JP2006146621A - Information management device and method, and information management program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To easily recognize the state of knowledge of each user. <P>SOLUTION: An information management device includes input/output data storage means 31, 32 for storing input/output data inputted and outputted by users and input/output condition data indicative of the input/output conditions of the input/output data; a feature value extraction means 22 for extracting feature values based on the predetermined standard of the input/output data; and an individual knowledge data creation means 23 for creating, based on the feature value of the input/output data and the input/output condition data, individual knowledge data indicative of the feature of the input/output data for each of the users who inputted or outputted the input/output data. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an information management apparatus, and more particularly to an information management apparatus that characterizes input / output data by a user and manages it for each user. The present invention also relates to an information management method and an information management program.

  In recent years, with a large amount of information being supplied, individuals can easily search for or receive desired information (information search and information distribution), and search for experts with the desired knowledge. (Expert search), etc. are increasing.

  On the other hand, personal information is often input / output on a computer as electronic information such as electronic documents, e-mails, and electronic images, and personal input / output information is stored as personal interests and expertise. It is becoming possible.

  Here, the input / output information refers to, for example, input information that has been copied or browsed and output information that has been created or edited. Conventional techniques have been proposed in which such input / output information of an individual is used for search and distribution. These prior arts include the following. For example, (1) Knowledge management service that accumulates input / output information of each individual and later uses it for information search and expert search, and (2) Stores input / output of personal information, and was in that individual Information distribution personalization and recommendation service (for example, Patent Document 1) enabling information provision.

JP 2001-175676 A

  However, as a first problem of the above-described conventional example, there arises a problem that old data is buried in a database and it is difficult to search. In other words, individual interests and expertise usually change over time, and individuals have multiple interests and expertise over time, but conventional methods use the entire period from the past to the present. Although it is not so noticeable, personal interests and expertise extracted from information intensively input and output during a certain past period, such as themes handled only in a certain past period, are buried in the whole, It becomes difficult to extract.

  The above situation will be described with reference to FIG. A solid line (horizontal line) in FIG. 21 indicates that a certain individual P was engaged in an activity related to a specific interest or expertise in a specific period on time t (horizontal axis). For example, the three solid lines (horizontal lines) located at the top in FIG. 21 indicate that the individual P is involved in the interests and expertise a during the period from time t1 to t2, t3 to t5, and t7 to t9. It means that he was doing. Similarly, two lower solid lines (horizontal lines) in FIG. 21 indicate that the individual P is engaged in activities related to interests and expertise b in the period from time t2 to t4, and is interested in the period from time t6 to t8. And activities related to expertise c.

  In the conventional example described above, since one individual's interests and expertise are extracted from the entire input / output information, only the individual interests and expertise a can be extracted and concentrated on a certain period in the past. There arises a problem that the handled interests and expertise b and c cannot be extracted. Similarly, in the conventional example, it is difficult to extract personal interests and expertise from input / output information that is old in time. Information that is far away from the time is difficult to extract. As a result, although a that has been dealt with a lot throughout and c that has recently been handled intensively can be extracted, there is a problem that interests and expertise b that are older than c and that have been handled intensively at a certain time in the past cannot be extracted. .

  Further, as a second problem of the conventional example, there arises a problem that it becomes difficult to accurately extract personal interests and expertise. This is because the individual's interests and expertise are extracted from the entire input / output of personal information without distinguishing whether the individual has just seen or deeply understood. To elaborate further, we do not distinguish between whether the information was just copied / viewed or whether it was treated as a high level of interest or expertise so that it could be created / edited, and features were extracted from both of them. Because.

  Therefore, the object of the present invention is to make it possible to easily recognize the knowledge state of each user by performing data management in accordance with the personal input / output status, thereby facilitating all information retrieval. And Specifically, in response to the first problem described above, by extracting the interests and specialized knowledge of each individual period using time information, it looks like a theme handled only at a certain time in the past An object is to make it possible to search for personal interests and expertise extracted from information intensively input and output during a past period. Another object of the present invention is to accurately extract the interests and expertise of an individual, whether they have only seen it or understood it deeply in response to the second problem.

Therefore, an information management apparatus according to one aspect of the present invention is
Input / output data storage means for storing input / output data input / output by the user and input / output status data representing the input / output status of the input / output data;
Feature value extraction means for extracting feature values based on predetermined criteria of input / output data;
Personal knowledge data generating means for generating personal knowledge data representing the characteristics of the input / output data for each user who has input / output the input / output data based on the feature values of the input / output data and the input / output status data;
It is characterized by having.

  As a result, personal knowledge data representing the characteristics of the input / output data is generated according to the input / output status of each user, so that the knowledge state of each user can be easily recognized, and accurate information search can be realized.

  The feature value extracting means is characterized in that the feature value is extracted based on at least one of the appearance frequency, appearance position, and emphasis information of each term included in the input / output data. Thereby, since personal knowledge data for each user is generated according to the frequency of input / output terms, the knowledge state of each user can be recognized more clearly.

  In addition, the input / output status data is data representing the time when the input / output data was input / output. At this time, the personal knowledge data generating means generates personal knowledge data based on a plurality of input / output data that are input / output within a predetermined period. Thereby, personal knowledge data can be generated for each input / output time of the input / output data, and the knowledge state of each user for each time can be recognized. Therefore, it is possible to search without being conscious of new and old information.

  The input / output status data is input / output distinction data representing input / output distinction of the input / output data. At this time, the personal knowledge data generating means generates the personal knowledge data by weighting the input / output data according to the input / output distinction data. Further, at this time, the personal knowledge data generating means is characterized in that the output data of the input / output data is weighted higher than the input data. Further, the input / output status data is reference status data representing a reference status of other data in the input / output data. At this time, the personal knowledge data generating means generates the personal knowledge data by weighting the input / output data according to the reference situation data. In particular, the weighting is characterized in that weighting is performed on data serving as a reference destination of predetermined data among input / output data. As a result, the characteristics of the personal knowledge data for each user are generated according to the input / output distinction of the input / output data or the reference state, so that it can be generated more reflecting the individual knowledge state. Accurate information search is possible.

  In addition, a similarity calculation unit that calculates a similarity between feature values of a plurality of input / output data is provided, and the personal knowledge data generation unit is configured to store personal knowledge data for a plurality of input / output data having a similarity equal to or greater than a predetermined value. It is characterized by generating. As a result, personal knowledge data for input / output data having similar characteristics can be generated, and personal knowledge data can be generated in a group of similar data, and more personal characteristics can be reflected.

