronic mail are often used these days.

However, in such information transmission using a document, miscommunication is caused due to the difference between the knowledge of the creator of the document and the knowledge of the user of the document, and the communication between the document creator and the user. However, there is a big problem that the mutual load on each other increases. For example, when a user of a document reads the document, it often happens that he / she does not fully understand the contents of the document without knowing the definition of technical terms and related information in the document. As a result, the value of the content of a document cannot be accurately determined, and important information is often overlooked. In addition, the content of the document may be misunderstood, and the subsequent actions may be disturbed. In order to prevent such miscommunication, the user must be the document creator or
It is necessary to make various inquiries to another person who has the same knowledge as that of the document creator, which is a heavy load for both the document creator and the user.

Therefore, the document creator assumes the knowledge level of the user of the document in order to prevent miscommunication and unnecessary load on the user of the document.
Select terms and add term definitions and related information to the document. Therefore, the less the knowledge shared between the creator of the document and the user, the more information has to be put in the document, which is a great burden on the document creator.

Further, when there are many users of the document, the knowledge possessed by the users is various, and the document creator needs to create a document that the user can understand for each user, or that there are many users. Assuming the level of knowledge you think, you will create a single document for that level. However, in the former case, the load of creating multiple documents to convey the same information is imposed on the creator. In the latter case, a document user who shares a lot of knowledge with the document creator has a redundant sentence with a lot of unnecessary information, and it takes more time than necessary to read it. On the contrary, the document cannot be fully understood by the document user who does not have the level of knowledge assumed by the document creator.

In order to solve such a problem, in the invention disclosed in Japanese Patent Laid-Open No. 7-93334, a computer simulates a known or unknown state of a word held by a user, and Among them, a step of automatically identifying a word that the user does not know, a step of automatically supplementing the unknown word with a synonym known by the user, and automatically displaying a document supplementing the synonym And a document processing method including the steps of:

Further, the above-mentioned publication discloses a means for simulating a known or unknown state of a word possessed by a user and automatically discriminating a word which is not known by the user among words in a document, A document processing apparatus is also disclosed, which comprises means for automatically supplementing unknown words with synonyms known to the user, and means for automatically displaying a document supplementing the synonyms.

With such a document processing method and document processing device, a synonym known to the user of the document is automatically added to a term unknown to the user of the document for each document user. Helps users understand and reduces the burden on the creator of the document.

[0009]

However, Japanese Patent Laid-Open No. 7-9 / 1995
In the invention disclosed in Japanese Patent No. 3334, it is helpless to let the user understand the terms and expressions expressing a relatively new concept without synonyms. Moreover, what actually becomes a problem when the user understands the document is the term or expression expressing a new concept. It is for the following reasons.

The first reason is that the new concept has an overwhelmingly small number of users who are familiar with the concept, as compared with the general concept where synonyms exist. The second reason is that in terms of valuable information that includes new information, terms and expressions that express new concepts often appear, and it has an important role in the document. There are many.

In order to avoid the above problem, the creator of a document should use terms and expressions that are easy to understand to represent a new concept to the extent that the intended user can understand the knowledge. Is common. However, in this case, unless the intended user is a person who is familiar with the creator of the document, it is difficult to assume the knowledge that the user has. Also, when the number of users is large, it is very difficult to create a document in consideration of the tendency of the knowledge of all users. As a result, the document creator creates a document that is difficult for the user to understand, causes miscommunication, or puts an extra load on both the creator and the user for mutual understanding after transmitting the document. Either it will take more time than is necessary to create a polite document, and moreover, it will cause many users to spend more than necessary to understand the document.

As described above, when a document including a relatively new concept is distributed, it is difficult to compose a sentence in terms that the user can easily understand, because the degree of knowledge of the user is unknown. there were.

The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a document processing apparatus that predicts the degree of understanding of a user regarding terms and expressions included in an existing document.

[0014]

FIG. 1 is a block diagram showing the principle of the present invention. In order to solve the above problems, according to the present invention, in a document processing apparatus that assists in creating a document to be provided to a user, a knowledge state management unit that is provided for each of a plurality of users and manages the user's degree of understanding of terms. 1a, 1b, 1c
... and information indicating the degree of comprehension of each user with respect to an arbitrary term is collected from the knowledge state management means 1a, 1b, 1c, ... And the user comprehension degree prediction means 4 for collecting statistics of the collected information. There is provided a document processing device having:

According to this configuration, the degree of understanding of the terms of a plurality of users is managed by the knowledge status management means 1a, 1b, 1c, ... Corresponding to each user. The user comprehension degree prediction unit 4 provides information indicating the degree of comprehension of each user with respect to an arbitrary term to the knowledge state management unit 1a of each user,
1b, 1c ... and collect statistics of the collected information. As a result, the document creator can predict the user's understanding level from the statistical result calculated by the user understanding level prediction means 4.

