JP2002140441A - Information exchanging method and its device and computer readable recording medium with information exchanging program recorded - Google Patents

Abstract

PROBLEM TO BE SOLVED: To organize the proper group of even participants with various careers and hobbies through a network, and to improve the effect of group learning (discussion). SOLUTION: A plurality of group patterns are prepared by dividing participants based on a participant registration file, and stored in a group pattern file (S105), and mean difference level dtotal for each group pattern is calculated based on an answer file, a difference level file, an attribute file, and the group pattern file, and the group pattern is selected based on the mean difference level dtotal, and stored in a result output file (S113, 115), and the group pattern stored in the result output file is read out, and the result of the group organization of the participants of group learning or discussion is displayed at a terminal according to the group pattern (S117).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information exchange method, a computer-readable recording medium recording an information exchange program, and an information exchange apparatus. The present invention relates to an information exchange method, a computer-readable recording medium storing an information exchange program, and an information exchange device, which are organized into appropriate groups according to the themes and enhance the efficiency of discussion and learning between participants.

[0002]

2. Description of the Related Art In recent years, information exchange between individuals, for example,
It is performed by various methods such as telephone, e-mail, and bulletin board. These are generally systems suitable for exchanging information between an unspecified number of individuals.

[0003] On the other hand, for exchanging information between pre-registered individuals (terminals), for example, TV conferences, network learning, and the like have attracted attention. In these, a fixed network was established between the participants at the scheduled date and time, and on this network, for example, meetings, education,
Learning and the like are performed.

As a pre-registered information exchange system between individuals, for example, in the field of learning, there is a one-way learning system (teaching machine or the like) from the teacher side to the student side. In addition, an educational method has been proposed that allows not only one-way transmission of teaching materials but also questions from students to teachers (Japanese Patent Publication No. 8-21891: University of Illinois). As a learning and teaching method, recently, so-called group learning based on teaching between participants (learners) has become effective.

[0005] This group learning system, for example, in the case of language education (LL system), in addition to simultaneous transmission of teaching materials, enables two-way information exchange by voice between students, A system has been disclosed in which a combination of students can be arbitrarily set (Japanese Patent Application Laid-Open No. 53-135730: Sony).

[0006]

However, the above-mentioned group learning system is effective for mutual teaching with a small number of people, but cannot obtain an appropriate learning effect when the number of participants (learners) increases. A case is assumed. In addition, the designation of grouping of learners for performing group learning is performed by a teacher's switch operation, and does not describe a so-called group formation method.

Further, the present inventors have studied a method of group formation in order to effectively perform group learning for learners such as school education and workplace education.
A learning instruction system for examining test answers by learner's mutual consultation (Japanese Educational Technology Magazine 5, 23-33,
1980) discloses a group formation method in which the correctness of a learner's answer is compared by a computer, and effective discussion is performed based on a mutual combination of correctness and wrongness and a seating arrangement in a classroom.

[0008] The above-mentioned learning instruction system included in the study of the present inventors can be said to be an effective method in that it can compensate for mutual deficiencies in knowledge. In addition, it has been actually proved that, when a plurality of pre-registered learners are organized into several groups, the effect of discussion is enhanced by performing appropriate grouping.

However, in the above-mentioned learning instruction system, it is difficult to apply to learning on a network, and when many learners with different backgrounds perform group learning on the network, the motivation of these learners is high. May be difficult to extract sufficiently.

[0010] As described above, when a group of learners (the number of learners is very large or the learner's background is different) is to be discussed at a school through a computer terminal on a specific theme, for example, , To ensure that all learners are able to express their opinions and engage in lively discussions, properly organize the group of learners into small groups,
It is hoped that arrangements will be made to provide an opportunity for discussion. The information exchange method and apparatus for performing the above-described processing for a group of learners is a new educational method, and further, a computer-readable recording medium that records this information exchange program is It is expected to contribute to the development of education-related software industry that makes full use of network technology.

[0011] Further, as an application form of information exchange by a group, there are, for example, a subcommittee-type debate and a debate (controversial group debate). In these cases, efficient information exchange is performed. There is an eager need for a group formation method that can be performed.

In view of the above, the present invention enables a participant, for example, a participant with multiple backgrounds and multiple preferences, to form an appropriate group and to achieve the effect of group learning (discussion) via a network. An object of the present invention is to provide an information exchange method, a computer-readable recording medium on which an information exchange program is recorded, and an information exchange device. Another object of the present invention is, for example, to increase the motivation of learners to learn and to reduce the burden on specialized educators.

[0013]

According to a first solution of the present invention, a question presentation step of reading a question corresponding to a learning or discussion theme from a question file and presenting it to a participant of the learning or discussion; An answer input step of creating an answer file based on the answer from the participant to the question, and a group pattern number and a group number based on the participant registration file storing the participant identifier and the number of persons, the number of persons in the group, and the selection criteria. A group pattern dividing step of creating a group pattern dividing file storing participant identifiers corresponding to the above, the answer file, a dissimilarity file preliminarily storing relevance between question answer options, and the group pattern dividing file Is read out, and for each group pattern based on each data, Calculate the average of the dissimilarity representing the continuity, calculate the average of all dissimilarities that is the average of all groups, store the group number corresponding to the group pattern number, the average of dissimilarity, and the average of all dissimilarities, and save the group pattern In the case where a group formation with a large or small difference is requested according to the selection criteria in the participant registration file, referring to the group pattern difference file, Select the group pattern number whose degree average is maximum or minimum,
An information exchange method and information including: a group pattern selecting step of storing a selected group pattern in a result output file; and a group displaying step of reading a group pattern stored in the result output file and outputting a group number to each participant. Provided is a computer-readable recording medium recording an exchange program.