  Further, in addition to the above configuration, the information storage means for storing the personal knowledge data generated by the personal knowledge data generating means is provided. And a search means for receiving search request data including data corresponding to the input / output data from another user and searching the personal knowledge data stored in the information storage means based on the search request data. It is a feature. At this time, the search means is characterized in that it searches for other personal knowledge data similar to the searched personal knowledge data based on a predetermined criterion. Further, in addition to the above configuration, a search information output means for outputting personal knowledge data searched by the search means is provided. Further, the present invention is characterized in that it includes search information output means for reading out and outputting input / output data corresponding to the personal knowledge data searched by the search means from the input / output data storage means. Further, the present invention is characterized in that search information output means for outputting user specifying information for specifying a user corresponding to the personal knowledge data searched by the search means is provided. Accordingly, information can be searched based on the characteristics of input / output data for each individual, and more accurate information can be searched.

In addition, an information management method according to another aspect of the present invention includes:
An information management method executed by a computer connected to input / output data storage means for storing input / output data input / output by a user and input / output status data representing the input / output status of the input / output data. ,
A feature value extraction step of extracting feature values of the input / output data based on a predetermined criterion;
Based on the characteristic value of the input / output data and the input / output status data stored in the input / output data storage means, personal knowledge data representing the characteristics of the input / output data for each user who has input / output the input / output data is generated. Personal knowledge data generation process,
It is characterized by having.

  The feature value extraction step is characterized in that feature values are extracted based on at least one of the appearance frequency, appearance position, and emphasis information of each term included in the input / output data.

The input / output status data is data representing the time when the input / output data was input / output.
The personal knowledge data generating step is characterized in that personal knowledge data is generated based on a plurality of input / output data input / output within a predetermined period.

The input / output status data is input / output distinction data indicating the input / output distinction of the input / output data.
The personal knowledge data generation step is characterized in that the personal knowledge data is generated by weighting the input / output data according to the input / output distinction data.

  Furthermore, the personal knowledge data generation step calculates similarity between feature values of a plurality of input / output data, and generates personal knowledge data for a plurality of input / output data having the similarity equal to or higher than a predetermined value. It is characterized by that.

An information management program according to another aspect of the present invention is
To a computer connected to input / output data storage means for storing input / output data input / output by the user and input / output status data representing the input / output status of the input / output data,
Feature value extraction means for extracting feature values of the input / output data based on a predetermined criterion;
Based on the characteristic value of the input / output data and the input / output status data stored in the input / output data storage means, personal knowledge data representing the characteristics of the input / output data for each user who has input / output the input / output data is generated. Personal knowledge data generation means;
It adopts the configuration of an information management program for realizing the above.

  An information management program for realizing, on the computer, feature value extraction means for extracting a feature value based on at least one of the appearance frequency, appearance position, and emphasis information of each term included in the input / output data But there is.

Also on the computer,
Personal knowledge data for generating personal knowledge data based on a plurality of input / output data input / output within a predetermined period when the input / output status data is data representing the input / output time of the input / output data. It is also an information management program for realizing the generation means.

Also on the computer,
Personal knowledge data generating means for generating personal knowledge data by weighting the input / output data according to the input / output distinction data when the input / output status data is input / output distinction data representing the input / output distinction of the input / output data It is also an information management program for realizing.

In addition,
Similarity calculation means for calculating the similarity between feature values of a plurality of input / output data;
It is also an information management program for realizing personal knowledge data generating means for generating personal knowledge data for a plurality of input / output data having a similarity equal to or greater than a predetermined value.

  Even the information management method and the information management program having the above-described configuration operate in the same manner as the information management apparatus described above, and thus the above-described object can be achieved.

  Since the present invention is configured and functions as described above, according to this, the history of input / output data by the user is generated as personal knowledge data representing the characteristics of the input / output data according to the input / output situation for each user. Therefore, it is possible to easily recognize the knowledge status of each user, and it is possible to easily and accurately search information by managing such personal knowledge data. .

  The present invention is an information management apparatus for accumulating and managing input / output data, which are personal interests and expertise input / output by a user, so that it can be easily retrieved later. For this reason, the present invention is characterized in that personal knowledge data representing the characteristics of input / output data is generated according to the input / output situation and managed.

  In the following, in the first embodiment, an input / output time is given as an example of an input / output state, and an information management apparatus that manages by dividing input / output data for each period during which input / output is performed will be described. In the second embodiment, an input / output distinction is given as another example of the input / output situation, and an information management apparatus that generates and manages features according to input / output differences and the presence or absence of other references will be described. In the third embodiment, a case where the above period and input / output distinction are combined will be described.

  A first embodiment of the present invention will be described with reference to FIGS. FIG. 1 is a block diagram showing the overall configuration of the present invention. FIG. 2 is a functional block diagram showing the configuration of the information management apparatus. 3 to 5 are explanatory diagrams showing the state of information processing in the information management apparatus. 6 to 12 are flowcharts showing the operation of the information management apparatus.

[Constitution]
The present invention is an information management apparatus 2 that manages data input and output by a user, and is configured by an information processing apparatus such as a server computer. Then, it is connected via a network N to user terminals 11, 12, and 13 which are computers operated by users A, B, and C that input and output data such as data registration and browsing with respect to the information management apparatus 2. Has been. Further, a user terminal 14 which is a computer operated by a user D who makes a data search request to the information management apparatus 2 is also connected via the network N.

  Note that the information management apparatus 2 is not necessarily configured by a single computer, and may be configured by a plurality of computers. Data input / output and search are not limited to being performed by the user terminals 11, 12, 13, and 14, but may be performed by a user operating the information management apparatus 2 itself. Further, the “user” is not limited to being an individual, and may refer to an organization or group composed of a plurality of subjects. Hereinafter, a specific configuration of the information management apparatus 2 will be described.

  The information management device 2 includes a calculation unit 20 that performs calculation processing of information and a storage unit 30 that stores information. The calculation unit 20 incorporates an information management program, which is shown in FIG. Thus, the data input processing unit 21, the feature value extraction processing unit 22, and the personal knowledge generation processing unit 23 are constructed, and input / output data by the user terminals 11, 12, and 13 are based on the features. Has the function to manage. Further, the calculation unit 20 includes the above-described program, so that a search request reception processing unit 24, a search processing unit 25, and a search result output processing unit 26 are constructed. Has a function of performing a search process in response to the search request. In response to this, the storage unit 30 includes an input / output data storage unit 31, a time data storage unit 32, a period setting data storage unit 33, and a personal knowledge data storage unit 34. Hereinafter, the processing units 21 to 26 and the storage units 31 to 34 will be described in detail.

