JP3914082B2 - Information presentation system and information presentation method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an information presentation system and an information presentation method such as a knowledge management system that supports a plurality of users to share and manage useful knowledge and information.
[0002]
[Prior art]
In recent years, electronic communication means for exchanging information among a plurality of users using a computer such as a personal computer and a network has become widespread. Communication means represented by e-mails, mailing lists, electronic bulletin boards, news groups, etc. are already indispensable means for daily work and exchanges between individuals.
[0003]
On the other hand, in organizations such as business groups, public organizations, and non-profit organizations, knowledge and information are regarded as an asset of the organization from the standpoint that knowledge and information are an important basis for organizational activities, just like people, goods and money. Knowledge management techniques are being introduced to efficiently create, store and share this.
[0004]
Messages such as emails and articles that are exchanged on a daily basis by multiple users (hereinafter referred to as “communities”) who are active with a common purpose and interest are: Contains valuable knowledge. Communication is a method that anyone can use easily and flexibly. From the active discussions that take place, knowledge such as new ideas, detailed verification by many members, and flexible solutions to urgent problems. Often produced.
[0005]
Such knowledge is usually mixed in a large number of messages that are not useful. In particular, messages such as electronic mail are often distributed and stored in computers such as personal computers of individual users.
[0006]
Normally, a user belongs to multiple communities and performs activities according to the purpose of each community. However, there are multiple communities with similar purposes and interests in the organization, and the formation and disappearance of communities. And member participation and withdrawal are fluid. For this reason, it is difficult for individual users and the entire organization to handle information media called messages exchanged by communication means as knowledge.
[0007]
From the standpoint of knowledge management, it is important to understand the quantity and quality of knowledge distributed in the organization and identify knowledge that is often used by many people, new knowledge, lack of knowledge, etc. is there. Furthermore, it is necessary to identify a person who has contributed to the creation and dissemination of useful knowledge, and to treat him / her as an expert in the field of knowledge, or to use the degree of contribution as an assessment material.
[0008]
However, as described above, it is difficult and labor intensive to specify only valuable knowledge or specify a contributor to dissemination of knowledge from a large number of messages exchanged daily by a plurality of users. Traditionally, such work has been done manually by experts called knowledge managers or individual members of the organization, but in the first place, the value of knowledge and the contribution of members are evaluated objectively and quantitatively. In many cases, the method was not clearly defined.
[0009]
As a known technique for extracting a topic (target topic) from a message such as an electronic mail, there is "Electronic information receiving apparatus and electronic information processing method" disclosed in Japanese Patent Laid-Open No. 2000-331018. This method is a method for specifying a topic of an individual message, and can be used for the purpose of knowing whether or not the message transmitted to the user has contents of interest to the person.
[0010]
However, it cannot be used for the purpose of evaluating whether a message is important for a plurality of users in the organization. Further, since processing based on the relationship between messages is not performed and only processing based on words appearing in individual messages is performed, it cannot be used for the purpose of grasping the dynamic transition of topics.
[0011]
As a publicly known technique for extracting a user's work characteristics and contributions from a user-created document or a conversation between users, there is an "optimum actor selection support system" disclosed in Japanese Patent Application Laid-Open No. 2000-113064. This method determines a person's specialization, role, and contribution based on a document created in a certain business, conversations exchanged, business assignments and requests, and selects the right person for the business. It can be used for various purposes.
[0012]
However, this method cannot be used for the purpose of grasping the influence of people and their knowledge in the entire organization beyond business. The contribution and suitability of a person in an explicit job are mostly self-evident, whereas the contribution of a person through activities such as discussions and information dissemination voluntarily by means of communication It is difficult to measure.
[0013]
In the method of this publicly known technique, it is possible to determine whether or not the sender plays an important role based on the number and contents of email transmissions. However, the content of the message itself is effectively used in the organization. It cannot be used for the purpose of grasping how important the topic targeted by the message is in the organization or finding other related topics.
[0014]
[Problems to be solved by the invention]
Therefore, in view of the above problems, the present invention provides an information presentation system and an information presentation method capable of specifying valuable knowledge from messages exchanged on a daily basis by electronic communication means. Objective.
[0015]
Another object of the present invention is to provide an information presentation system and an information presentation method that can identify members who have contributed to the creation and spread of valuable knowledge.
[0016]
Another object of the present invention is to provide an information presentation system and an information presentation method capable of quantitatively measuring the degree of spread of knowledge within an organization for each field or topic (topic) targeted by the knowledge. And
[0017]
[Means for Solving the Problems]
The information presentation system according to the present invention is an information presentation system in which a plurality of terminal devices and a server device are connected to a network, and the server device performs a message between the plurality of terminal devices via the network. Transmission / reception means for transmitting / receiving information, first storage means for storing message information received by the transmission / reception means, and types of operations in the terminal device that is a transmission / reception source of message information transmitted / received by the transmission / reception means Based on the second storage means stored as the operation history, the operation history stored in the second storage means, and the similarity of the contents between the message information in the message information stored in the first storage means Extraction means for extracting the strength of the relationship between each message information, and strong with other message information extracted by this extraction means And it is configured message information with a lot of engagement from one or more selected presenting means for presenting.
[0018]
The information presentation method of the present invention is an information presentation method of an information presentation system configured by connecting a plurality of terminal devices and a server device to a network, and is performed between the plurality of terminal devices via the network. Stores message information, stores the operation type in the terminal device that is a transmission / reception source for the stored message information as an operation history, and the contents between the stored operation history and the message information in the stored message information The strength of the relationship between each message information is extracted based on the similarity of the message, and one or more message information having a strong relationship with the other extracted message information is selected and presented. It is characterized by.