According to a second solution of the present invention, an answer file storing answers from participants to a question, a participant registration file storing a participant identifier and the number of persons, the number of persons in a group, and a selection criterion; A group pattern classification file that stores participant identifiers corresponding to pattern numbers and group numbers, a dissimilarity file that stores in advance the relevance between question answer options, a group number corresponding to the group pattern numbers, and an average dissimilarity A group pattern dissimilarity file that stores all dissimilarity averages, a storage unit that stores a result output file that stores the selected group pattern, and an arithmetic unit that reads, writes, or creates each of the files, The arithmetic unit reads out a question corresponding to the learning or discussion theme from the question file and performs the learning or discussion. The answer file is created based on the answer from the participant to the question, and the group pattern division file is created based on the participant registration file. The file and the group pattern division file are read out, a dissimilarity average representing the relevance between the participants in the group is obtained for each group pattern based on each data, and an average of all dissimilarities, which is an average of all groups, is calculated. ,
The group pattern dissimilarity file is created, and if a group formation with a large or small dissimilarity is requested according to the selection criteria in the participant registration file, the group dissimilarity file is referred to and the total dissimilarity is determined. Selecting a group pattern number whose average is maximum or minimum, storing the selected group pattern in the result output file, reading the group pattern stored in the result output file, and outputting the group number to each participant. An information exchange device characterized by the following.

One of the features of the present invention is an information exchange system for a plurality of individuals registered in advance, which allows each individual to answer questions related to the theme of discussion or learning, Choose attributes as choices,
The information processing apparatus further includes group formation means including a formation algorithm for forming the individual into a group based on the information of the answered options, a personal terminal for transmitting and receiving information, and information exchange means between personal terminals in the group.

According to another feature of the present invention, in the above-mentioned composition algorithm, for example, the creator of the discussion sets the degree of association between the options or how to determine the degree of association, and sets the degree of association to group organization. I try to reflect it.

According to another feature of the present invention, there is further provided an information exchange system for a plurality of pre-registered individuals, wherein
A personal terminal for transmitting and receiving information, information exchange means between personal terminals in the group, and decision simulation means based on input of a representative opinion of each group.

According to another feature of the present invention, in the composition algorithm, group composition,
The arrangement and role in the group are determined.

As a result, the learner's knowledge of the learning and the discussion theme is different, and the learner's experience, career,
Even if individuality (individual attributes) related to learning backgrounds such as motivation and hopes are different, group formation taking into account the attributes of the individual will increase the motivation of learners to participate and improve group learning, Discussions can be held.

More specifically, in a management-related course,
In particular, management simulations are often performed in which one class is divided into groups and the two companies compete with each other for profit. When this is systematized as one of the decision-making simulations, the role assignment / arrangement utilizing the learner's attributes is performed, thereby increasing the willingness to compete and leading to efficient learning. In the information exchange according to the present invention, information necessary for group formation of learners is input, a learner group is organized into groups based on the information, and information (opinions) can be exchanged only within each group. At the same time, the group can exchange opinions as needed, thereby enhancing the effects of group learning and discussion.

[0021]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a schematic configuration diagram of an information exchange system according to the present invention. This information exchange system is, for example, a client / server model for performing group learning (discussion) on an appropriate network such as the Internet, and for example, a learner terminal 10 for a learner (participant) of group learning or discussion. And network 2
0, and an information exchange device 100 which is a device for group formation connected via a network 20. The information exchange device 100 is, for example, a teacher monitor terminal 30.
, Storage unit 40, arithmetic unit 50, transmission / reception server 55
And a simulator 60, each of which has a bus 6
5 are connected.

The learner terminal 10 is capable of inputting and outputting voices, characters, screens, and the like. For example, a normal personal computer terminal including an earphone 11, a microphone 12, a display 13, and the like can be applied. In addition, since the group learning is performed by two-way voice communication, the present invention can be applied even at a communication speed of a normal telephone line. In addition, the presence of character data / image information further enhances the learning effect. Therefore, a communication speed of about an integrated digital communication network (ISDN) is preferable. Also, if ADSL, optical fiber communication, etc. can be used, an advanced multimedia system can be constructed, and higher quality learning can be performed.

The teacher monitor terminal 30 further includes, for example, a display 31 for switching and setting a group learning flow for the learner terminal 10.

FIG. 2 is a schematic configuration diagram of the storage unit 40.
The storage unit 40 stores, for example, a question file 41, an answer file 42, a dissimilarity file 43, and an attribute file 44.
, A participant registration file 45, a group pattern classification file 46, and a group pattern difference degree file 47
And a result output file 48.

Here, returning to FIG.
0 executes, for example, an algorithm for performing an appropriate grouping based on data of various files included in the storage unit 40 (details will be described later).

The transmission / reception server 55 is connected to the learner terminal 10 and the teacher monitor terminal 3 via the network 20, for example.
It transmits and receives information between 0 and receives information such as voices, characters and images transmitted by the learner, and further specifies the transmission destination specified by the sender (learner and teacher) and the designation of the teacher The received information is automatically transmitted to another learner or teacher according to the group distribution list that holds the information on the learning mode according to. The learning mode includes a shared mode (question answer mode) used for starting and ending the learning, and a group discussion mode indicating that the group is being discussed. Switch setting.