[0016]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing the principle of the present invention. The document processing apparatus of the present invention is provided with knowledge state management means 1a, 1b, 1c, ... For each of a plurality of users. The knowledge state management means 1a, 1b, 1c, ...
It manages the level of understanding of arbitrary terms by the corresponding users. The term extracting means 3 extracts a predetermined term 3a from the input document 2. The comprehension level analysis unit 4 provides information about the comprehension level of the user with respect to the term 3a extracted by the term extraction unit 3, to each of the knowledge state management units 1a, 1b, 1c, ...
・ Collect statistics from the collected information and create forecast data. The prediction data is stored in the storage unit 4a in the comprehension degree analysis unit 4. The prediction data stored in the storage unit 4a can be displayed on the display device in response to a request from a person who creates a document to be distributed to the user (document creator).

With such a configuration, for example, when the document 2 created by the document creator includes a relatively new term "VOD (Video On Demand)", the term extracting means 3 extracts the "VOD" from the document 2. 3 ”is extracted. At this time, the knowledge state management means 1a has a known degree of "VOD" of "20", the knowledge state management means 1b has a known degree of "VOD" of "30", and the knowledge state management means 1c has "VOD". The known degree of is assumed to be “0”. It should be noted that the higher the known value, the better the term is known.

The comprehension degree predicting means 4 collects the degree of knowledge of "VOD" from each knowledge state managing means 1a, 1b, 1c, .... Then, the prediction data is created based on the collected information of a large number of known degrees, and the prediction data is stored in the storage unit 4.
Stored in a. In this example, the degree of understanding of the user is divided into three levels of "highly known person", "known person", and "unknown person". Then, the known degree for "VOD" is collected for 50 persons, and 10 are highly known persons, 25 are known persons, and 1 are unknown persons.
It is assumed that there were five people. Then, the ratio of the number of people included in each understanding level to the total number of people is calculated, and the number of highly-known people is 2
0%, 50% known, 30% unknown. This is stored in the storage unit 4a as prediction data of the degree of understanding of the user with respect to the term "VOD". If the same process is performed on a large number of terms in the document 2, a prediction result of the degree of understanding of the user for the entire document 2 is created.

By displaying the prediction data in the storage unit 4a on the screen of the display device, the document creator can understand how much the user understands the contents of the document he is creating.
You can know before distributing the document to the user. As a result, it becomes possible to create a document that satisfies the necessary and sufficient details, politeness, and simplicity for the user.

FIG. 2 is a block diagram showing a concrete example of a document processing apparatus for carrying out the present invention. This is because, when the user inputs the first document to be understood, the document processing apparatus automatically converts the first document into the second document that the user can understand.
This is an example in which the function of the present invention is added.

In FIG. 2, a related information management section 40 is a section for accumulating / managing / retrieving information (hereinafter referred to as related information) necessary for a user to understand the contents of a document.
Registration of related information managed by the related information management unit 40 /
Deletion / update is managed by the related information creator. In addition,
The document creator 34 can also serve as the related information creator.

Further, knowledge state management units 11 and 21, additional information extraction units 12 and 22, document creation units 13 and 23, and user input / output i / f 14 and 24 are provided for the users 15 and 25, respectively. . The knowledge state management units 11 and 21 are units that accumulate / manage information representing the knowledge states of the corresponding users 15 and 25.

The additional information extracting units 12 and 22 include technical term extracting units 12a and 22a and a condition narrowing unit 12b, respectively.
22b and 22b. Technical term extractor 12a, 22a
Analyzes a document and extracts terminology from the document.
Then, a technical term list is created based on the extracted technical terms. The condition narrowing units 12b and 22b determine, for the terms extracted by the technical term extracting units 12a and 22a, information related to the terms and the related species of information necessary for the user. In the additional information extracting units 12 and 22 having such a configuration,
The related information necessary for the user to understand the first document is taken out from the related information management unit 40, and the list of the first document and the taken out related information is sent to the document creation units 13 and 23. As the first document, the document being edited by the document editing unit 32 is sent when the destination of the document is designated by the document creator 34.

The document creation units 13 and 23 create a second document from the documents and related information sent from the additional information extraction units 12 and 22. The users 15 and 25 can read the second document through the input / output i / f. Also, at this time, the knowledge status management units 11 and 22 update the knowledge status of the users 15 and 25 according to the content of the first document.