[0019]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
[0020]
FIG. 1 shows the overall configuration of an information presentation system according to the present invention. This information presentation system is configured by connecting a plurality of user terminals 12 to a server device 10 via a communication network 11.
[0021]
The server device 10 is configured by a computer that is communicably connected to a plurality of user terminals 12 via a communication network 11. The server device 10 includes a message transmission / reception unit 1, a community management unit 2, a message relationship extraction unit 3, a community information presentation unit 4, a message search unit 5, and a message topic extraction unit 6.
[0022]
The message transmission / reception unit 1 is means for a plurality of users to transmit and receive messages. Messages (message information) are transmitted and received in a conventional manner such as e-mail, electronic bulletin board, and chat.
[0023]
The community management unit 2 is means for managing a plurality of communities (user groups) composed of a plurality of users. The community management unit 2 includes a user information storage unit 21, an operation history storage unit 22, a message storage unit 23, and a community information storage unit 24.
[0024]
The user information storage unit 21 is means for storing information of a user who is a user of the present information presentation system.
[0025]
The operation history storage unit 22 is a means for storing a history of operations such as creation, browsing, replying, forwarding, and deletion of messages by the user, and all operations for all users for messages performed by the message transmission / reception unit 1. Remember.
[0026]
The message storage unit 23 is a means for storing messages transmitted and received by the message transmission / reception unit 1.
[0027]
The community information storage unit 24 is means for storing a plurality of information on communities to which the user belongs.
[0028]
The message relationship extraction unit 3 is a means for extracting a relationship between each of a plurality of messages stored in the message storage unit 23. The extraction process is performed based on information stored in the user information storage unit 21, the operation history storage unit 22, and the message storage unit 23, which will be described in detail later.
[0029]
The community information presentation unit 4 contributes to the relationship between the messages extracted by the message relationship extraction unit 3, the importance of the message obtained based on the relationship, the degree of spread of the topic targeted by the message, and the spread This is a means for presenting information on the user and the like to the user.
[0030]
Although the present embodiment is configured by the above-described means, it has a configuration including a message search unit 5 and a message topic extraction unit 6 in order to present more useful information to the user.
[0031]
The message search unit 5 is a means for searching for a message that satisfies a condition specified by the user from the messages stored in the message storage unit 23. This means is mainly used for searching for messages related to a topic that the user is interested in by a condition such as a keyword, as well as a conventional search means capable of searching by a condition such as a message creation date and a creator It is.
[0032]
The message topic extraction unit 6 is a means for extracting a topic common to a plurality of messages. As will be described in detail later, the message topic extraction unit 6 finds a common topic from the contents discussed in a plurality of messages, and changes in the degree of spread in each topic over time (every hour) It is used to present the relationship between topics in a form that the user can grasp.
[0033]
The user terminal 12 uses a personal computer, a mobile phone, a PDA, or the like to send and receive messages such as e-mails, electronic bulletin boards, and chats via the communication network 11.
[0034]
The communication network 11 is, for example, a LAN, an intranet, the Internet, or the like.
[0035]
The configuration shown in FIG. 1 is an example in which a community, a group of users having the same purpose and interest, provides a place for performing intellectual activities through message exchange. Exchange of messages in the community is performed in the form of a conventional mailing list or electronic bulletin board, and access rights such as creation / viewing of messages are managed by the community management unit 2 so as to be given only to members of the community.
[0036]
On the other hand, the configuration shown in FIG. 2 is the minimum configuration of the embodiment according to the present invention, and is a configuration without a message exchange frame called community.
[0037]
The configuration of FIG. 2 is different from the configuration of FIG. 1 and is a simple configuration that does not include the community management unit 2 and the community information storage unit 24.
[0038]
That is, the server device 20 having this configuration includes a message transmission / reception unit 1, a message relationship extraction unit 3, an information presentation unit 4, a user information storage unit 21, an operation history storage unit 22, and a message storage unit 23. This configuration includes a message relation extraction unit 3 which is the most essential means in the present invention. The identification of important messages, which will be described later, and the identification of contributors in the dissemination of knowledge can also be performed with the configuration of FIG. In this configuration, the community information presenting unit 4 shown in FIG. 1 is simply the information presenting unit 4 because no message exchange frame called community is provided.
[0039]
In the following description of the present invention, the processing flow and effects will be described mainly based on the more complicated and effective configuration of FIG.
[0040]
FIG. 3 shows an example of user information stored in the user information storage unit 21. As shown in the figure, a user is identified by a unique user ID 31, and items of a password 32, a name 33, a mail address 34, and a community 35 are stored and managed for each user. However, as described above, in the embodiment shown in FIG.
[0041]
In FIG. 3, for example, the user ID “u1” has a name “Mori Aoki”, a mail address “aoki@aaa.net”, and a community “c1, c2,...”.
[0042]
FIG. 4 shows an example of community information stored in the community information storage unit 24. A community is identified by a unique community ID 41, and items of a community name 42, a mailing list address 43, a bulletin board URL 44, and a member 45 are stored and managed for each community.
[0043]
The mailing list address 43 is an address for a community member to send a message to all other members. On the other hand, the bulletin board URL 44 is a URL (unique resource location) of an electronic bulletin board for a community member to share a message with other members.