The simulator 60 is, for example, a connecting means for allowing a group of learners of another group to learn or discuss the representative opinion of each group when learning or discussing a group. The management simulation can be performed based on the transmission data of the learner, and the result can be notified to the learner. Simulator 60
Has an intermediary function that mediates between groups of learners,
According to an instruction from the teacher monitor terminal (teacher console) 30, a connection for a conference call between groups and a connection between a certain participant and a predetermined participant are made.

First, an information exchange apparatus 100 according to the present invention
An example of a data configuration required to perform group organization in will be described. Required files are mainly provided in the storage unit 40, and are written or read by the arithmetic unit 50 as appropriate.

FIG. 4 is an explanatory diagram of the question file 41. The question file 41 is a file that stores question data in accordance with a learning or discussion theme and questions to be asked to a learner when forming a group. The question file 41 includes, for example, a discussion theme (here,
"How does the first half of the 21st century change compared to the second half of the 20th century?"), Three questions on this subject (here q
1. What happens to crime? , Q2. What about unemployment? , Q3. What happens to social security? ), Options for answering these three questions (here, a.
b. Unchanged, c. Etc.) are shown.

FIG. 5A is an explanatory diagram of the answer file 42. The answer file 42 is a file in which the answer of each participant to each question is stored, which is response data in which the answers from the participants are aggregated. In the answer file 42, for example, the question numbers and choices answered by the learners s1 to s4 are respectively shown.

FIG. 5B is an explanatory diagram of the difference degree file 43. The difference degree file 43 includes options a to
A value indicating the relevance between c and a value indicating the degree of difference is stored. The difference file 43 includes, for example, the option a
The difference between b is 0.5, and the difference between the options ac is
1.0, and the difference between the options bc is 0.5.

FIG. 5C is an explanatory diagram of the attribute file 44. This attribute file 44 indicates the attributes of the participants. The attribute file 44 indicates, for example, the degree of willingness to discuss included in the attributes of the individual. The learners s1 and s4 have 1.0, and the learners s2 and s3 have 0.
5 Note that the personal attribute is, for example, data in consideration of the learning background of the learner, and refers to information accompanying the individual learner, such as personality, strength / dissatisfaction, gender, and learning history. The attribute file 44 is, specifically, a personality survey result related to each learner's opinion and willingness to discuss the theme of the discussion, or each learner's opinion and management role related to management issues and management roles. Here are the results of a personality survey. When such attributes are not considered, the attribute file 44 is unnecessary.

FIG. 6 is an explanatory diagram of the participant registration file 45. The participant registration file 45 is a file that stores the participant identifiers and the number of participants of the learner terminal 10 on the network 20, the number of persons in the group, and selection criteria. The participant registration file 45 includes, for example, a participant ID (participant identifier) for identifying a participant in the group learning or the discussion, a participant name corresponding to the participant ID, and the number of participants (here, 30). Name) Number of learners in one group S
n (for example, three), the magnitude of the difference as a selection criterion, and whether or not the attribute is weighted are shown. further,
Appropriate data such as age and gender may be added.

Here, an outline of a group of learners (group learning courses 1 and 2, management game course) obtained by implementing an optimum group formation method described later according to the learning or discussion theme will be described.

In the group learning course 1, a criterion is such that the greater the number of pairs having different opinions within the group, the easier it is to get a trigger for discussion. At this time, the above-mentioned value of the degree of difference is used as a pair value as a basis for calculation.
Then, all the pair values in the group are averaged, and the value is regarded as “good grouping”. Therefore, the average value d _{total of} “good group characteristics” for all groups d _total
Can be requested. That is, if a group is formed such that d _total is maximized, it becomes an optimal group organization. At this time, personal attribute data such as personality survey results indicating the degree of willingness to discuss (attribute file 4
If there is 4), weighting may be performed using that value. Next, in the simulator 60, after performing the optimal group formation, communication control is performed so that information can be exchanged only within each group. In the first screen of the discussion, as a trigger for discussion, the opinions in the group are cross-referenced, and instructions such as "Please ask other people with different opinions to answer in more detail" etc. Is also good.

In the group study course 2, the criterion is that the greater the number of pairs having similar opinions in the group, the easier it is to get a chance for discussion. In addition, the group organization here is the same as the above except that the dissimilarity is changed to the similarity.
This is the same as the group learning course 1, and if a group is formed so that d _total is minimized, it becomes an optimal group formation. In the first screen of the discussion, as a trigger for the discussion, the opinions in the group were cross-referenced. "Everybody is a group of opinions somewhat similar, but please try to exchange opinions so that they can be combined into one opinion. "Or the like.

In the management game course, for example, a group of learners is divided into groups and a management game is performed. In this case, it is necessary to form a group so that roles are assigned in consideration of the learner's attributes. The number of learners in the group may be determined depending on the content of the management game. In a management game, one group corresponds to one company, and these groups compete for profit. In this case, a criterion is used in which a group is formed based on information on the personality of each individual so that the personality and the way of thinking differ as much as possible within one group (here, similar to the group learning course 1,
Grouping is performed using the degree of difference). In addition, if there is a personality survey result such as leadership, a sense of responsibility, or a so-called wife role, the role of the president or vice president is designated based on the personality survey result or discussion within the group.
This information can be stored by providing a storage area in the attribute file 44 or the like. At this time, the content may be referred to on the teacher monitor terminal 30 and the corresponding participant may be selected. Once the management game has begun, exchange information only within the group when planning a strategy within the group. In addition, the president can also be set to exchange information not only within the group but also with other groups in accordance with the commencement of commercial business.