On the other hand, on the side of the document creator 34, the user comprehension degree prediction unit 31 connected to the knowledge state management units 11 and 21,
Document editing unit 3 connected to the additional information extracting units 12 and 22
2, and a document creator input / output i / f 33 that manages the input / output of the document creator 34. Document editing unit 32
Is a part where the document creator edits the document. Also, from the document creator of the document to the document creator input / output i / f33
When the destination of the document is designated through the additional information extracting unit 12 or 22 for each user who has designated the document currently being edited.
Send to

The user comprehension degree prediction unit 31 has a technical term extraction unit 31a and a term comprehension degree prediction unit 31b. The technical term extraction unit 31a extracts a predetermined technical term from the document being edited by the document editing unit. The term comprehension degree prediction unit 31b
Based on the knowledge states of the users 15 and 25, the tendency of how much the specified users 15 and 25 have the understanding of the specified terms and documents is predicted.

Details of the processing functions executed by the above configuration will be described below. In the document processing apparatus of FIG. 2, the user comprehension degree prediction unit 31 is the main part of the present invention. However, in order to make accurate prediction in the user comprehension degree prediction unit 31, understanding of the user's term It is important that the degree is always simulated by the knowledge state manager.
Therefore, the function for simulating the knowledge state of each user 15, 25 will be described.

In the related information management unit 40, the related information is body data indicating the information body, a data type attribute indicating what type of body data the data is, and a character string indicating what the information is about. A key attribute that stores
It is managed as data consisting of a related species attribute indicating what kind of relationship is related to the target specified by the key part. Data of a data type such as a character string or various figures is stored as the main body data. The data types that can be used in this example are string type, table type, geometric type, mathematical type, image type, and composite type thereof. The type to which the body data belongs is indicated by the data type attribute.

The related species attribute includes "definition information" indicating that the related information is the definition of the object indicated by the key attribute, and "definition information" indicating that the related information is the details of the object indicated by the key attribute. Any one of three values of “detailed information” and “additional information” indicating that the related information is associated with the object indicated by the key attribute is stored.

The related information management section 40 has a search function. This retrieval function retrieves the related information having the specified key attribute value and related species attribute value when the retrieval of the related information is specified along with the value of the key attribute, the value of the related species attribute, and the related information. If the information exists, it will be retrieved.

FIG. 3 is a diagram showing an example of related information managed by the related information management unit. Related information is key attribute 4
1, related species attribute 42, data type attribute 43, body data 4
It is managed by 4 items. In this figure, six types of related information 40a-40f are shown. Note that all the related information shown in this example is a "character string type" data type attribute.

There are two types of related information whose key attribute is "related information", the related type attribute of the first related information 40a is "definition information", and the related type attribute of the second related information 40b is "detailed information". Is. There are three types of related information whose key attribute is "knowledge state", the related type attribute of the first related information 40c is "definition information", the related type attribute of the second related information 40d is "detailed information", and the third type. The related species attribute of the related information 40e is “incidental information”. There is one type of related information whose key attribute is “additional information”, and the related species attribute of the related information 40f is “definition information”.

Next, the knowledge state management unit 1 shown in the example of FIG.
1, 21 will be described. Knowledge state management units 11 and 21
In, known records that represent information that the user knows are managed. Each known record consists of a category of information that the user knows, and a degree of knownness (an integer value of 0 to 30) that indicates the degree to which the user knows the information of the category of the information.

The category of information in this example is a key portion holding a character string indicating what the information relates to, and a related seed portion indicating what kind of information the object designated by the key portion has. And consists of. In the related seed part, "definition information" indicating that the information is the definition of the target indicated by the key attribute, and "detailed information" indicating that the information is the details of the target indicated by the key attribute, Any one of three types of values of “incidental information” indicating that the information is associated with the object indicated by the key attribute is stored.

When the instruction to refer to the knowledge state is input together with the character string, the knowledge state management units 11 and 21 output a known record having the character string as a key portion. Since there can be multiple known records that have the same character string in the key part,
If there are multiple known records corresponding to the designated character string, all of them are output.

Further, the knowledge status management units 11 and 21 are provided with a character string representing a key, a value of a related species, and an update mode ("known",
If the update of the knowledge state is specified together with "unknown"), the known record being managed is registered / updated. When the update mode is known, if there is a known record having the character string representing the key and the value of the related species, the known degree thereof is set to 30. If the corresponding known record does not exist, a known record having the character string representing the key and the value of the related species is created, and the known degree thereof is set to 30.

When the update mode is unknown, if there is a known record having the character string representing the key and the value of the related species, the known degree is set to 0. If the corresponding known record does not exist, a known record having the character string representing the key and the value of the related species is created, and the known degree thereof is set to 0.