  The data input / output processing unit 21 receives input / output data input / output by the user to / from the information management apparatus 2 via the user terminals 11, 12, and 13 and time data representing the time input / output by the user. The input / output data storage unit 31 and the time data storage unit 32 (input / output data storage unit) in the storage unit 30 have a function of storing each. Here, the time data represents, for example, the time when the user accesses the input / output data or the time when the input / output data is updated. The input / output data is data that has been stored in the storage unit of the information management apparatus 2 in advance. Therefore, in such a case, the time data refers to the time when the user saved the input / output data in the storage unit 30. The data input / output by the user may be data input / output to / from a storage unit of another computer. In that case, the information management device 2 monitors such an input / output state and operates to store the input / output data and the time data. The input / output data is data representing the interests and expertise of the user, and is extracted, stored, and shared as personal knowledge, as will be described later.

  The state when the input / output data and the time data are stored as described above will be described with reference to FIGS. For example, when the user A is interested in the programming language “C ++” and reads a document to acquire knowledge, the data input / output by the user A via the user terminal 11 is stored as needed. The data is stored in the 30 input / output data storage unit 31. At the same time, the time data (time information) when the user A inputs / outputs the input / output data is stored in the time data storage unit 32 of the storage unit 30. Then, as shown in FIG. 3, the input / output data by the user A is stored in association with the ID of the input / output data and the input / output time. The input / output data includes “C ++”, “library”, “class”, “variable” and the like shown in FIG. 4 as described later.

  Next, the feature value extraction processing unit 22 (feature value extraction means) will be described. The feature value extraction processing unit 22 operates to calculate the feature value of the input / output data. At this time, the feature value calculation rule (reference) is set in advance and incorporated in the processing unit 22. Hereinafter, an example of feature value extraction will be described. Here, the input / output data is assumed to be text information, and the feature value is extracted based on the appearance frequency of each term included in the text information.

The feature value extraction processing unit 22 reads data input / output by the user A stored in the input / output data storage unit 31 of the storage unit 30 and analyzes the data. At this time, the analysis unit may be arbitrary and may not be an information unit such as one file. For example, when the input / output data Aj (j> = 1) for a certain period is text information, the analysis method performs weighting using a TF / IDF technique, which is a well-known technique, and features words (terms). This is done by extracting More specifically, words are extracted from each input / output data Aj by morphological analysis processing, and the frequency of appearance of content words such as nouns, verbs, and adjectives is counted from the extracted words, and each content word is input. The TF / IDF value of the content word is calculated using the appearance frequency in the output data. Then, the top N feature words of the TF / IDF value and the TF / IDF value are used as the feature amount of the input / output data. For the TF / IDF method, refer to the document “Sparck”.
Jones, Karen. (1972). A statistical interpretation of term specificity and its
Journal of Documentation 28: 11-21 "and so on, and is well known, so detailed explanation is omitted.

  In the above description, the case where the input / output data is text information has been described. However, the input data is not limited thereto, and may be, for example, text information, annotated image information or audio information. . The feature value extraction method is not limited to the above method, and the feature value may be extracted based on the appearance position of each term included in the input / output data, emphasis information such as font information and prosodic information. And you may extract a feature value combining these.

  Here, an example of the result of the feature value extraction process 22 is shown in FIGS. In FIG. 4, morphological analysis is performed on input / output data whose input / output data ID is A3, content words such as nouns, verbs, and adjectives are extracted, and compared with the frequency of content words in other input / output information, TF / IDF In this example, the top four feature words “C ++”, “library”, “class”, “variable” and their TF / IDF values are extracted as feature values. FIG. 5A shows an example in which input / output data A1 to Aj at each time are displayed as feature values.

  Next, the personal knowledge generation processing unit 23 (personal knowledge data generation means) will be described. The personal knowledge generation processing unit 23 obtains personal knowledge data representing the characteristics of the input / output data for each user who has input / output the input / output data, based on the time data which is the input / output status data, from the feature values of the input / output data. Operates to generate. In this embodiment, personal knowledge data representing the characteristics of individual user input / output data is created by calculating the characteristics of a plurality of input / output data that are particularly close in input / output time.

  Further, the personal knowledge generation processing unit 23 will be specifically described. First, the period setting data stored in the period setting data storage unit 33 is read. The period setting data is, for example, data that sets a plurality of input / output data that are close in time as data within one period, or data that sets input / output data within a predetermined time as one period. For example, it is data serving as a reference for setting a predetermined period. The period is set based on the period setting data. In this embodiment, the period A1 to A3 shown in FIG. 5A is set as the specific period At1. Then, personal knowledge data representing further features is generated from the feature values of the input / output data within the period. For example, the TF / IDF values of the top N feature words of each input / output data are used as feature word vectors, and the sum is generated. And let the produced | generated data be personal knowledge data, as shown in FIG.5 (b). Thereafter, personal knowledge data for another period is also created, and personal knowledge data is also created for input / output data by the users B and C. The personal knowledge generation processing unit 23 stores the personal knowledge data generated as described above in the personal knowledge data storage unit 34 (information storage unit).

  Here, the method of generating personal identification data by the personal knowledge generation processing unit 23 is not limited to the above-described method. For example, the personal identification data may be generated by adding the feature words that are the feature values of the input / output data and their TF / IDF values within a predetermined time and taking the average. Further, the setting of the period is not limited to that described above, and the period may be dynamic rather than a fixed time.

  Next, the search request receiving unit 24 will be described. The search request receiving unit 24 operates to receive a search request for data managed by the information management device 2 from the user terminal 14 and notify the search processing unit 25 of the request. At this time, it is assumed that the user D operates the user terminal 14 that makes a search request. Then, it is assumed that the user D transmits search request data “C ++ or its library” and makes a search request.

  Upon receiving notification of the search request data from the search request receiving unit 24, the search processing unit 25 (search means) analyzes the content of the search request data and searches for at least one of the stored personal knowledge data. As a specific search process, for example, a value of “1.0” is assigned to “C ++” and “library” which are extracted content words through a morphological analysis process. At the same time, a value of “0” is assigned to words that are included in the search target and are not included in the search request. As a result, for example, a feature word vector for a search request such as (C ++, library, class, variable) = (1.0, 1.0, 0, 0) is created. Then, the similarity between the word feature vector of the search request data and one or more personal knowledge data is calculated. Although the calculation method of the similarity may be arbitrary, for example, when the calculation is performed using the space vector method, the similarity between the above-described personal knowledge data At1 and the search request data becomes “0.91”. The degree of similarity between the search requests is calculated based on the personal knowledge data for each period of all users A, B, and C, and the personal knowledge data having the highest similarity to the search request data is identified from the result. To do.