FIG. 3 shows an information exchange apparatus 10 according to the present invention.
It is a flowchart regarding the group formation processing of 0.
First, in step S101, the question read from the question file 41 included in the storage unit 40 is stored in the learner terminal 1
0 is displayed on the display 13. At this time, the display 13 is in the question answering mode.

FIG. 12A is an example of an explanatory diagram showing the learner screen 70 in the question answering mode. Screen 70 is
For example, at the start of learning, the display screen in the question answer mode in which group formation is performed, the question answer top 71,
Frames are laid out (layout) in the middle stage 72 and the bottom stage 73. The top row 71 is assigned to a row for displaying a message from the teacher, and here the meaning of the question is explained. The middle stage 72 is a stage for displaying specific question items and performing an answer input (here, checking the corresponding items), and includes a transmission button 74. Further, the learner automatically sends the answer to the teacher monitor terminal 30 by pressing the send button 74 after the individual answers the answer.

The lowermost row 73 represents a declaration of group formation determined by a group formation algorithm described later (here, A child is group g1) after repeating the above-mentioned questions. On the left side of the screen 70, a small screen 75 indicating that a group is being formed is displayed.
And small screen 7 showing the number of participants (45 in this case)
6 is displayed. The screen 70 can have an appropriate layout as long as necessary items are displayed.

Here, whether or not transmission to the transmission destination specified by the sender (learner or teacher) in the question answering mode will be described. FIG. 12B is an explanatory diagram of the group distribution list 90 in the question answering mode. According to this list, it is determined whether transmission is possible.

The group distribution list 90 includes, for example, items as a source terminal 91, a destination 92, whether transmission is possible 93 and remarks 94 as items. The transmission source terminal 91 includes a teacher monitor terminal (here, T1) 30 and a learner terminal (here, s1) 10. Monitor terminal 3 for teachers
In the case of 0, the information is sent to all learners, such as "Send questions to all", a specific group, "Group configuration notification, etc.", and individual participants "Individual communication". Image information can be transmitted. Thus, the teacher can specify the transmission destination and transmit the answer to the question, his / her own opinion in group learning, and the like. Further, in the learning terminal 10, the information transmission destination is any one of the teacher monitor terminal 30 "reply to the teacher" and all the "self-introduction", and the voice / character / image information is transmitted to the selected transmission destination. be able to. However, if other groups such as “group configuration notification etc.” and simulator 60 “send to simulator” are specified as the transmission destination, transmission may not be possible.

In step S103, the question file 41
Answers to the questions read from
0, and is stored in the answer file 42.

In step S 105, the arithmetic section 50 creates a plurality of group patterns for each participant based on the participant registration file 45, and stores them in the group pattern division file 46.

FIG. 7 shows a group pattern division file 4
FIG. 6 is an explanatory diagram of the sixth embodiment. Group pattern division file
File 46 corresponds to the group pattern number and the group number.
This is a file storing a corresponding participant identifier. glue
The file pattern classification file 46 is, for example, a participant registration file.
Participation in group study or discussion read from file 45
Divide 30 participants into 10 groups (including 3 people)
This shows a group pattern in the case of digit. Specifically
Is a group corresponding to the group pattern classification number C (1).
Loop pattern (group pattern total number Cn)
Group numbers 1 to n_g(N_g: Total number of groups, where
Is n_g= 10), one group (s1, s
2, s3), 2 groups (s4, s5, s6), 3 groups
Group (s7, s8, s9) ... 10 groups (s2
8, s29, s30). As well
The group pattern C (2) includes one group (s
1, s2, s6), 2 groups (s4, s5, s9),
3 groups (s7, s8, s12) ... 10 groups
(S28, s29, s3)
You. Note that the total number Cn of group pattern divisions is
Is₃₀C₃×₂₇C ₃×₂₄C₃...₆C₃By
Is calculated.

In step S107, the arithmetic unit 50 determines whether the answer file 42, the dissimilarity file 43, the attribute file 4
4, the average difference d _total for each group pattern is calculated based on the group pattern classification file 46 and stored in the group pattern difference file 47.

For this purpose, first, the arithmetic unit 50 reads information on the degree of association between options from the difference degree file 43. The degree of association between options is, specifically, the distance (difference) or closeness (similarity) of the content between options.
These values may be arbitrarily determined by the inventor of the group formation, the teacher, the operator of the information exchange apparatus 100 between 0 and 1,
In the case of an option representing a place name, it is determined by its unique attribute. As a method of giving the numerical value, for example, as to the degree of difference, 1 is given between alternatives of a pair whose opinions are completely different, while a value close to 0 is given to a pair whose opinions are quite similar.

Here, the basic principle for calculating the average difference d _total will be described. First, two learners (for example, si, sj, i ≠ j) in the group are selected. As this choice, so choose the two people from the group in the learners the number of Sn in any, there is a choice of _two ways _{S n} C. Learners (si, s) for a certain question (question number q)
j) The degree of difference d ₁ (q, s) between options based on the answer to j)
i, sj) are arbitrarily given by a person (for example, a teacher) who sets a discussion theme between 0 and 1 in the dissimilarity file 43.

Next, the difference d ₂ (si, sj) between the learners (si, sj) is the difference d ₁ (q, si, sj) obtained when the two students answer all the questions q. Average.