Further, the knowledge status management units 11 and 21 calculate the knownness values of all the known records managed by the knowledge status management units 24 and 24.
It includes a function that decrements by -1 with the passage of time. This is to simulate human characteristics that are forgotten over time. However, the degree of knownness of 0 does not decrease any further.

FIG. 4 is a diagram showing an example of known records managed by the knowledge status management unit 11. Known records are
The key section 111, the related seed section 112, and the known degree 113 are managed. In this figure, five types of known records 11a to 11e are shown.

There are two types of known records whose key part is "related information", the related seed part of the first known record 11a is "definition information", the degree of knownness is "20", and the second known record 11b. The related species part of is "detailed information", and the degree of knownness is "10". There are three types of known records whose key part is “knowledge state”, the related species part of the first known record 11c is “definition information”, the degree of known is “25”, and the related species of the second known record 11d. The part is "detailed information", the degree of knownness is "20", the related species attribute of the third known record 11e is "incidental information", and the degree of knownness is "0".

Next, the additional information extracting units 12 and 22 will be described. FIG. 5 is a flowchart showing a procedure for extracting related information and updating the knowledge state when the first document is input to the additional information extraction unit.

First, in the technical term extraction units 12a and 22a,
The first document sent from the document editing unit 32 is analyzed, and a noun having a professional meaning in the document (hereinafter referred to as a technical term)
To make a list (step S1). Therefore, the technical term extraction units 12a and 22a have a morphological analysis unit that divides sentences in a document and a noun dictionary of general terms. The technical term extracting units 12a and 22a examine the nouns extracted as a result of the morphological analysis in the noun dictionary of general terms, and determine that the terms that are not there are technical terms.

Next, it is checked whether the technical term list is empty (step S2). If it is empty, the list of the first document and the related information is sent to the document creating unit (step S3), and the process ends. If the technical term list is not empty, one technical term is extracted from the list (step S4). At this time, the extracted technical term is deleted from the list.

Next, with respect to the acquired terms, the condition narrowing unit determines the related species of information necessary for the user to understand the first document (step S5). As a result of the determination of the related species, it is determined whether or not the user has the necessary related information, and if there is the necessary related information, the related species is determined (step S6). As a result of the determination of the related species, if it is determined that the user does not need information of any related species, the process returns to the check of the technical term list (step S2).

When it is determined that the user needs the definition information, the processing for the case where the definition information is necessary, which will be described later, is performed (step S7), and the process returns to the check of the technical term list (step S2). When it is determined that the user needs the detailed information, a process for the detailed information described later is performed (step S8), and the process returns to the check of the technical term list (step S2). When it is determined that the user needs the supplementary information, the processing for the case where the supplementary information is required, which will be described later, is performed (step S9), and the technical term list is checked (step S2).

FIG. 6 is a flowchart showing the procedure for determining the related species (step S5) performed by the condition narrowing unit. In the related species determination process, first, an instruction to refer to the knowledge state is sent to the knowledge state management unit together with the extracted technical term, and a known record having the technical term in the key section is extracted (step S11). It is determined whether or not there is a known record (step S12), and if there is no known record in the knowledge state management unit, all information regarding the technical term is known to the user, and information on any related species is required. It is determined that there is not (step S19), and the process ends.

When the known record can be extracted, the known degree of the record of which the value of the related seed part is "definition information" is checked (step S13). Here, when the known degree is 15 or less, it is determined that the user needs the definition information (step S14), and the process ends.

When the degree of knownness of the record whose related seed part value is "definition information" is greater than 15, it is determined that the definition information of the technical term is known to the user, and the relationship among the known records extracted is determined. The degree of knowledge of the record whose seed value is "detailed information" is checked (step S15). Here, when the known degree is 15 or less, it is determined that the user needs detailed information (step S16), and the process ends.

If the degree of knownness of the record whose related seed part value is "detailed information" is greater than 15, it is determined that the detailed information of the technical term is known to the user, and the relationship among the known records extracted is determined. The degree of knownness of the record in which the value of the seed portion is “additional information” is checked (step S17). Here, when the known degree is 15 or less, it is determined that the user needs the supplementary information (step S18), and the process ends.

When the degree of knownness of the record in which the value of the related species part is "incidental information" is greater than 15, it is determined that all the information related to the technical term is known to the user and the information of any related species is unnecessary. Then (step S19), the process ends.

FIG. 7 is a flow chart showing the procedure of the processing (step S7) when definition information is required. First, using the technical term as a key attribute value, the related information management unit is instructed to take out the related information whose related seed attribute value is "definition information" (step S21). Here, it is determined whether or not the definition information exists (step S22). If the extraction of the definition information fails, the user remains unknown about the technical term, and the knowledge state is updated with the related term as “definition information” and the update mode as “unknown”, since the technical term remains unknown. Is issued to the knowledge state management unit.
As a result, the degree of knownness of the definition information is updated to 0 (step S23).