  Then, the specified personal knowledge data itself may be extracted from the personal knowledge data storage unit 34, and other personal knowledge data similar to this may be specified and extracted. For example, the second highest personal knowledge data may be extracted by referring to the similarity. In addition, input / output data corresponding to the identified personal knowledge data may be extracted from the input / output data storage unit 31. Furthermore, user specifying information for specifying a user corresponding to the specified personal knowledge data may be extracted. That is, in the above-described example, information for identifying the user A (for example, the personal name of the user A) is extracted. Such a search may be performed when a person search request is made from the user terminal 14 at the time of the search request.

  The extracted information is passed to the search result output processing unit 26 (search result output unit), and is transmitted from the search result output processing unit 26 to the user terminal 14 that has made the search request.

[Operation]
Next, the operation of the information management apparatus 2 described above will be described with reference to FIGS. FIG. 6 is a flowchart showing information storage processing by the information management apparatus for input / output data by the user terminals 11, 12, and 13 of the users A, B, and C, and FIGS. It is a flowchart which shows. FIG. 9 is a flowchart showing search processing by the information management apparatus in response to a search request from the user terminal 14 of the user D, and FIGS. 10 to 11 are flowcharts showing detailed operations of a part thereof. In particular, the user A is interested in the programming language C ++, reads a document to acquire knowledge, creates a memo, and the user D inputs and outputs such knowledge data. Suppose you want to know about "C ++ and its library" and search for the personal knowledge of others. That is, the description will be given with reference to FIGS. 3 to 5 described above.

  First, the data input / output processing unit 21 of the information management device 2 stores the input / output data input / output at the user terminal 11 of the user A in the input / output data storage unit 31 (step S1). At the same time, the data input / output processing unit 21 of the information management apparatus 2 stores the time data when the user A inputs / outputs the input / output data in the time data storage unit 32 (step S2). Then, as shown in FIG. 3, the ID of the input / output data from the user terminal 11 by the user A and the time data for accessing the input / output data are stored.

  Subsequently, the feature value extraction processing unit 22 of the information management device 2 reads the input / output data stored in the storage unit 30 and extracts the feature value (step S3, feature value extraction step). The feature value extraction process will be described in detail with reference to FIG. When the input / output data is text information, the feature value extraction processing unit 22 uses a known technique of TF / IDF for the input / output data Aj (j> = 1) for a predetermined period. Weight. Specifically, as shown in FIG. 4, morphological analysis processing is performed on the input / output data A3 to extract content words such as nouns, verbs, and adjectives (step S11). Then, the frequency of appearance of content words such as nouns, verbs, and adjectives is counted from the extracted words (step S12), and using the frequency of appearance of each content word in each input / output data, TF / An IDF value is calculated (step S13). Subsequently, feature words “C ++”, “library”, “class”, and “variable” are extracted from the top four TF / IDF values (step S14), the TF / IDF values are extracted, and the feature words and The TF / IDF value is set as a feature value (step S15).

  Subsequently, the personal knowledge generation processing unit 23 uses the feature value of the input / output data obtained by the feature value extraction processing unit 22 and the time data stored in the time data storage unit 32 of the storage unit 30. Representative characteristics between input / output data with high temporal proximity are generated as personal knowledge data representing the interests and expertise of the individual during that period (step S4, personal knowledge data generation step). This operation will be further described in detail with reference to FIG.

  The personal knowledge generation processing unit 23 first reads the period setting data from the period setting data storage unit 33 (step S21), and sets a period during which input / output data for generating personal knowledge data is input / output based on the period setting data. (Step S22). In this case, as shown in FIG. 5A, a period t1 is set, and the features of the input / output data A1 to A3 are the targets for generating personal knowledge data. Then, the TF / IDF values of the top four feature words of each input / output data are used as feature word vectors, and their sum is created and used as personal knowledge data (step S23). Specifically, when the sum of the word feature vectors in the period t1 is At1, At1 = A1 + A2 + A3 = (0.96,0.84,0) + (0,0.92,0,064) + (0.93,0.86,0.74, 0.55) = (1.89, 2.62, 0.74, 1.19). Similarly to the above, personal knowledge data is created for each period tk (k> = 1).

  Thereafter, the generated personal knowledge data is stored in the personal knowledge data storage unit 34 (step S5). Then, not only the personal knowledge data of the user A but also the personal knowledge data for each of the users B, C, etc. are created and stored in the personal knowledge data storage unit 34 in the same manner.

  Next, a search processing operation by the information management apparatus 2 when a search request is made from the user terminal 14 of the user D will be described with reference to FIG. First, when a search request is made from the user terminal 14, the search request reception processing unit 24 receives the search request (step S41) and notifies the search processing unit 25 of it. At this time, consider a case where user D transmits a search request “C ++ or its library”.

Subsequently, the search processing unit 25 analyzes the content of the search request (step S42). For example, the search request data is subjected to a morphological analysis process, and a value of “1.0” is assigned to the extracted content words “C ++” and “library”. At the same time, a value of “0” is assigned to words that are included in the search target and are not included in the search request. As a result, for example, (C ++, library, class, variable) = (1.0,
A feature word vector for a search request such as 1.0, 0, 0) is created.

  Based on this, the personal knowledge data storage unit is searched (step S43). Specifically, as shown in FIG. 10, the similarity between the word feature vector of the search request data and the personal knowledge data is calculated. For example, when the calculation is performed using the space vector method, the similarity between the personal knowledge data At1 shown in FIG. 5A and the search request is “0.91”. The degree of similarity with the search request data is calculated for the personal knowledge data for each period such as user A, user B, and user C. From this result, the personal knowledge data having the highest similarity with the search request is specified (step S52). For example, when it is determined that the personal knowledge data At1 is the personal data having the highest similarity, the personal knowledge data At1 is extracted as a search result (step S44). The detailed operation of step S44 is also shown in FIG. 11A (step S61). And the extracted personal knowledge data At1 is output with respect to the user terminal 14 of the user D (step S45). Note that the data to be output as the search result may be extracted and output not only the personal knowledge data At1 but also personal knowledge data having the second and third highest similarity to the search request data. Furthermore, not only the personal knowledge data but also the input / output data that is the basis of the data is extracted from the input / output data storage unit 31 (see step S71 in FIG. 11B) and output to the user terminal 14 of the user D. May be.

  When the request from the user D who made the search request is not only the personal knowledge data but also a request to search for “detailed person” for “C ++ and its library”, as described above, the personal knowledge data After specifying At1, the user A who inputs / outputs the input / output data that is the basis of the personal knowledge data is specified (see step S81 in FIG. 11C), and the search result is “A that is an individual name”. To the user terminal 14 of the user D.