[0044]
Exchange of messages within the community in this embodiment can be performed using either or both of a mailing list and an electronic bulletin board. When both means are used, the messages sent to the mailing list are posted on the electronic bulletin board, and conversely, the messages written on the electronic bulletin board are linked together so that they are also sent to the mailing list.
[0045]
The item of the community member 45 is described by the user ID shown in FIG. 3, and conversely, the item of the user's community 35 shown in FIG. 3 is described by the community ID.
[0046]
In FIG. 4, for example, the community 46 has a community ID “c2”, a community name “mobile technology study group”, a mailing list address “mobile-ml@xxx.net”, and a bulletin board URL “http: // xxx. net / mobile-bbs "and members are" u1, u2, u3, u19, u26, u68, ... ".
[0047]
Further, the community 47 has a community ID “c4”, a community name “Palm Friends Association”, a mailing list address “palm-ml@xxx.net”, no bulletin board URL, and members “u2, u3, u13, u26, … ”.
[0048]
FIG. 5 shows an example of an operation history stored in the operation history storage unit 22. As shown in the figure, items of operation date 51, user ID 52 who performed the operation, operation type 53, message ID (in the case of an operation on an existing message) 54, and message ID (in the case of creating a new message) 55 Are sequentially stored.
[0049]
For example, the operation 56 is a history that the user “ul” “created” a new message “m1” on the date “2002/2/8 10:24:17”.
[0050]
The operation 57 is a history that the user “u2” browses the message “m1” on the date “2002/2/8 10:25:34”.
[0051]
Further, the operation 58 is a history that the user “u2” returns the message “m2” to the message “m1” on the date “2002/2/8 10:29:24”.
[0052]
In this way, the user operation history is sequentially stored in the operation history storage unit 22.
[0053]
In the case of this embodiment, there are five standard types of operations for messages: “create”, “browse”, “reply”, “forward”, and “delete”, but the present invention includes these five types. The present invention is not limited, and the intention of the operation can be extracted as will be described later.
[0054]
FIG. 6 shows an example of community list information presented to the user from the community information presentation unit 4. In the case of the configuration illustrated in FIG. 1, the community information presentation unit 4 presents a list of communities in which the user is participating and non-participating communities to the user after performing user authentication.
[0055]
In the case of participating communities, each community name 61 and important messages 62 within the community are presented.
[0056]
In the case of the non-participating community 63, an important topic 64 in the community is presented. Here, the case where the important message is not presented to the non-participating community is a case where the access right to the message itself is not given to the non-participating user.
[0057]
In this embodiment, the important message is not presented, but the important message itself may be presented to non-participating users.
[0058]
For example, when operating on an intranet in a company, it is preferable to set the access right strictly. When targeting any user on the intranet, all users are set without setting the access right. Message viewing authority may be given.
[0059]
The process of automatically identifying important messages and important topics in the community is a characteristic process of the present invention, the details of which will be described later. By such presentation, the user can identify important messages and topics in the community. You will be able to conduct activities such as knowing and browsing immediately, knowing the existence of non-participating communities that are discussing topics that interest you, and participating.
[0060]
FIG. 7 shows an example of mail list information presented to the user terminal 12 from the community information presentation unit 4. In the configuration shown in FIG. 2, although a forum for community discussion is not provided explicitly, it is used for exchanging useful information between users who share the same purpose and interest using communication means such as e-mail. Messages are exchanged.
[0061]
The example shown in FIG. 7 presents the importance 71 for each message transmitted to a certain user terminal 12. By such presentation, the user can easily distinguish important ones from a large number of messages transmitted every day. The process of specifying the important message is realized by the same process as that of the embodiment shown in FIG. 1, and details thereof will be described later.
[0062]
FIG. 8 shows an example of a message stored in the message storage unit 23.
[0063]
A message 81 shown in FIG. 8A is a new message created by the user “Aoki”. On the operation history shown in FIG. 5, the operation that created this message corresponds to the operation 56.
[0064]
A message 82 shown in FIG. 8B and a message 83 shown in FIG. 8C are reply messages to the message 81.
[0065]
Further, the message 84 shown in FIG. 8D is a reply message to the message 83.
[0066]
The examples of FIGS. 8A to 8D show a part of a message exchanged by a member whose community name is “Mobile Technology Study Group” (operation 46 in FIG. 4) using the mailing list of the community. Yes.
[0067]
On the other hand, FIG. 9 shows an example in which messages are exchanged using an electronic bulletin board. Here, an example of messages exchanged on the electronic bulletin board in the community name “Internet home appliance planning team” (operation 47 shown in FIG. 4) is shown.
[0068]
An electronic bulletin board 91 shown in FIG. 9A is a screen for browsing messages. Messages 92, 93, and 94 respectively indicate the title portion of the message. The message 93 is a reply message to the message 92. Message 95 is the body of message 94.
[0069]
The members of the community browse the message using such a screen, and create a new message or create a reply message as necessary.
[0070]
An electronic bulletin board 96 shown in FIG. 9B shows a screen for creating a reply message to the message 94. The message 97 is the title of the reply message to be created, and the message 98 is its body. In this example, the user “Ito” who browsed the message 81 shown in FIG. 8A creates a message for transferring the message.
[0071]
It is the first object of the present invention to identify important messages from the messages shown in FIGS. 8A to 8D and FIGS. 9A and 9B. For this purpose, first, Extract relationships between messages.