When the extraction of the related information has succeeded, the related information is added to the related information list (step S24), and then the related term is defined as "definition information" and the update mode is set as "known" with the technical term as a key. The knowledge state management unit is instructed to update the knowledge state. That is, since the definition information is presented to the user, it becomes “known”. As a result, the degree of knownness of the definition information is updated to 30 (step S25).

The processing when the detailed information is required and the incidental information is required is the same procedure as the flowchart of FIG. However, the "definition information" portion becomes "detailed information" and "incidental information", respectively.

Next, the document creating sections 13 and 23 will be described. The document creation units 13 and 23 receive the first document and the related information list described later from the additional information extraction unit, and add the related information to an appropriate position of the first document to generate the second document. To do.

FIG. 8 is a flow chart showing the procedure for creating the second document in the document creating section. First, it is checked whether or not the related information list is empty (step S31).
If the related information list is empty, the second document is complete and the process ends. If the related info list is not empty,
Relevant information is retrieved from the list (step S3)
2). At this time, the retrieved related information is deleted from the list.

Next, in the first document, the position where the same character string as the value of the key attribute of the related information appears (more precisely, the position of the tail end of the matched character string) is checked (step S3).
3). When there are multiple corresponding character strings, the position that appears first is adopted.

Next, the data type attribute of the related information is checked (step S34). If the value of the data type attribute is not the character string type, the body data of the related information is used as the annotation of the position of the character string (step S38).

If the value of the data type attribute is a character string, the value of the related species attribute of the related information is checked (step S3).
5). When the value of the related species attribute is "definition information", the contents of the main body data are put in parentheses, placed at the position detected in step S33 (step S37), and the process returns to the related information list check (step S31).

If the value of the related species attribute is "detailed information", the length of the body data is checked (step S3).
6). If the length of the body data is 20 characters or less,
The contents of the body data are put in parentheses, placed at the position detected in step S33 (step S37), and the process returns to the check of the related information list (step S31). When the length of the body data is more than 20 characters, the body data is used as the annotation of the position (step S38), and the process returns to the check of the related information list. If the value of the related species attribute is "incidental information", the body data is used as the annotation of the position detected in step S33 (step S38), and the process returns to the related information list check (step S31).

The annotation process of this example (step S3)
The process 8) is a process of placing a note symbol at a position where an annotation is placed and placing related information together with the note symbol at the end of the document. Here, how the various types of information change when the first document is input to the additional information extraction unit 12 of the document processing apparatus as described above will be described. FIG. 9 is a diagram showing the first document 50 sent from the document editing unit. When such a first document is input, a technical term list is created by the technical term list creation processing. FIG. 10 is a diagram showing the created technical term list 51. And this technical term list 5
Since 1 is not empty, the first term "related information" is retrieved. Here, it is assumed that the related information management unit 40 manages the related information shown in FIG. 3 and the knowledge state management unit 11 manages the knowledge state shown in FIG.

Next, a known record having "related information" in the key part is taken out. In this case, known records 11a and 11b having "related information" in the key part are extracted from the known records shown in FIG. FIG. 11 is a diagram showing two retrieved known records 11a and 11b. Of the retrieved known records, the known degree of the known record 11a whose related seed part value is "definition information" is greater than 15, and the known record 1 whose related seed part value is "detailed information".
Since the known degree of 1b is 15 or less, it is determined that the user needs “detailed information” of the term “related information”.

Then, the related information management unit is instructed to take out the related information whose key attribute value is "related information" and related seed attribute value is "detailed information". Then, the related information whose key attribute value is “related information” and related species attribute value is “detailed information” is extracted from the related information shown in FIG. FIG.
FIG. 8 is a diagram showing the extracted related information 40b.

After adding the related information 40b shown in FIG. 12 to the related information list, an instruction to update the knowledge state with "related information" as a key, related species as "detailed information" and update mode as "known" is given. This is performed for the knowledge status management unit. Then, the process returns to the check of the technical term list.