  In this way, according to the information management device 2 described above, the information input / output intensively in a certain past period is retrieved by extracting the interests and expertise of each individual period using the time information. It is an expert who has personal interests and expertise extracted from information intensively input and output during a certain past period, such as themes handled only in a certain past period. You can search for other individuals and their information. Similarly, it is also possible to search what interests and expertise other individuals who have current matching interests and expertise have in the past.

  Furthermore, it is now possible to extract interests and expertise for each individual period using time information, and to search for changes in interests and expertise from the past to the present of the individual, so that the individual who is the searcher wants When searching for information, it is possible to search for desired information from information held by a person closer to the interests of the past and the present and the transition of expertise, so that information can be reused it can.

[Modification]
Next, a modified example of the information management apparatus 2 configured as described above will be described with reference to FIG. In this modification, the basic configuration is the same as that of the information management device 2 described above, but the method of creating personal knowledge data at the time of information storage processing is different. That is, the personal knowledge generation processing unit 23 has the following functions, and the operation in step S4 in FIG. 6 is as shown in the flowchart in FIG.

[Constitution]
First, the personal knowledge generation processing unit 23 in this modification has a similarity calculation function (similarity calculation means) for calculating the similarity between feature values of a plurality of input / output data. And based on this calculated similarity, it has the function of producing | generating personal knowledge data using the feature value of several input / output data whose said similarity is more than predetermined value. In this embodiment, since the personal knowledge data is generated based on the feature values of the input / output data within the set period, as described above, the degree of similarity is high and the input / output The personal knowledge data is generated from the feature values of the input / output data having high time proximity.

[Operation]
A specific processing operation will be described with reference to the flowchart of FIG. First, when the feature value of each input / output data is calculated, the similarity between the feature values is calculated (step S31). At this time, the method for calculating the degree of similarity may be arbitrary, but the degree of similarity is determined based on, for example, overlapping of sets of feature words. In addition, using the well-known Vector Space Model (vector space method) etc., the feature of the information is expressed as a word matrix vector (feature word vector) based on the word data that appears in the text information. The degree of similarity between two pieces of information can be obtained from the distance between word vectors, the inner product, and the like. The vector space method is described in the document "Salton,
G. (1989) Automatic Text Processing. Addison-Wesley Publishing Company. Another known technology, Latent
Using the Semantic Space (latent semantic space method) or the like, it is also possible to approximate the created feature word vector by low-order approximation by singular value decomposition and reduce the vector dimension. The latent semantic space method is described in the document “Deerwester,
S. et. Al. (1990): Indexing by latent semantic analysis.Journal of the American
"Society for Information Science, 41 (7), 391-407."

  Then, using the calculated similarity, the feature values of the input / output data having a high similarity to each other are extracted (step S32), and the feature values between the input / output data having a high temporal proximity are obtained for the individual. Knowledge data is generated (step S35). That is, as described above, the period setting data is read out, a predetermined period is set (step S34), and the personal knowledge data representing the feature of only the feature value having high similarity among the input / output data existing in the period. Is generated.

  Next, a second embodiment of the present invention will be described with reference to FIGS. FIG. 13 is a functional block diagram showing the configuration of the information management apparatus. FIG. 14 to FIG. 15 are explanatory diagrams showing the state of information processing in the information management apparatus. 16 to 17 are flowcharts showing the operation of the information management apparatus.

[Constitution]
The information processing apparatus 102 in the present embodiment basically has the same configuration as that in the first embodiment described above. Therefore, the information management apparatus 102 is configured by an information processing apparatus such as one or a plurality of server computers, and is a computer operated by users A, B, and C that input and output data such as data registration and browsing. Are connected via a network N to user terminals 11, 12, and 13. Further, a user terminal 14 which is a computer operated by a user D who makes a data search request to the information management apparatus 102 is also connected via the network N.

  The information management apparatus 102 includes a calculation unit 120 that performs calculation processing of information, and a storage unit 130 that stores information. The information management program is incorporated in the calculation unit 120, so that FIG. As shown in FIG. 4, a data input processing unit 121, a feature value extraction processing unit 122, and a personal knowledge generation processing unit 123 are constructed, and input / output data by the user terminals 11, 12, and 13 are used as characteristics thereof. Based on the management function. In the present embodiment, in particular, since the method of generating personal knowledge data of input / output data is different from that in the first embodiment, the functions of the processing units 121, 122, and 123 are different. Accordingly, the storage unit 130 includes an input / output data storage unit 131, an input / output classification data storage unit 132, a weighting setting data storage unit 133, and a personal knowledge data storage unit 134. . These will be described in detail below.

  In addition, the calculation unit 120 incorporates the above-described program, so that a search request reception processing unit 124, a search processing unit 125, and a search result output processing unit 126 are constructed. However, since the functions of the processing units 124, 125, and 126 are almost the same as those in the first embodiment, the description thereof is omitted.

  The data input / output processing unit 121 inputs / outputs data that is input / output to / from the information management apparatus 102 by the user via the user terminals 11, 12, and 13 and data indicating the input / output status of the input / output data. It has a function of receiving input / output distinction data representing distinction and storing them in the input / output data storage unit 131 and the input / output distinction data storage unit 132 in the storage unit 130, respectively. The data input / output processing unit 121 further has a function of receiving reference status data representing the reference status of other data in the input / output data and storing it in the input / output segmented data storage unit 132.

  Here, each data mentioned above is explained in full detail. The input / output data is personal input / output data stored in the information management apparatus 102 in advance and capable of distinguishing input / output such as e-mail transmission mail and reception mail. The input / output distinction data is data indicating whether the input / output data is input data or output data. For example, the sent mail is provided with distinction data “1” indicating output data. The received mail is stored with the distinction data “0” indicating the input data. A storage example at that time is shown in FIG. This figure is an example in which the input / output data ID of user A, the input / output distinction data, and the reference relationship are recorded. First, an ID is assigned to the input / output data and stored together with the contents of the data. Further, if the input / output data user A is output data, “1” is assigned, and if the input data is input data, “0” is assigned. Store as shown.

  The reference status data is data indicating a reference relationship between input and output data. The reference relationship indicates that it can be obtained, for example, by specifying the input data in which the content of the description is written from the description such as the reference document specified in the output data or the description enclosed in quotation marks. . Then, for example, as shown in FIG. 14, when the output data A3 is created by referring to the other input data A2 (see the arrow in FIG. 14), the reference relationship is created. Is assigned to the data of A3 as “A3-> A2” and stored.

  The feature value extraction processing unit 122 operates in the same manner as described in the first embodiment. For example, the feature value extraction processing unit 122 extracts words from each input / output data Aj by morpheme analysis processing, and extracts nouns from the extracted words. The frequency of appearance of content words such as verbs and adjectives is counted, and the TF / IDF value of the content word is calculated using the frequency of appearance of the content word in each input / output data. As a result, as shown in FIG. 4 described in the first embodiment, the top N feature words and TF / IDF values are extracted as feature values.