[0072]
Next, processing for extracting the relationship between messages by the message relationship extracting unit 3 in such a configuration will be described with reference to the flowchart of FIG. The message relationship extraction unit 3 quantitatively calculates the relationship between two messages by using a term of dependency between messages and a term of text similarity in the message contents.
[0073]
For any two messages mi and mj, if the author of mi is ui and the author of mj is uj (step 101), the message relation extracting unit 3 first determines whether mj is a reply message of mi. Confirm (step 102).
[0074]
When mj is a reply message of mi, the dependency Sdep (mi, mj) of mj on mi is set as a constant Srep (step 103).
[0075]
If mj is not a reply message of mi in step 102, it is confirmed whether mj is a transfer message of mi (step 104).
[0076]
When mj is a transfer message of mi, the dependency Sdep (mi, mj) of mj on mi is set as a constant Sfwd (step 105).
[0077]
Here, the similarity between the texts of the contents of the messages mi and mj is obtained, and this is set as Text (mi, mj) (step 108), and the process proceeds to step 1011. The similarity of text is obtained by a conventional method based on the similarity of word vectors included in the text.
[0078]
If mj is not a transfer message with mi in step 104, the operation history storage stores whether or not the message uj was created within time t (0 ≦ t ≦ Tt) after the user uj viewed the message mi. The operation history stored in the unit 22 is used to check (step 106).
[0079]
Here, Tt is a threshold value, and when the message mj is created after a time longer than the time t, or when the user uj is not browsing the message mi in the first place, the interval between mi and mj Assume that there is no relationship, and the degree of association between mi and mj is R (mi, mj) = 0 (step 1012).
[0080]
If the time t (0 ≦ t ≦ Tt) is satisfied in step 106, the dependency Sdep (mi, mj) of the message mj on the message mi is set as a monotonously decreasing function Sref (t) of the time t (step 107). That is, when the user uj creates the message mj without passing a large time after viewing the message mi, it is considered that the relationship between mi and mj is large. What is required in this step is not an explicit relationship such as reply or forwarding, but an implicit relationship in which another message is created by referring to or quoting one message.
[0081]
Subsequently, the similarity between the texts of the messages mi and mj is obtained, and this is set as Text (mi, mj) (step 109).
[0082]
Here, the obtained Text (mi, mj) is compared with the threshold value St (step 1010). If Text (mi, mj) is smaller than the threshold value St, mi and mj are not related, and mi and mj Assume that the relevance is R (mi, mj) = 0 (step 1012). That is, it is considered that the implicit relationship estimated in steps 106 and 107 cannot be supported by the text of the content.
[0083]
In step 1010, if Text (mi, mj) is equal to or greater than the threshold value St, the process proceeds to step 1011.
[0084]
In step 1011, the dependency Sdep (mi, mj) based on the operation history between the messages obtained in the processing up to step 108 or step 1010 and the similarity of the content text Text (mi, mj) are respectively expressed as constants C1. , C2 multiplied by R (mi, mj) = C1 × Sdep (mi, mj) + C2 × Text (mi, mj) is set as the relevance between mi and mj.
[0085]
It is considered that the relationship between mi and mj is stronger as the value of R (mi, mj) is larger.
[0086]
Note that R (mi, mj) has directionality, and R (mi, mj) ≠ R (mj, mi). This is because Sdep (mi, mj) ≠ Sdep (mj, mi).
[0087]
Next, processing for specifying an important message by the message relationship extraction unit 3 based on the strength of the relationship between individual messages obtained by the above processing will be described with reference to the flowchart of FIG.
[0088]
When obtaining the importance level of the message mi, the message relation extracting unit 3 first obtains the number of times of browsing Fref (mi) and the number of times of deletion Fdel (mi) based on the operation history (step 111).
[0089]
Here, the number of deletions is the number of times a message transmitted to a plurality of members by means such as a mailing list is deleted by each recipient. It can be considered that a message is more useful as the number of times it is viewed, and conversely, it can be regarded as an unnecessary message as the number of deletions is increased.
[0090]
Using the values of the browsed number of times Fref (mi) and the number of times of deletion Fdel (mi) and the degree of association R (i, j) between mi and another message mj obtained in FIG. Next) is calculated (step 112). It can be obtained by the following formula.
[0091]
I (mi, 0) = C3 × Fref (mi) −C4 × Fdel (mi) + C5 × ΣR (mi, mj)
Here, Σ means that the sum of R (mi, mj) is taken for all messages mj related to the message mi. C3, C4, and C5 are positive constants. The more frequently viewed and the fewer deleted times, the more important messages are considered. Further, as the number of related messages, typically, reply or transfer, or implicit reference or quotation, is increased, the message is considered to be an important message.
[0092]
In the above equation, the importance of the message mi is obtained by taking into account the relevance with the message mj that is directly related to the message mi. Based on this, the nth degree of importance is recursively obtained by the following equation (step 113, 114).
[0093]
I (mi, n) = I (mi, n−1) + C6 × ΣI (mj, n−1)
Here, C6 is a constant, and Σ means summing up all the messages mj related to the message mi as described above. The n-th degree of importance is nothing but the importance of not only the importance of messages directly related to the message mi but also the importance of messages related through n stages. Therefore, the constant C6 is set to a value larger than “0” and smaller than “1”, and the influence on the importance is reduced as n increases.
[0094]
The important messages shown in FIGS. 6 and 7 are automatically and appropriately identified by the processing from steps 111 to 114 in FIG. 11, and the important messages shown in FIGS. 6 and 7 are presented to the user terminal 12 via the community information presentation unit 4. be able to.