Such processing is performed for all the technical terms in the technical term list. As a result, the related information list is completed. FIG. 13 shows the created related information list 52.
FIG. In FIG. 13, the related information 40b whose key attribute is "related information" and related species attribute is "detailed information"
And the related information 40e whose key attribute is "knowledge state" and related species attribute is "incidental information" are listed. At this time, the knowledge status managed by the knowledge status management unit is updated. FIG. 14 is a diagram showing the updated knowledge state. FIG. 14 is updated from the knowledge state of FIG. What is updated is a known record 11f in which the key part is "related information" and the related seed part is "detailed information", and a known record 11 in which the key part is "knowledge state" and the related seed part is "incidental information".
g. The known record 11f has a known degree of 10 to 3
It has been changed to 0. The known degree of the known record 11g is changed from 0 to 30. Thus, the knowledge state of the user 15 is simulated by updating the value of the known record at any time.

When the first document 50 is as shown in FIG. 9 and the related information list 52 is as shown in FIG. 13, the second document 53 as shown in FIG. Created with (The underlined part is the first
Is the part added to the document).

The user comprehension degree predicting unit 31 predicts the degree of comprehension of the user with respect to an arbitrary technical term by using the knowledge states of the users 15 and 25 simulated as described above. The user understanding level prediction method will be described in detail below.

The document editing section 32 is a section where the document creator edits the document. Further, when the destination of the document is designated by the document creator of the document through the document creator input / output i / f 33, the currently edited document is sent to the additional information extracting units 12 and 22 for each designated user. Further, when a user is specified and an instruction to predict the user comprehension degree is given from the document creator 34 through the document creator input / output i / f 33 during editing of the document, the document editing unit 32 is currently editing. The user comprehension degree prediction unit 3 is provided with a list of documents and designated users.
Send to 1.

The user comprehension degree predicting unit 31 has a function of predicting the situation of the user's understanding of a specific term or a specific document. The document creator uses the document comprehension input / output i / f33 as the comprehension degree prediction result by the user comprehension degree prediction unit
The document being edited can be revised so that it is easier for users to understand.

Further, the technical term extracting section 31a in the user comprehension degree predicting section 31 analyzes the document and extracts the technical term contained in the document. The term comprehension degree prediction unit 31b predicts the comprehension degree of the user of the term extracted by the technical term extraction unit.

In this embodiment, when a document and a list of users are sent from the document editing unit together with an instruction for predicting comprehension, how many users are in the technical terms in the document and how much the technical terms are. Predict what you are doing. The prediction is
The degree of comprehension of the user group designated for each of the definition information, detailed information, and supplementary information for the term is predicted. The degree of comprehension is predicted by calculating the ratios of highly known persons with a known degree of 21 or more, known persons with a known degree of 11 or more and 20 or less or no corresponding known record, unknown persons with a known degree of 10 or less, respectively. Be seen.

Therefore, the user comprehension degree prediction unit 31 includes a storage unit, in which the number of highly known persons in each of the technical term storage area (character string), definition information for each term, detailed information, and incidental information. (Integer value of 0 or more), number of known people (0
The above integer value), the number of unknown persons (integer value of 0 or more), advanced known person ratio (integer value of 0 to 100), known person ratio (integer value of 0 to 100), unknown person ratio (from 0 to 1
(Integer value up to 00).

FIG. 16 is a flow chart showing a procedure of predicting the degree of comprehension with respect to a specific term performed by the user comprehension degree predicting unit. When the understanding degree instruction is input to the user comprehension degree prediction unit together with the document and the user list, the technical term extraction unit 31a first creates a list of technical terms in the document (step S41). Therefore, the technical term extraction unit 31a has a morphological analysis unit that divides a sentence in a document and a noun dictionary of general terms. Technical term extractor 31a
Examines the nouns extracted as a result of morphological analysis in the noun dictionary of general terms, and judges those that are not there as technical terms.

Next, it is checked whether the technical term list is empty (step S42). If the list is empty, the process ends. If the list is not empty, the term is retrieved from the list and stored in the term storage area (step S4).
3). At this time, the retrieved term is deleted from the list.

Next, the term comprehension level predicting unit 31b predicts the comprehension level of the user designated by the user list with respect to the retrieved term (step S44), and then checks the technical term list (step S44). Return to S42).

FIG. 17 is a flow chart showing the procedure of predicting the degree of comprehension of a term performed by the term comprehension degree predicting unit 31b. First, the degree of understanding of definition information for a term is predicted (step S51). Next, the degree of understanding of detailed information about the term is predicted (step S5).
2). Finally, the degree of comprehension of the supplementary information for the term is predicted, and the process ends (step S53).

FIG. 18 is a diagram showing the first half of a flow chart showing the procedure of predicting the degree of understanding of definition information for a term performed by the term comprehension degree predicting unit. First, 0 is stored in each storage area of the number of highly known persons, the number of known persons, and the number of unknown persons of the definition information of the term (step S61). Next, the first user in the user list is selected (step S62). From the knowledge status management unit of the selected user, a known record of definition information for the term being processed is retrieved (step S6).
3). Here, it is judged whether or not the known record can be extracted (step S64), and if the known record cannot be extracted, it is assumed that the selected user knows the definition information of the term being processed, and the definition information of the term. The value stored in the storage area for the known number of is incremented by 1 (step S68).