  Next, the personal knowledge generation processing unit 123 (personal knowledge data generation means) will be described. The personal knowledge generation processing unit 123 operates to generate personal knowledge data by weighting the feature values of the input / output data according to the input / output distinction data. At this time, for example, the weighting for the output data is higher than that for the input data. Further, the input / output data is also weighted according to the reference status data. For example, a predetermined weight is added to the input data being referred to.

  Specific processing of personal knowledge data creation by the personal knowledge generation processing unit 123 will be described with reference to FIGS. First, FIG. 15A shows an example of input / output data. As shown in FIG. 14, A1 and A2 are input data, and A3 is output data. A3 refers to A2. In such a case, an arbitrary weight, for example, w = 0.8 (0 <= w <= 1) is added to the feature value of the output data A3 among the feature values of the input / output data, and the input data A1, A2 A weight of 0.2 (1-w) is added to the feature value. Further, for example, a weight of 0.5 is added to the input data A2 that is the reference destination of A3. Therefore, a weight of (1−w) + 0.5 = 0.7 is added to A2. Then, the feature value of each input / output data to which the weight is added is used as a feature word vector, and the sum Aw of the word feature vectors in each input / output data is taken. Then, Aw = (0.94, 1.50, 0.59, 0.89), which is generated as personal knowledge data representing typical characteristics of input / output data by the user A. Such a result more strongly reflects the characteristics of the information handled as output data by the user A.

  The personal knowledge data generated as described above is stored in the personal knowledge data storage unit 134. Similarly, personal knowledge data is generated and stored for input / output data of other users.

  Here, the above-described weighting value is an example, and is not limited to the above value. In the above-described method for calculating the personal knowledge data, the example in which the weight is added to the feature value of the input / output data based on the reference relationship has been described. However, the weight based on the reference relationship is not necessarily limited thereto. It may not be added. Conversely, only the weight based on the reference relationship is added, and the weight based on the input / output distinction may not be added.

  When there is a search request from another user for the personal knowledge data generated and stored as described above, the information management apparatus 102 outputs the search request receiving unit 124, the search processing unit 125, and the search result output. A search process is executed by the action of the processing unit 126. However, the function of each of the processing units 124, 125, 126 is almost the same as that of the first embodiment, and the description thereof is omitted.

[Operation]
Next, the operation of the information management apparatus 102 described above will be described with reference to FIGS. FIG. 16 is a flowchart showing information accumulation processing by the information management apparatus for input / output data by the user terminals 11, 12, and 13 of the users A, B, and C, and FIG. 17 is a flowchart showing a part of detailed operation thereof. It is. Similar to the first embodiment, the operation of the information accumulation process when the user A is interested in the programming language C ++ and reads a document or creates a memo to acquire knowledge will be described.

  First, the data input / output processing unit 121 of the information management apparatus 102 stores the input / output data input / output on the user terminal 11 of the user A in the input / output data storage unit 131 (step S101). At the same time, the data input / output processing unit 121 of the information management apparatus 102 stores the input / output distinction data, which is the distinction between acquisition and output of data input / output by the user A, in the input / output distinction data storage unit 132 ( Step S102). At this time, the input / output data reference relationship is also stored in the input / output distinction data storage unit 132. An example stored in this way is shown in FIG. As described above, in this storage example, the input / output data ID is assigned to the data input / output by the user A, and the data content and ID are stored. Further, if the data accessed or stored by the user A is output data, “1” is assigned, and if the data is input data, “0” is given and stored. Further, a reference relationship that is created by referring to the input data A2 when creating certain output data A3 is assigned to A3 and saved as "A3-> A2". The contents of input / output data are, for example, data that can be distinguished from input / output such as e-mail transmission mail and reception mail.

  Subsequently, the feature value extraction processing unit 122 of the information management apparatus 102 reads the stored input / output data, and extracts the feature value as in the case of the first embodiment (step S103).

  Subsequently, the personal knowledge generation processing unit 123 uses the feature value of the input / output data obtained by the feature value extraction processing unit 122 and the input / output distinction data stored in the input / output distinction data storage unit 132. Representative characteristics representing personal interests and expertise corresponding to the input / output situation are generated as personal knowledge data (step S104). This operation will be further described in detail with reference to FIG.

  The personal knowledge generation processing unit 123 first reads the weight setting data from the weight setting data storage unit 133 (step S111), and also reads the input / output distinction data and the reference situation data from the input / output distinction data storage unit 132 (step S111). S112). Based on these, the feature value of each input / output data is weighted (step S113). In this embodiment, the output data is set to have a high weight, and the reference destination data is also set to have a high weight. Specifically, weights are added as shown in FIG. 15A, and the sum of word feature vectors of each input / output data A1, A2, A3 is calculated, so that the individual shown as Aw in FIG. Knowledge data is generated (step S114).

  Thereafter, the generated personal knowledge data is stored in the personal knowledge data storage unit 134 (step S105). Then, not only the personal knowledge data of the user A but also the personal knowledge data is created for the users B, C, etc., and stored in the personal knowledge data storage unit 134.

  Note that the search processing operation by the information management apparatus 102 when a search request is made later from the user terminal 14 of the user D is the same as that in the first embodiment, and the description thereof is omitted.

  In this way, since the personal knowledge data is created by weighting the knowledge created based on the output data of the personal knowledge using the input / output distinction data, the user outputs the output data Since it reflects the characteristics of the information handled as a more strongly, it is possible to accurately extract personal interests and expertise during subsequent searches. This is because the input data that the individual browses and copies is simply viewed by the individual, and the degree of association with the individual's interests and expertise may be low, whereas the individual This is because output data such as creation and editing is not realized unless the individual goes through the selection and creation steps, and it is considered that the degree of association with the individual's interests and expertise is strong.

  In addition, by creating a personal knowledge by weighting the input data that has a reference relationship with the output data, as with the case of creating personal knowledge by weighting the output data, individual interests and expertise Can be stored accurately. That is, as with the output data, the input data, which is the information that was referenced when creating the output data, is used by the individual in comparison to the entire input data that the individual may have just viewed. This is because the output data is created, so it is considered that there is a strong degree of association with individual interests and expertise.

  Next, a third embodiment of the present invention will be described with reference to FIGS. FIG. 18 is a functional block diagram illustrating the configuration of the information management apparatus. 19 to 20 are flowcharts showing the operation of the information management apparatus.