[0095]
Further, it is possible to grasp how many messages related to a certain topic are exchanged by using the processing for obtaining the strength of the relationship between messages described in FIG.
[0096]
For example, when the user of the user terminal 12 wants to grasp the degree of spread of the technical topic “BlueTooth (TM)”, the information of “BlueTooth (TM)” is transmitted from the user terminal 12 to the community information presentation unit 4 via the communication network 11. Request.
[0097]
First, the community information presentation unit 4 uses the message search unit 5 to search for a message including this word using the word “BlueTooth (TM)” as a search condition. Subsequently, the community information presentation unit 4 obtains a set of messages having a degree of relevance greater than a certain threshold with the search result message, and transmits information on the temporal transition of the number of messages to the user terminal 12 via the communication network 11. And present it.
[0098]
FIG. 12 shows an example in which “BlueTooth (TM)” information is presented to the user terminal 12 in response to a request from the user terminal 12.
[0099]
A black circle 131 in FIG. 12 indicates a message including the word “BlueTooth (TM)” (that is, a message directly searched by the message search unit 5), and a white circle 132 includes the word “BlueTooth (TM)”. Indicates a message that is not related to this topic.
[0100]
The contents of the message may be presented to the user in response to the user instructing each circle using a pointing device such as a mouse. For example, as shown at 133 in FIG. 12, “Do you know BlueTooth (2/8 Aoki)” is displayed.
[0101]
The time axis may be displayed in the vertical direction (every hour), and the number of messages may be displayed in the horizontal direction.
[0102]
According to the present invention, it is possible to present not only messages that satisfy the search condition but also related messages by using the relationship between messages, so it is more accurate and can cope with fluctuations in topic terms. To grasp the degree of popularization of selected topics.
[0103]
FIG. 13 shows an example in which the degree of topic spread is presented to the user terminal 12 for each community. In FIG. 13, communities such as “Mobile Technology Study Group”, “Internet Home Appliance Planning Team”, and “Palm Friendship Club” are presented.
[0104]
In this way, by presenting the number of messages for each community, the user can easily grasp which community has been interested in the topic from which time.
[0105]
In addition, it is possible to easily identify a person who introduced the topic for the first time in the community, a community that is interested in the topic for the first time in the entire organization, a community that is most interested in the topic, and the like.
[0106]
FIG. 14 shows an example in which the time transition of the number of users who have performed an operation on a message is presented on the user terminal 12.
[0107]
A bar graph 151 indicated by diagonal lines indicates the number of users who performed an operation on a message including the word “BlueTooth (TM)”, and a bar graph 152 indicated by white indicates an operation on this message including a related message. It shows the number of users who went.
[0108]
By indicating the number of users, not the number of messages or the number of operations, it is possible to accurately know how many users the topic is popular with. That is, it is possible to distinguish and grasp when the same user performs a large number of operations and when many users perform a small number of operations.
[0109]
FIG. 15 shows an example in which the time transition of the number of communities that have performed the operation is presented on the user terminal 12.
[0110]
In this case, it is possible to distinguish whether the topic is a topic that is interested in a specific small number of communities in the organization or a topic that is widespread in many communities. .
[0111]
FIG. 16 shows an example in which the temporal transition of the number of operations for each type of operation is presented on the user terminal 12. The meanings of the shaded and white bar graphs are the same as in FIGS.
[0112]
For example, a topic that is viewed only by a large number of users and has a small number of operations such as message creation / reply / forward and a topic that is frequently created / reply / forward a message have different topic characteristics.
[0113]
The former can be regarded as a typical topic for administrative communication matters, and the latter can be regarded as an intellectual activity with active discussion. The presentation shown in FIG. 16 makes it possible to grasp such a difference in topic properties. In addition, it can be determined that the message has a smaller value as the number of deleted messages increases.
[0114]
FIG. 16 described above is an example in which the types of operations for messages are defined as five basic and relatively simple types of creation, browsing, reply, transfer, and deletion, and the trend is presented. It is also possible to estimate the time and present the temporal transition.
[0115]
For this purpose, the message relation extraction unit 3 performs processing based on the history of the five types of operations and the text of the message content. For example, it is estimated that a message in which text including a question mark or a question word is described has an intention of a question, and a reply message to the message is likely to have an intention of an answer.
[0116]
A new message that does not contain a question mark or question word is intended to be introduced, and a reply message that quotes and denies a part of the message can be assumed to be a correction.
[0117]
In addition, it is possible to estimate that a reply message that does not include a question mark, a question word, or a negative sentence but includes reference information (for example, URL) to new information or a new term has an intention of augmentation.
[0118]
Furthermore, the intention of evaluating whether a message is regarded as unnecessary or useful can be drawn from operations such as reading a message and deleting it immediately or repetitively reading it without deleting it.
[0119]
FIG. 17 shows an example in which the time transition of the number of messages is presented to the user terminal 12 for each operation intention estimated in this way.
[0120]
By presenting the number of messages for each operation intention, a more detailed understanding can be performed. In other words, it is possible to know whether the discussion on the topic is actively conducted and whether the knowledge on the topic is refined in a favorable manner by interactive activities between users such as questions, answers, corrections, and augmentations. it can.
[0121]
FIG. 18 shows an example in which the number of messages of the user is presented for each operation intention. According to this, it is possible to easily grasp what activities the user performs on a certain topic and how the user contributes to the community.
[0122]
For example, the user “Aoki” 191 often introduces new knowledge, and in comparison to this, the “Ito” 192 answers other users' questions and corrects or augments the knowledge. I understand that there are many.