The dissimilarity d ₃ (g) in the group is represented by d ₂ (si, s) for all combinations of the _two pairs si, sj in the group.
j) is the average. Further, the average of all the dissimilarity d ₃ (g) over all groups g (average of all dissimilarities) is represented by d
_total

Therefore, d₁(Q, si, sj), d
₂(Si, sj), d₃(G) is defined as follows:
It is. Note that unless otherwise confused,
The relevant addition range is omitted and described. d₁(Q, si, sj) = [0, 1] (ie,
0 ≦ d₁(Q, si, sj) ≦ 1) d₂(Si, sj) = Σd₁(Q, si, sj) / n_q  d₃(G) = Σd₂(Si, sj) / n_p  d_total= Σd₃(G) / n_g

Here, q: question number, si, sj (i ≠
j): learner (participant) number, n _q : total number of questions, n _p =
_Sn C ₂ , Sn: number of learners in a group, g: group number, _ng : total number of groups.

When Sn = 2, 2 in the group
There is only one human pair, but Sn = 3,4,5. . In this case, it can be seen that there are _np pairs of two persons in the group. When the number of learners in the group is 2 (Sn
= 2) is n _p = 1, so d ₂ (si, sj) and d 2
₃ (g) has the same value.

FIG. 8 is an explanatory diagram of the group pattern dissimilarity file 47. The group pattern difference file 47 is a file that stores a group number corresponding to a group pattern number, a difference average, and a total difference average. The group pattern dissimilarity file 47 corresponds to the group pattern division file 46 described above.
In particular, d ₁ (q, si, sj), d calculated for the group pattern number C (1) in the above calculation process
_{2 (si, sj), d} 3 (g), d tota l (1) is specifically shown for the other group pattern number C
Regarding various values calculated for (2) to (n),
Omitted.

First, for one group, which is group number 1,
Two students selected from the included learners (s1, s2, s3)
Learners answered three questions (q1, q2, q3)
In the case of union of each two pairs as shown in the figure
Questions (s1-s3, s1-s2, s2-s3)
q = 1-3 different degree d₁(1, s1, s3), d ₁
(2, s1, s3) d₁(3, s1, s3) and d
₁(1, s1, s2), d ₁(2, s1, s2), d₁
(3, s1, s2) and d₁(1, s2, s3), d₁
(2, s2, s3), d₁(3, s2, s3) is calculated
Is done.

Further, based on the obtained degree of difference for each question, the degree of difference d ₂ (s1, s3), d ₂ (s
1, s2) and d ₂ (s2, s3) are calculated. Further, the difference d ₃ (1) within the group is calculated based on the difference between the learners. Here, group number 2
-10 are subjected to the same calculation process as that of one group having the group number 1, and the dissimilarity d in each group is obtained.
₃ (2) to d ₃ (10) are calculated. Based on the dissimilarity _d 3 ₍₁₎ to d 3 in this group (10), for the group pattern number C (1), the total dissimilarity average d across all groups (10 groups)
_total (1) is calculated.

Here, the above d ₁ (q, si, sj),
The correlation among d ₂ (si, sj), d ₃ (g), and d _total (1) will be described using a schematic diagram.

FIG. 9 shows d ₁ (q, si, sj), d
₂ (si, _sj), is a correlation diagram of _{d 3 (g), d tota} l (1). The correlation 49 is, for example, d ₁ (q, si, sj), d ₂ (si, s) in the group pattern number C (1).
j), d ₃ (g) and d _total (1) are shown. The correlation 49 is, for example, a group number g = 1 to 1
Each of the groups is divided into 10 groups, and each group includes three learners. Specifically, group 1 of group number 1 includes learners s1, s2, and s3, and group 2 of group number 2 includes learners s
4, s5, and s6, and the group 10 includes the learner s
28, s29, and s30, respectively. In addition,
The difference d ₂ (si, sj) between the learners is represented by, for example, d (s1, s2) as indicated by an arrow in the figure.
D _total (1) is shown as an average of d ₃ (1) to d ₃ (10).

In step S109, step S107
All dissimilarity average d _{tota l} obtained by determines whether calculated for all groups pattern. In step S111, when the total difference average d _total for all group patterns is calculated, the calculation unit 50 determines whether a group formation with a large / small difference is set based on the participant registration file 45. I do. Incidentally, if the total dissimilarity average d _total there are group pattern that has not been calculated, again, the process returns to step S107, and calculates the total dissimilarity average d _total per group pattern.

In step S 113, when a group formation having a large degree of difference is set, the arithmetic unit 50 selects a group pattern having the maximum value of the average degree of difference d _total and stores it in the result output file 48. In step S115, when a group formation with a small degree of difference is set, the calculation unit 50 selects a group pattern having the minimum value of the average degree of difference d _total and stores it in the result output file 48.

The arithmetic unit 50 determines whether or not a group formation having a large degree of difference is requested according to the group learning or the discussion theme based on the above-mentioned participant registration file 45. When a group formation with a large difference is requested, the arithmetic unit 50 sets the group pattern number C calculated in the group pattern difference file 47.
Based on the values of d _total (1) to (n) for (1) to (n), a group pattern corresponding to the group pattern number C that maximizes d _total (C) is selected. Further, the result of the selected group formation is recorded in the result output file 48.