[Constitution]
The information processing apparatus 202 according to the present embodiment employs a configuration in which the processing units included in the information processing apparatuses 2 and 102 described in the first and second embodiments are combined. The information management apparatus 202 includes a calculation unit 220 that performs calculation processing of information and a storage unit 230 that stores information, and the information management program is incorporated in the calculation unit 220, so that FIG. As shown in FIG. 4, a data input processing unit 221, a feature value extraction processing unit 222, and a personal knowledge generation processing unit 223 are constructed, and input / output data by the user terminals 11, 12, and 13 are used as the characteristics. Based on the management function. In this embodiment, in particular, since the method of generating personal knowledge data of input / output data is different, the functions of the data input / output processing unit 221 and the personal knowledge generation processing unit 223 are different. Accordingly, the storage unit 230 includes an input / output data storage unit 231, a time data / input / output distinction data storage unit 232, a period setting data / weighting setting data storage unit 233, and a personal knowledge data storage unit 234. And are formed. These will be described in detail below.

  In addition, the calculation unit 220 includes a search request reception processing unit 224, a search processing unit 225, and a search result output processing unit 226 by incorporating the above program. However, since the functions of the processing units 224, 225, and 226 are almost the same as those in the first embodiment, the description thereof is omitted.

  The data input / output processing unit 221 receives the input / output data and stores the input / output data in the input / output data storage unit 231, and distinguishes the time data indicating the input / output time of the input / output data from the input / output of the input / output data. It has a function of accepting input / output distinction data and reference status data representing a reference relationship between the input / output data and storing them in the time data / input / output distinction data storage unit 232.

  In addition, the personal knowledge generation processing unit 223 outputs, when calculating the characteristics of a plurality of input / output data within a predetermined period in which the proximity of input / output time is high, based on the time data and the input / output distinction data. The personal knowledge data is generated by adding weights in consideration of the input / output situation such as adding weights to the data.

[Operation]
Next, the operation of the information management apparatus 202 described above will be described with reference to FIGS. FIG. 19 is a flowchart showing information storage processing by the information management apparatus for input / output data by the user terminals 11, 12, and 13 of the users A, B, and C, and FIG. 20 is a flowchart showing a part of the detailed operation. It is. As in the first and second embodiments, the operation of the information accumulation process when the user A is interested in the programming language C ++ and reads a document or creates a memo to acquire knowledge is described. To do.

  First, the data input / output processing unit 221 of the information management apparatus 202 stores the input / output data input / output at the user terminal 11 of the user A in the input / output data storage unit 231 (step S201). At the same time, the data input / output processing unit 221 of the information management apparatus 202 performs the input / output distinction data which is the distinction between the time data indicating the input / output time of the input / output data and the input / output of the data input / output by the user A. Are stored in the time data / input / output distinction data storage unit 232 (step S202). At this time, the reference relationship of the input / output data is also stored in the time data / input / output distinction data storage unit 232.

  Subsequently, the feature value extraction processing unit 222 of the information management apparatus 202 reads the stored input / output data, and extracts the feature value as in the case of the first embodiment (step S203).

  Subsequently, the personal knowledge generation processing unit 223 performs the feature value of the input / output data obtained by the feature value extraction processing unit 222 and the time data and input / output stored in the time data / input / output distinction data storage unit 232. Using the distinction data, representative features representing personal interests and expertise corresponding to the input / output situation in the input / output data in a predetermined period are generated as personal knowledge data (step S204). This operation will be further described in detail with reference to FIG.

  First, the personal knowledge generation processing unit 223 reads the period setting data from the period setting data / weighting setting storage unit 233 (step S211), and input / output of input / output data for generating personal knowledge data is performed based on the period setting data. A period is set (step S212). Subsequently, the weight setting data is read from the period setting data / weight setting data storage unit 233 (step S213), and based on this, as described in the second embodiment, each input / output data within the set period is read. The feature value is weighted (step S214). Then, the personal knowledge data is generated by calculating the sum of the word feature vectors of the feature values that are the input / output data within the set period and to which the weight is added (step S215). The method for generating the personal knowledge data is an example, and is not limited to the above method.

  Thereafter, the generated personal knowledge data is stored in the personal knowledge data storage unit 234 (step S205). Then, not only the personal knowledge data of the user A but also the personal knowledge data is created for the users B, C, etc., and stored in the personal knowledge data storage unit 234.

  Note that the search processing operation by the information management apparatus 202 when a search request is made later from the user terminal 14 of the user D is the same as that in the first embodiment, and a description thereof will be omitted.

  By doing this, personal knowledge data is created in consideration of the period during which the input / output data is input / output and the input / output distinction, so that the characteristics of the user who handled the input / output data can be more reflected. This makes it easier to use the data later. In particular, information input / output intensively during a certain past period can be searched, and personal interests and expertise can be extracted with higher accuracy.

  The present invention has industrial applicability because it can be used as a database for storing and managing input / output data by a user and enabling the search.

It is a block diagram which shows the whole structure of this invention. 1 is a functional block diagram illustrating a configuration of an information management device in Embodiment 1. FIG. It is explanatory drawing which shows an example when input / output data and time data are memorize | stored. It is explanatory drawing which shows an example when the feature value of input / output data is calculated. FIG. 5A is an explanatory diagram showing an example when generating personal knowledge data, FIG. 5A shows an example of a plurality of input / output data that are temporally continuous, and FIG. 5B shows input / output data for a predetermined period. An example when personal knowledge data is generated by using the data is shown. 3 is a flowchart illustrating an operation of information accumulation processing by the information management apparatus according to the first exemplary embodiment. It is a flowchart which shows the detailed operation | movement of step S3 of FIG. It is a flowchart which shows the detailed operation | movement of step S4 of FIG. 6 is a flowchart illustrating an operation of a search process performed by the information management apparatus according to the first embodiment. It is a flowchart which shows the detailed operation | movement of step S43 of FIG. FIGS. 11A, 11B, and 11C are flowcharts showing the detailed operation of step S44 of FIG. 9, respectively. It is a flowchart which shows the other detailed operation example of step S4 of FIG. FIG. 10 is a functional block diagram illustrating a configuration of an information management device according to a second embodiment. It is explanatory drawing which shows an example when input / output data and input / output distinction data are memorize | stored. It is explanatory drawing which shows an example when producing | generating personal knowledge data, FIG.15 (a) shows an example which added the weighting to input / output data based on the input / output distinction, FIG.15 (b) shows personal knowledge data. An example is shown when it is generated. 10 is a flowchart illustrating an operation of information accumulation processing by the information management apparatus according to the second embodiment. It is a flowchart which shows the detailed operation | movement of step S104 of FIG. FIG. 10 is a functional block diagram illustrating a configuration of an information management device according to a third embodiment. 14 is a flowchart illustrating an operation of information accumulation processing by the information management apparatus according to the third embodiment. It is a flowchart which shows the detailed operation | movement of step S204 of FIG. It is explanatory drawing for demonstrating the inconvenience in a prior art example.