[0123]
The hatched bar graph is the number of messages related to the topic of interest (for example, “BlueTooth (TM)”) itself, as in FIGS. 14 and 15 and the like, and the white bar graph is a direct line about the topic. This indicates the number of messages that are not related to each other but are somewhat related.
[0124]
For example, the user “Aoki” 191 is not limited to those directly related to the topic, but introduces a wide range of new information on this field (for example, wireless technology). On the other hand, it can be seen that the user “Ito” 192 pays particular attention to the topic and contributes to activities that make the knowledge detailed and accurate.
[0125]
On the other hand, although the user “Yamada” 193 contributes little to the refinement of knowledge like “Ito”, it actively absorbs knowledge on this topic by asking other members questions and confirming them actively. I understand that.
[0126]
It can be seen that the user “Sato” 194 contributes little to this topic.
[0127]
Thus, according to the present invention, the contribution degree of members in the community can be quantitatively evaluated. The user's contribution depends on what topic field the user specializes in, but by presenting the total number of messages related to a plurality of topics in the format shown in FIG. Can understand the degree of contribution.
[0128]
Furthermore, by presenting the number of messages for each community, not for each user, it is possible to grasp the degree of contribution of the community to the entire organization.
[0129]
Next, the topics that community members focus on and discuss often change dynamically over time while being related to other topics. If the temporal transition of a plurality of topics related to each other can be presented to the user, the user can know a new topic to be noticed related to the topic he is interested in. It is also possible to grasp the overall trend of the field to which the topic group belongs.
[0130]
Therefore, in the present invention, first, a common topic is automatically extracted from a plurality of mutually related messages, and further, temporal transitions of the plurality of mutually related topics are converted into quantitative values such as the number of messages and the number of users. Present based on.
[0131]
Next, a process of automatically extracting a common topic from a plurality of messages by the message topic extraction unit 6 will be described with reference to a flowchart of FIG.
[0132]
The message topic extraction unit 6 first sets a certain message set M as a topic extraction target (step 121). Since the elements are automatically added to the message set M in the subsequent steps 122 to 124, it is not necessary for the user to provide all elements from the user terminal 12 from the beginning. As an effective method, let M be a message set in which the important message identified in the process of FIG.
[0133]
Subsequently, for a message mj that is not an element of M, the relevance level R (mi, mj) or R (mj, mi) between the messages obtained in the process of FIG. 10 described above is greater than or equal to a certain threshold value Rt. It is checked whether or not the element message mi exists (step 122).
[0134]
If such a message mi exists, the message mj is newly added to M (step 123). In other words, the message mj is considered to be highly related to the messages in the message set M and likely to have a common topic.
[0135]
Steps 122 and 123 are repeated until there are no more messages to be added to the message set M (step 124). As a result, the message set M becomes a set of messages that are strongly related to each other.
[0136]
When there are no more messages to be added to M in step 124, words that are statistically significantly included in the messages of the message set M are selected as words representing the topic “M”. This process is realized by a conventional statistical method such as an x-square test.
[0137]
Through the above-described processing, one set of messages related to the common topic is obtained. By creating a plurality of such sets and examining whether there is a duplication of messages between each other, it is possible to examine whether there is a relationship between a plurality of topics. That is, it can be considered that topics having duplicate messages are related to each other.
[0138]
FIG. 20 is an example in which the time transition of the number of messages belonging to each of the topics related to each other is presented to the user terminal 12 by the processing described above.
[0139]
For example, it can be seen that a topic indicating the technology “wireless LAN” 202 has been widely interested for a long time, compared to a relatively new topic about the technology of wireless communication “BlueTooth (TM)” 201.
[0140]
In addition, the user knows the existence of the topic of the product to be applied to the “BlueTooth (TM)” technology called “home server” 203 and the topic as the component element of “home server” called “DVD-RAM” 204, The trend of the spread can be grasped.
[0141]
As described above, according to the present invention, valuable knowledge can be identified from messages exchanged on a daily basis using communication means. The value of a message is accurately and quantitatively evaluated by the type and number of explicit and implicit operations performed on that message. By appropriately presenting the valuable message obtained in this way to the user, knowledge created and refined by a plurality of communities in the organization can be spread across the boundaries of the community.
[0142]
In addition, the members who have contributed to the creation and diffusion of valuable knowledge can be identified by the type and number of operations on the message. Considering the intention of the operation of the message, it is possible to quantitatively evaluate the contribution of members in the process of refinement and dissemination as well as knowledge creation. It can be done easily.
[0143]
Furthermore, it is possible to quantitatively measure the degree of knowledge spread within an organization for each target topic of the knowledge, and to present a user with an overview of relationships between multiple topics and temporal transitions. As a result, it is possible to confirm whether knowledge has been disseminated without excess or deficiency, and it becomes easy to know a new topic to be noted.
[0144]
In the information presentation system of the present invention, information may be presented to the user terminal 12 in response to a search request from the user terminal 12, and information is periodically presented to each user terminal 12 from the community information presentation unit 4. You may make it do.
[0145]
As described above, according to the embodiment of the present invention, valuable knowledge can be specified from messages exchanged on a daily basis by electronic communication means.
[0146]
It is also possible to identify members who have contributed to the creation and dissemination of valuable knowledge.
[0147]
Furthermore, it is possible to quantitatively measure the spread of knowledge within an organization for each field or topic targeted by the knowledge.