When the known record can be taken out in step S64, the known degree of the known record is checked (step S65). When the degree of knowledge is 21 or more, it is determined that the user is familiar with the definition information of the term, and the value stored in the storage area of the number of highly known persons of the definition information of the term is incremented by 1 (step S66). If the degree of knowledge is 11 or more and 20 or less, it is determined that the user knows the definition information of the term, and the value stored in the storage area of the number of known persons of the definition information of the term is incremented by 1 (step S6
8). If the degree of knowledge is 10 or less, the user determines that he / she does not know the definition information of the term, and increments the value stored in the storage area of the number of unknown persons of the definition information of the term by 1 (step S67).

Next, it is checked whether or not there is the next user in the user list (step S69). If there is a user, the user is selected (step S70), and the known record is taken out (step S63). If there is no next user, the process of creating prediction data is started.

FIG. 19 is a diagram showing the latter half of the flowchart showing the procedure of predicting the understanding degree of the definition information for the term performed by the term understanding degree predicting unit. This is a prediction data creation process that is executed when it is determined in step S69 in FIG. 18 that there is no next user. If there is no next user, the result of “the number of highly known persons of definition information / the number of users × 100” is stored in the storage area of the degree of highly known persons of definition information (step S71). Next, the result of “the number of known persons of definition information / the number of users × 100” is stored in the storage area of the known person ratio of definition information (step S72).
Finally, the result of “the number of unknown persons in the definition information / the number of users × 100” is stored in the storage area of the unknown person ratio of the definition information (step S73), and the process ends.

The process of predicting the degree of understanding of the detailed information of the term and the process of predicting the degree of understanding of the additional information of the term are almost the same as the processes of FIGS. 18 and 19. The difference is that in the process of predicting the degree of understanding of detailed information of a term, "definition information" in the procedure of FIGS.
Is "detailed information", and in the process of predicting the degree of comprehension of additional information of a term, "definition information" in the procedure of FIGS. 18 and 19 is "additional information".

As described above, the degree of understanding of the user can be predicted. That is, the user comprehension degree prediction unit 31
Is input from the document creator input / output i / f33 together with the list of users who are the target of the understanding level prediction, the term for checking the understanding level, and the instruction for predicting the understanding level of the term. Calculates how much each type of information is understood by the users listed in the specified user list. And for the term specified at this time,
The term comprehension degree prediction unit 31b predicts comprehension degree according to the procedure shown in FIG.

FIG. 20 is a diagram showing an example of the inside of the storage unit of the user comprehension degree prediction unit. FIG. 20 shows the results of the prediction of the understanding level for the four technical terms. FIG.
Then, as a storage area of related information about each technical term,
Definition information 110, detailed information 120, and incidental information 13
A storage area of 0 is provided. The storage area of each related information is divided into items of highly known persons, known persons, and unknown persons. Further, in each item, a value of the number of persons corresponding to an arbitrary term and a ratio of the number of persons corresponding to the total number of persons can be stored.

In this example, the prediction data 101 to 10 for 20 persons relating to the four types of technical terms "related information", "knowledge state", "additional information", and "related information management section".
4 is stored.

Prediction data 101 regarding "related information"
Is 10 people (50%) who know the definition information
There are 7 known people (35%) and 3 unknown people (15%). There are 7 highly known persons (35%), 8 known persons (40%), and 5 unknown persons (25%) for detailed information. There are 0 (0%) highly known persons, 7 (35%) known persons, and 13 (65%) unknown persons with respect to the incidental information.

Prediction data 102 regarding "knowledge state"
Is 17 people (85%) who know advanced definition information,
There are 3 known people (15%) and 0 unknown people (0%). There are 10 people (50%) with advanced knowledge for detailed information,
There are 8 known people (40%) and 2 unknown people (10%). There are 5 highly known persons (25%), 4 known persons (20%), and 11 unknown persons (55%) with respect to the incidental information.

Prediction data 103 regarding "additional information"
Indicates that there are 8 (40%) highly known persons, 9 (45%) known persons, and 3 (15%) unknown persons with respect to the definition information. There are 0 (0%) highly known persons, 14 (70%) known persons, and 6 (60%) unknown persons for the detailed information. There are 0 (0%) highly known persons, 8 (40%) known persons, and 12 (60%) unknown persons with respect to the incidental information.