Explanation of symbols

2 Information management device 11, 12, 13, 14 User terminal 21, 121 Data input / output processing unit 22, 122 Feature value extraction processing unit (feature value extraction means)
23,123 Personal knowledge generation processing unit (personal knowledge data generation means)
24,124 Search request reception processing unit (search means)
25, 125 Search processing unit (search means)
26, 126 Search result output processing unit (search result output means)
31, 131 Input / output data storage unit (input / output data storage means)
32 hour data storage (input / output data storage means)
33 Period setting data storage unit 34, 134 Personal knowledge data storage unit (information storage means)
132 Input / output distinction data storage means (input / output data storage means)
133 Weight setting data storage unit

Claims (29)

Input / output data storage means for storing input / output data input / output by the user and input / output status data representing the input / output status of the input / output data;
Feature value extraction means for extracting feature values based on predetermined criteria of the input / output data;
Personal knowledge data generating means for generating personal knowledge data representing the characteristics of the input / output data for each of the users who input / output the input / output data based on the feature values of the input / output data and the input / output status data; ,
An information management device comprising:   The information management apparatus according to claim 1, wherein the feature value extraction unit extracts a feature value based on an appearance frequency of each term included in the input / output data.   3. The information management apparatus according to claim 1, wherein the feature value extraction unit extracts a feature value based on an appearance position of each term included in the input / output data.   4. The information management apparatus according to claim 1, wherein the feature value extracting unit extracts a feature value based on emphasis information of each term included in the input / output data.   5. The information management apparatus according to claim 1, wherein the input / output status data is data representing a time when the input / output data is input / output.   6. The information management according to claim 5, wherein the personal knowledge data generating means generates the personal knowledge data based on a plurality of the input / output data input / output within a predetermined period. apparatus.   7. The information management apparatus according to claim 1, wherein the input / output status data is input / output distinction data representing input / output distinction of the input / output data.   The information management apparatus according to claim 7, wherein the personal knowledge data generating unit generates the personal knowledge data by weighting the input / output data according to the input / output distinction data.   9. The information management apparatus according to claim 8, wherein the personal knowledge data generating means weights the output data of the input / output data higher than the input data.   10. The input / output status data is reference status data representing a reference status of other data in the input / output data. The information management device described.   11. The information management apparatus according to claim 10, wherein the personal knowledge data generating means generates the personal knowledge data by weighting the input / output data in accordance with the reference situation data.   The information management apparatus according to claim 11, wherein the personal knowledge data generation unit weights data that is a reference destination of predetermined data among the input / output data. A similarity calculating means for calculating a similarity between feature values of the plurality of input / output data;
6. The personal knowledge data generating means generates the personal knowledge data for a plurality of the input / output data having a similarity equal to or greater than a predetermined value. , 6, 7, 8, 9, 10, 11 or 12.   An information storage means for storing the personal knowledge data generated by the personal knowledge data generation means is provided, wherein the personal knowledge data is provided. 1, 2, 3, 4, 5, 6, 7, 8, 9, The information management device according to 10, 11, 12, or 13.   Retrieval means for accepting search request data including data corresponding to the input / output data from another user and searching the personal knowledge data stored in the information storage means based on the search request data The information management apparatus according to claim 14.   16. The information management apparatus according to claim 15, wherein the search means searches for other personal knowledge data similar to the searched personal knowledge data based on a predetermined criterion.   The information management apparatus according to claim 15 or 16, further comprising search result output means for outputting the personal knowledge data searched by the search means.   The search result output means for reading out and outputting the input / output data corresponding to the personal knowledge data searched by the search means from the input / output data storage means, or 17. The information management device according to 17.   19. Information according to claim 15, 16, 17 or 18, further comprising search result output means for outputting user specifying information for specifying the user corresponding to the personal knowledge data searched by the search means. Management device. An information management method executed by a computer connected to input / output data storage means for storing input / output data input / output by a user and input / output status data representing the input / output status of the input / output data. ,
A feature value extracting step of extracting the feature value of the input / output data based on a predetermined criterion;
Based on the characteristic value of the input / output data and the input / output status data stored in the input / output data storage unit, the characteristic of the input / output data for each user who has input / output the input / output data is represented. Personal knowledge data generation process for generating personal knowledge data;
An information management method characterized by comprising:   21. The information according to claim 20, wherein the feature value extracting step extracts a feature value based on at least one of an appearance frequency, an appearance position, and emphasis information of each term included in the input / output data. Management method. The input / output status data is data representing the time when the input / output data was input / output,
22. The personal knowledge data generation step generates the personal knowledge data based on a plurality of the input / output data input / output within a predetermined period. Information management method. The input / output status data is input / output distinction data representing input / output distinction of the input / output data,
23. The information management method according to claim 20, wherein the personal knowledge data generating step generates the personal knowledge data by weighting the input / output data according to the input / output distinction data. .   The personal knowledge data generation step calculates similarity between feature values of the plurality of input / output data, and generates the personal knowledge data for the plurality of input / output data having the similarity equal to or greater than a predetermined value. 24. The information management method according to claim 20, 21, 22, or 23. To a computer connected to input / output data storage means for storing input / output data input / output by the user and input / output status data representing the input / output status of the input / output data,
Feature value extracting means for extracting feature values of the input / output data based on a predetermined criterion;
Based on the characteristic value of the input / output data and the input / output status data stored in the input / output data storage unit, the characteristic of the input / output data for each user who has input / output the input / output data is represented. Personal knowledge data generating means for generating personal knowledge data;
Information management program for realizing In the computer,
26. For information management according to claim 25, for realizing the feature value extracting means for extracting a feature value based on at least one of appearance frequency, appearance position, and emphasis information of each term included in the input / output data. program. In the computer,
When the input / output status data is data representing the input / output time of the input / output data, the personal knowledge data is generated based on a plurality of the input / output data input / output within a predetermined period. 27. The information management program according to claim 25 or 26 for realizing the personal knowledge data generating means. In the computer,
When the input / output status data is input / output distinction data representing the input / output distinction of the input / output data, the input / output distinction data is weighted to generate the personal knowledge data. 28. The information management program according to claim 25, 26 or 27 for realizing the personal knowledge data generating means. In the computer,
Similarity calculating means for calculating the similarity between feature values of the plurality of input / output data;
29. Information according to claim 25, 26, 27, or 28, for realizing the personal knowledge data generating means for generating the personal knowledge data for a plurality of the input / output data having a similarity equal to or greater than a predetermined value. Administrative program.

Images