[0148]
Note that the present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the scope of the invention in the implementation stage. In addition, the embodiments may be appropriately combined as much as possible, and in that case, the combined effect can be obtained. Furthermore, the above embodiments include inventions at various stages, and various inventions can be extracted by appropriately combining a plurality of disclosed constituent elements. For example, even if some constituent requirements are deleted from all the constituent requirements shown in the embodiment, the problem described in the column of the problem to be solved by the invention can be solved, and the effect described in the column of the effect of the invention Can be obtained as an invention.
[0149]
【The invention's effect】
As described above in detail, according to the present invention, valuable knowledge can be specified from messages exchanged on a daily basis by electronic communication means, contributing to the creation and dissemination of valuable knowledge. To provide an information presentation system and an information presentation method that can identify members and quantitatively measure the degree of knowledge spread within an organization for each field or topic (topic) targeted by the knowledge. it can.
[Brief description of the drawings]
FIG. 1 is a block diagram showing the overall configuration of an information presentation system according to the present invention.
FIG. 2 is a block diagram showing another configuration of the information presentation system according to the present invention.
FIG. 3 is a view showing an example of user information stored and managed in a user information storage unit;
FIG. 4 is a diagram showing an example of community information stored and managed in a community information storage unit.
FIG. 5 is a diagram showing an operation history for a message stored and managed in an operation history storage unit;
FIG. 6 is a diagram showing an example of presenting a list of communities, important messages, and important topics.
FIG. 7 is a diagram showing an example in which a list of messages and their importance levels are presented.
FIG. 8 is a diagram showing an example of an e-mail message.
FIG. 9 is a diagram showing an example of a message from an electronic bulletin board.
FIG. 10 is a flowchart for explaining a process of extracting a relationship between messages.
FIG. 11 is a flowchart for explaining processing for specifying an important message;
FIG. 12 is a diagram showing an example in which a temporal transition of the number of messages related to a certain topic is displayed.
FIG. 13 is a diagram showing an example in which the temporal transition of the number of messages related to a topic is displayed for each community.
FIG. 14 is a diagram showing an example in which a temporal transition of the number of users who performed an operation on a message is displayed.
FIG. 15 is a diagram showing an example in which a temporal transition of the number of communities that have been operated is displayed.
FIG. 16 is a diagram showing an example in which the temporal transition of the number of operations is displayed for each type of operation.
FIG. 17 is a diagram showing an example in which the time transition of the number of messages is displayed according to intention.
FIG. 18 is a diagram showing an example in which the number of messages for each user and each intention is displayed.
FIG. 19 is a flowchart for explaining processing for extracting a group of messages and topics related to each other from a plurality of messages;
FIG. 20 is a diagram showing an example in which the time transition of the number of messages is displayed for each of a plurality of topics related to each other.
[Explanation of symbols]
1 ... Message transmission / reception unit (transmission / reception means)
2 ... Community Management Department
3 ... Message relation extraction unit (extraction means, estimation means)
4 ... Community information presentation part (presentation means)
5 ... Message search part (search means)
6 ... Message topic extraction unit (second extraction means)
10: Server device
11. Communication network
12 ... User terminal (terminal device)
21 ... User information storage unit (storage means)
22 ... Operation history storage unit (storage means)
23. Message storage unit (storage means)
24. Community information storage unit (storage means)

Claims (20)

In an information presentation system configured by connecting a plurality of terminal devices and server devices to a network,
The server device is
Transmitting / receiving means for transmitting / receiving message information performed between the plurality of terminal devices via the network;
First storage means for storing message information received by the transmission / reception means;
In the terminal device that is a transmission / reception source of message information transmitted / received by the transmission / reception means, (a) “creation” of message information, “browsing” of existing message information, “reply”, “forwarding”, and “existing” of existing message information Operation type indicating which operation among a plurality of operations including “Delete” has been performed, (b) Operation target when the operation type is “Browse”, “Reply”, “Transfer”, and “Delete” ID of the existing message information, (c) ID of the created message information when the type of operation is “Create”, “Reply”, and “Transfer”, (d) Date and time of operation, and (e) The terminal Second storage means for storing an operation history including a user ID of the device;
Based on the degree of dependence of the created message information obtained from the operation history on the existing message information according to the type of operation, and the similarity of the contents between the created message information and the existing message information Relevance calculating means for calculating relevance indicating the strength of the relationship from the message information to other message information for each message information;
When a search request is made from the terminal device, search means for searching message information stored in the first storage means based on a search condition included in the search request ;
Means for obtaining a message set including message information as a search result of the search means and message information having a degree of association between the message information as the search result and a predetermined threshold value or more;
A first presentation means for presenting the message set and information indicating a temporal transition of the number of message information in the message set;
An information presentation system.   2. The information presentation system according to claim 1, wherein the message information transmitted / received by the transmission / reception means is transmitted / received in the form of an electronic mail, an electronic bulletin board, or a chat.   The first presenting means presents the number of message information as a search result in the message set and the number of message information having a degree of relevance equal to or higher than a predetermined threshold in an identifiable manner. 1. The information presentation system according to 1. The server device stores information on each user of the plurality of terminal devices, and information on a plurality of user groups that each user transmits and receives message information with the same purpose and interest through the network. Further comprising storage means;
The first presenting means presents information indicating temporal transition of the message set and the number of message information in the message set for each user group stored in the third storage means. The information presentation system according to claim 1.   The first presenting means is characterized by presenting the number of message information as a search result in the message set for each user group and the number of message information having a degree of relevance equal to or higher than a predetermined threshold. The information presentation system according to claim 4.   The server device presents information indicating a temporal transition of the number of users who are transmission / reception sources of each message information in the message set based on the user and user group information stored in the third storage unit. 5. The information presenting system according to claim 4, further comprising second presenting means.   The server device presents information indicating a temporal transition of the number of user groups as transmission / reception sources of each message information in the message set based on the user and user group information stored in the third storage unit. 5. The information presenting system according to claim 4, further comprising third presenting means.   For each type of operation stored in the second storage unit, the server device presents the message set and information indicating temporal transition of the number of message information in the message set. The information presentation system according to claim 1, further comprising means.   The fourth presenting means presents the number of message information as a search result in the message set for each type of operation and the number of message information having a degree of relevance equal to or greater than a predetermined threshold in an identifiable manner. The information presentation system according to claim 8. The server device is