When a group formation having a small degree of difference is requested, the arithmetic unit 50 sets the group pattern number C calculated in the group pattern difference degree file 47.
Based on the values of d _total (1) to (n) for (1) to (n), a group pattern corresponding to the group pattern number C that minimizes d _total (C) is selected and stored in the storage unit. The result is stored in a result output file 48 included in 40. When a group pattern is selected, an appropriate rearrangement may be performed, such as sorting in a descending order or an ascending order.

FIG. 10 is an explanatory diagram of the result output file 48. The result output file 48 is a file for storing grouping data holding identification information of the learners of the group formed in the group. The result output file 48 indicates that a group pattern having a maximum value of d _total (C) is obtained when a group composition having a large degree of difference is requested, and a group pattern having a small degree of difference is requested when a group composition having a small degree of difference is requested. The group pattern in which d _total (C) is the minimum value is stored. Here, as an example, d _total (2) is stored in the result output file 48 as a group pattern in which d _total (C) is the maximum value. In addition, in order to execute the optimal group formation according to the theme of appropriate group learning or discussion,
d _Total (C) may be recorded in the result output file 48 with a group pattern other than the maximum or minimum. Furthermore, d
_Total (C) may record both maximum and minimum group patterns.

In step S117, the arithmetic unit 50 stores the participants of the group learning or the discussion in the result output file 48.
_Is read out, and the result of group formation is displayed on the terminals 10 and 30 in accordance with this group pattern.

When the group is formed, the mode can be shifted to the group discussion mode. This mode is switched by the teacher terminal 30. The simulator 60 is connected by the terminal 30 or 10 so that a call can be made between participants in a predetermined group.

FIG. 13A is an explanatory diagram showing the learner screen 80 in the group discussion mode. The screen 80 is, for example, a display screen in a group discussion mode indicating that a group discussion is being performed, and is laid out in an uppermost stage 81, a middle stage 82, and a lowermost stage 83. The top row 81 displays, for example, the remarks of members of the group g1.

The middle section 82 is a case of a member of the group g1, in which opinions are exchanged by voice of three persons A, B, and C of the same group, and the names of the speakers are displayed in the order in which they were made. Character string 82 and a face photograph 84 of the group member. Also, when someone in group g1 makes a decision input for simulation,
The progress is also displayed. In addition, since the face photograph 84 of the group member is posted on the middle section 82, a sense of presence can be obtained even if the user is away from the place.

The lowermost stage 83 shows the score status of each group in the simulation in a graph. For this reason, it is easy to understand the situation of other companies, and it can be expected that willingness to compete further increases. On the left side of the screen, a small screen 85 indicating that the group is under discussion and a small screen 86 indicating the number of times of speech of each group are displayed. For this reason, the teacher can transmit appropriate advice to the inactive group by looking at the statement status of each group.

FIG. 13B is an explanatory diagram showing a group distribution list in the group discussion mode. According to this list, it is determined whether transmission is possible.

The group distribution list 95 includes, for example, information transmission source terminal 96, transmission destination 97, transmission availability 98, and remarks 99 as items. The source terminal 96 includes a teacher monitor terminal (here, T1) 30 and a learner terminal (here, s1) 10. Monitor terminal 3 for teachers
In the case of 0, the destination of the information is any one of “send to all groups”, “send to specific group”, and the individual participant “individual contact” for all the learners. Information can be sent. This allows the teacher to specify the destination, answer questions,
It is possible to transmit their own opinions in group learning. In the learning terminal 10, the transmission destination of the information is the monitor terminal 30 for teachers such as "question to teacher" and the group g1.
Either “intra-group information exchange” or simulator 60 “send to simulator” can transmit voice, character, and image information to the selected destination. However,
As for the transmission destination, communication is not possible except for g1, which is another group, but the simulator 60 may be configured to enable information exchange for negotiation with another group via the intermediary function. As a result, the learner can select any one of the teacher (T1), the group to which the learner belongs, all the members, and the simulator as the information transmission destination, and transmit information such as voice, text, and images.

Hereinafter, in the group learning course 1, for example, four people (s1, s2, s3, s4) in one class, and two people (Sn = 2) in a group, two groups ( _ng =
The case of group formation in 2) will be specifically described.

First, the four learners set the question file 41
Are answered by the answer options a to c for the three questions q1 to q3 stored in. This answer pattern
This is shown in the answer file 42, and the difference d ₁ (q, si, sj) between the options is shown in the difference file 43.

Next, the group pattern division file 4
6, corresponding to the group pattern classification recorded in
For example, (s1, s2) is set as group 1 which is group number 1, and (s3, s4) is set as group 2 (here, the group pattern number is C = a). Here, the difference d ₂ (s1, s) between the learners in group 1
2) is 0 because s1 and s2 are the same option answer. Therefore, d ₃ (1) also becomes 0. Also for group 2, d ₃ (2) is 0. Therefore, d _total (a) also becomes 0. In another grouping pattern C = b, for example, grouping into group 1 (s1, s3) and group 2 (s2, s4) is performed. _{Here, d 3 (1) d 3} (2) , since both the 0.67, d _{to tal} (b) becomes 0.67. Therefore, the grouping pattern C = b is “better grouping” than the grouping pattern C = a.

Here, the grouping pattern C =
b and a group pattern number C = c of knitting group 1 (s1, s4) and group 2 (s2, s3), which are still other grouping patterns (when four learners form two groups, , The cases where grouping is possible are the above three cases), and d _total is 0.
67, and the “goodness of grouping” is about the same.
When the willingness to discuss stored in the attribute file 44 is taken into consideration, for example, when a group having a large variation in the willingness to discuss has an educational policy of “good grouping”, the grouping pattern C = b is "good grouping" than grouping pattern C = c.