Prediction data 1 for "related information management section"
No. 04 has 0 persons (0%) with high level knowledge of definition information,
There are 5 known people (25%) and 15 unknown people (75%). There are 0 (0%) highly known persons, 3 (15%) known persons, and 17 (85%) unknown persons for detailed information. There are 0 (0%) highly known persons, 0 (0%) known persons, and 20 (100%) unknown persons for the incidental information.

It can be seen from FIG. 20 that the user almost knows the definition information and detailed information of the term “knowledge state”. On the other hand, each information of the term "related information management section" is almost unknown to the user, and the description of the term "related information management section" is included in the text of the document or the term "related information management section" is added to the related information management section. It can be seen that the user cannot understand the information unless the information about "copy" is registered.

[0089]

As described above, in the first invention,
Since the user comprehension degree prediction means collects information indicating the degree of comprehension of the user with respect to an arbitrary term from a plurality of knowledge state management means and collects the collected information, the document creator can To what extent the user can understand the term can be predicted from the statistical result of the user comprehension degree prediction means.

Further, in the second invention, when a document is input to the term extracting means 3, a predetermined term is extracted from the document, and information indicating the degree of understanding of each user regarding the extracted term is collected. Since statistics are taken, the user comprehension degree predicting means determines how much the user can understand the specific term appearing in the document being created or being created by the document creator. It can be comprehensively predicted from the statistical results.

Further, in the third invention, the user is divided into a plurality of groups based on the degree of understanding of the term, and each group is divided.
By statistically classifying the users by calculating the percentage of users included in the group, how the creators of the documents disperse their understanding of the target terms Can be easily recognized.

[Brief description of drawings]

FIG. 1 is a principle configuration diagram of the present invention.

FIG. 2 is a configuration diagram showing a specific example of a document processing apparatus for carrying out the present invention.

FIG. 3 is a diagram showing an example of related information managed by a related information management unit.

FIG. 4 is a diagram showing an example of known records managed by a knowledge state management unit.

FIG. 5 is a flowchart showing a procedure of extracting related information and updating a knowledge state when the first document is input to the additional information extracting unit.

FIG. 6 is a flowchart showing a procedure of determining a related species performed by a condition narrowing unit.

FIG. 7 is a flowchart showing a procedure of processing when definition information is required.

FIG. 8 is a flowchart showing a procedure of creating a second document in the document creating unit.

FIG. 9 is a diagram showing a first document sent from a document editing unit.

FIG. 10 is a diagram showing a created technical term list.

FIG. 11 is a diagram showing two known records extracted.

FIG. 12 is a diagram showing extracted related information.

FIG. 13 is a diagram showing a created related information list.

FIG. 14 is a diagram showing an updated knowledge state.

FIG. 15 is a diagram showing a created second document.

FIG. 16 is a flowchart showing a procedure of predicting a degree of comprehension of a document performed by a user comprehension degree predicting unit.

FIG. 17 is a flowchart showing a procedure of predicting an understanding level for a term performed by a term understanding level prediction unit.

FIG. 18 is a diagram showing the first half of a flow chart showing a procedure of predicting an understanding level of definition information for a term performed by a term understanding level prediction unit.

FIG. 19 is a diagram showing the latter half of the flowchart showing the procedure of predicting the understanding level of definition information for a term performed by the term understanding level prediction unit.

FIG. 20 is a diagram illustrating an example of an inside of a storage unit included in a user comprehension degree prediction unit.

[Explanation of symbols]

 1a, 1b, 1c Knowledge status management means 2 Document 3 Term extraction means 3a Term 4 User comprehension degree prediction means 4a Storage section 11,21 Knowledge status management section 12,22 Additional information extraction section 13,23 Document creation section 14,24 User input / output i / f 15,25 User 31 User comprehension level prediction unit 31a Technical term extraction unit 31b Term comprehension level prediction unit 32 Document editing unit 33 Document creator input / output i / f 34 Document creator 40 Related information Management Department

Claims (3)

[Claims] 1. A document processing device for assisting in the creation of a document to be provided to a user, the knowledge state management means being provided for each of a plurality of users, for managing the user's degree of understanding of the term, and for any term. A document processing device comprising: a user comprehension level prediction unit that collects information indicating the comprehension level of each user from the knowledge state management unit and obtains statistics of the collected information. 2. The method further comprises a term extracting means for extracting a predetermined term from the input document, and the user comprehension degree predicting means collects information on the degree of comprehension of the user with respect to the term extracted by the term extracting means. The document processing apparatus according to claim 1, wherein: 3. The user comprehension degree predicting means divides the user into a plurality of groups based on the degree of comprehension of the term, and calculates the ratio of the users included in each group. The document processing device according to claim 1.