For each message information in the message set, estimation means for estimating the intention of the message information from the content of the message information and the type of operation for the message information stored in the second storage means;
For each intention of the message information estimated by the estimation means, the message set further includes fifth presentation means for presenting information indicating a temporal transition of the number of message information in the message set. The information presentation system according to claim 1.   The fifth presenting means is characterized by presenting the number of message information as a search result in the message set for each intention and the number of message information having a degree of relevance equal to or higher than a predetermined threshold in an identifiable manner. The information presentation system according to claim 10.   11. The information presentation system according to claim 10, wherein the intention of the message information estimated by the estimation means is a question, answer, introduction, correction, augmentation, confirmation, summary, illustration, evaluation, and the like. The server device further includes third storage means for storing information of each user of the plurality of terminal devices,
The fifth presenting means stores the message set for each intention of the message information and information indicating the temporal transition of the number of message information in the message set of the user stored in the third storage means. The information presentation system according to claim 10, wherein information is presented for each user. The server device is
For each message information stored in the first storage means, based on the degree of association between the message information and other message information, the number of times the message information is viewed, and the number of times the message information is deleted Importance calculating means for calculating the importance of information,
Among the message information stored in the first storage means, the message information that is greater than or equal to a predetermined threshold value between the important message information specified based on the importance and the importance message information. A means for obtaining a message set for topic extraction including:
A topic extraction means for extracting a topic word of the message set from the message set to be extracted;
Sixth presenting means for presenting information indicating the temporal transition of the number of message information in the message set from which the topic word is extracted for each topic word extracted by the topic extracting means;
The information presentation system according to claim 1, further comprising: An information presentation method in an information presentation system configured by connecting a plurality of terminal devices and a server device to a network,
The server device storing in a first storage means message information transmitted and received between the plurality of terminal devices via the network;
In the terminal device from which the server device transmits and receives each message information, (a) “create” message information, “browse” existing message information, “reply”, “forward” and “delete” to existing message information Type of operation indicating which operation among a plurality of operations including “” is performed, and (b) when the type of operation is “view”, “reply”, “forward”, and “delete”, ID of existing message information, (c) ID of created message information when operation type is “Create”, “Reply”, and “Transfer”, (d) Date and time of operation, and (e) Terminal device concerned Storing an operation history including the user ID in the second storage means;
The server device determines the dependency of the created message information obtained from the operation history on the existing message information according to the type of operation, and the content between the created message information and the existing message information. Calculating a relevance level indicating the strength of the relationship from the message information to other message information, based on the similarity,
When the server device makes a search request from the terminal device, searching the message information stored in the first storage means based on a search condition included in the search request ;
The server device obtains a message set including message information as a search result and message information having a degree of association between the message information as the search result and a predetermined threshold value or more;
A first presentation step in which the server device presents the message set and information indicating a temporal transition of the number of message information in the message set;
Information presentation method. The server device stores information on each user of the plurality of terminal devices, and information on a plurality of user groups that each user transmits and receives message information with the same purpose and interest via the network, as a third storage unit Further comprising the step of storing
The first presenting step presents, for each user group stored in the third storage unit, information indicating the time transition of the message set and the number of message information in the message set. The information presentation method according to claim 15.   A second presentation in which the server device presents information indicating a time transition of the message set and the number of message information in the message set for each type of operation stored in the second storage unit. The information presentation method according to claim 15, further comprising a step. For each message information in the message set, the server device estimates the intention of the message information from the content of the message information and the type of operation for the message information stored in the second storage means. An estimation step;
And further comprising a third presenting step for presenting information indicating temporal transition of the message set and the number of message information in the message set for each intention of the message information estimated in the estimating step, The information presentation method according to claim 15. The server device further includes a step of storing information of each user of the plurality of terminal devices in a third storage unit,
In the third presenting step, the message set for each intention of the message information and the information indicating the temporal transition of the number of message information in the message set are stored in the third storage unit. The information presentation method according to claim 18, wherein information is presented for each user. For each message information stored in the first storage means, the server device determines the degree of association between the message information and other message information, the number of times the message information is viewed, and the number of times the message information is deleted. Based on the importance calculation step for calculating the importance of the message information,
Among the message information stored in the first storage means, the message information that is greater than or equal to a predetermined threshold value between the important message information specified based on the importance and the importance message information. Obtaining a message set subject to topic extraction including:
A topic extraction step for extracting a topic word of the message set from the message set to be extracted;
A fourth presenting step for presenting information indicating temporal transition of the number of message information in the message set from which the topic word is extracted for each topic word extracted in the topic extraction step;
The information presentation method according to claim 15, further comprising:

Images