Next, good grouping is performed, and within the group, the names of the members of the group and the answer choices selected by each person are made public. Then, an instruction can be given on the screen to ask the opponents with different opinions in the group to further hear the details. It should be noted that the grouping in the group learning course 1 can also be used for exercises in ordinary subjects such as mathematics and science. In other words, a multiple-choice question was conducted, and one class was divided into groups based on the results of the evaluation of correct and incorrect answers so that the number of correct and incorrect combinations increased as much as possible, and the learner of the incorrect question obtained the correct answer. Have the learner teach the process of deriving the answer.

Thus, the wrong learner can listen to the other party until he / she is satisfied with the details, and the positive learner can explain his derivation process to the wrong learner, thereby making his own learning easier. It can also be used to enhance presentation and to learn presentations (eg how to explain things to others in a way that is easy to understand).

Next, a description will be given of a modification of the processing shown in FIG. 3 regarding the processing related to group formation in the above-described group learning course 2 and management game course.

(1) Group Study Course 2 First, in group study course 2, for example, the destination of a virtual school trip is determined for each group.
The final decision making is stored in the participant registration file 45 as a whole. In this case, the above-described step S103
Then, the user inputs his / her desired destination and the like as options and stores it in the answer file 42. In step S105, the information is divided into groups of two persons based on the information and stored in the group pattern division file 46. Specifically, for example, if the desired destinations of s1, s2, s3, and s4 are the United Kingdom, Sapporo, the United States, and Tokyo, respectively, the foreign group 1, s2, and s4 in which s1 and s3 are the so-called group number 1 are group The group is formed as the domestic group 2 with the number 2.

Here, a case where group formation is performed based on the similarity of answers stored in the difference file 43 in step S107 will be described. However, since the similarity-related variables in the following equation correspond to the group pattern dissimilarity file 47, they will be schematically described.

First, r₁(Q, si, sj), r₂(S
i, sj), r₃(G), r_totalAre each selected
Similarity between limbs, similarity between learners, similarity within groups
(Average of similarity) and average of all similarities. Here,
Is described in step S107 in the group learning course 1.
The only difference that was changed to the similarity was the calculation process.
Is omitted. r₁(Q, si, sj) = 1-d₁(Q, si, sj)
 (Ie, 0 ≦ r₁(Q, si, sj) ≦ 1) r₂(Si, sj) = Σr₁(Q, si, sj) / n_q  r₃(G) = Σr₂(Si, sj) / n_p  r_total= Σr₃(G) / n_g  Here, q: question number, si, sj (i ≠ j): learner
(Participant) number, n_q: Total number of questions, n_p= Combination
(Sn, 2), Sn: number of learners in the group, g: glue
Number, n_g: The total number of groups.

In this case, the similarity between the options may be indicated by 4 × 4 cells arranged vertically and horizontally similarly to the difference degree file 43. Also, d ₁ (q, si, sj)
Is 1, if they are matched by domestic and foreign correlation,
Otherwise, it is set to 0. Further, r ₃ (g) and r _total increase if the opinion is a close pair, so “good grouping” can be determined based on the magnitude of the value.

Next, after good grouping is performed,
The participants in the group, for example, discuss the destination and the like in each group, make a decision (representative opinion), and send the representative opinion to the teacher monitor terminal according to the group distribution list included in the simulator 60. 30. In addition, the agent agent agent (for example, a teacher) uses the teacher monitor terminal 30 to present an instruction to the learner terminal 10 to discuss and make a decision between groups. In addition, the outcome of the discussion between the groups is determined. At this time, the display 13 is in the group discussion mode.

(2) Management Game Course In the management game course, for example, two airlines A and B with four learners and two participants in a group (Sn = 2) discuss the seating rate between Hiroshima and Tokyo. Shall compete. In addition, since Sn = 2 in each group, it may be considered that there are only two persons, the president and the vice president. In step S105,
Ask the four learners in advance whether they would like to be the president or the vice president, and make appropriate grouping patterns, such as dividing them into groups to follow the intention as much as possible or dividing them based on the results of personality surveys. To do. For example, in Morseley's personality test, the degree of extraversion and nervousness of the subject is quantified, and can be used.

In this case, in step S107 described above,
The arithmetic unit 50 reads information on the attribute of the learner from the attribute file 44 (for example, data obtained by a personality survey result, and the degree of the personality is quantified by predetermined counting).

Here, a case where attribute weights w (si) and w (sj) are added to a certain learner (si, sj) will be described. However, here, it is the same as the calculation process for the degree of difference described in step S107 in the group learning course 1, and the duplicated portions are omitted.

The difference d between the learners₂(Si, sj),
D for all pairs in the group₂(Si, s
j) the intra-group difference d which is the average of j)₃(G) and phase
Average degree of dissimilarity d₃(G) flat across all groups
Average (average of all dissimilarities) is d_totalIs represented by the following equation. d₂(Si, sj) = Σw (si) w (sj) d
₁(Q, si, sj) / n_q  d₃(G) = Σd₂(Si, sj) / n_p  d_total= Σd₃(G) / n_g

FIG. 11 is an explanatory diagram of the gain matrix 67. This gain matrix 67 shows a simulation close to actual company management. For example, if both Company A and Company B are set at 20,000 yen, each will generate a profit of 4 billion yen per period. However, one of them is 1
If the price is reduced to 5,000 yen, the customer will flow more to that, and the profit will increase significantly to 6 billion yen, while the other company will decrease the profit to 500 million yen. If both are 15,000 yen, they will be 3 billion yen.

Also, this management game is both 20,0
It starts under the condition of 00 yen, but it is free to lower the price to 15,000 yen or return to 20,000 yen again in the subsequent period unit. From the second term onwards, ask the mediator agent for public information, such as the face photo of the other party, personality survey results, etc., as necessary, and consult with the other party's strength and feelings within the group, and think that it is optimal The move is determined, and the representative contacts the intermediary agent. Thereafter, the mediator agent transmits the profit of the period to each representative based on the gain matrix 120. This is repeated many times, and management simulations are performed to maximize profits.

In order to make this management game a more realistic simulation, for example, it is necessary to collect certain necessary expenses in any period, and to enable information exchange between groups through an intermediary agent. In order to prevent collusion, information on prices may not be handled, or a part of assets may be lent to a partner company at a fixed interest rate for a fixed period. If an opponent goes bankrupt on the way, the game with the opponent ends for the time being, but the loan does not return.

Next, another embodiment of the information exchange method according to the present invention will be described. This is especially true for steps S105 and S1 when there are many participants.
This is another embodiment relating to the process for 07.

First, for example, when the number of participants in group learning or discussion is 100 or less, the arithmetic unit 50 expands all cases (total number of group pattern divisions Cn) to determine a combination and is requested. The group pattern division having the average dissimilarity d _total according to the dissimilarity is selected by appropriate rearrangement. On the other hand, in a case where the number of participants exceeds 100, it is assumed that developing all cases requires a considerable amount of time on a normal small computer (the total number Cn of group pattern divisions is all calculated. In order to create the group pattern dissimilarity file 47, an enormous amount of calculation is required).

The grouping algorithm according to the present invention will be described for the case where the number of participants is large. This group formation algorithm (hereinafter, referred to as a sequential determination method: random combination) is a method of determining a sequential combination by grouping a large number of participants for each of a predetermined number of processing units K. An optimal combination is performed based on the degree of difference between the options selected by the participants. Here, as an example, a case where 1,000 participants are organized into a group of 2 to 6 persons will be described.

(1) Adjustment of the number of people First, 1,000 people are given consecutive numbers (1, 2,
3 and so on; Example: A random serial number, such as in ascending order of registration number, or ranking by random numbers assigned to attendees. The basic number of people in the group is Sn. That is, 1
The two groups are basically Sn people (Sn = 2, 3, 4,
5, 6). Sn is determined by the originator of the discussion. For convenience of explanation, the processing unit K is used as a calculation example.
(K means that every K columns from the left end of the table are used as a processing unit for grouping).

Therefore, assuming that the processing unit K is 10, for example, if Sn is 2, 20 people are sequentially subjected to grouping processing from the left (S
If n = 3, refer to the group pattern division file 46). Further, the value of the processing unit K is determined by the number of questions to be applied and the number of options (for example, in the case of three questions and three choices, K =
About 10). Note that the variable K when Sn = 2 can be paraphrased as the number of pairs.

Here, the case of K = 10 will be described. Specifically, the following number of persons is adjusted according to Sn and K so as to form a group of multiples of K in order to be established as a group (details will be described later). 500 when Sn = 2
Make a pair. In the case of Sn = 3, 332 pairs × 3 people + 1 pair × 4
333 pairs of people. In addition, one supermarket is excluded and 999 are set, and 21 dummies are added to make 340 sets.
In the case of Sn = 4, 250 sets are made. If Sn = 5, 2
Make 00 pairs. In the case of Sn = 6, there are 167 sets of 166 sets × 6 persons + 1 set × 4 persons. In addition, add 20 dummy and make 1,020 pairs to make 170 pairs. Note that an average value is set in advance for the dummy value.

(2) Creation of a Row A case will be described in which all the members (1,000 to 1,020 persons) are arranged in the Sn row according to the attribute 1. FIG. 14 (a)
FIG. 4 is an explanatory diagram showing a row for each attribute 1 (two types) created by a sequential determination method. The column 120 indicates, for example, a column in which the value of the attribute 1 is created in two types (sex). As an example, in the case of Sn = 2, if the number of persons does not match, the number is turned back to the shortage and arranged in two stages. If Sn = 3,
1,020 people are randomly arranged according to the composition ratio of attribute 1, and furthermore, 21 dummies are placed at the end of each stage and arranged in three stages. In the case of Sn = 4, 1,000 people are randomly arranged according to the composition ratio of attribute 1, and are arranged in four rows. When Sn = 5, they are arranged in five stages as in the case of Sn = 4. S
In the case of n = 6, 1,020 persons are randomly rearranged according to the composition ratio of attribute 1, and 20 dummy persons are placed at the end of each row, and are arranged in 6 rows.

As a result, Sn stages having the same length based on the attribute 1 of the participant are created. The number of people in each row is a multiple of the processing unit K. FIG. 14B is an explanatory diagram showing a row for each attribute 1 (five types) created by the sequential determination method. The row 130 shows a row in which, for example, the value of the attribute 1 is created in five types (participants' continent addresses).

As an example, when Sn = 2, 1000 people are randomly arranged according to the composition ratio of attributes and arranged in two rows.
In the case of Sn = 3, 1,020 persons are randomly arranged in accordance with the composition ratio of the attributes, and 21 dummy persons are placed at the end of each step, and are arranged in three steps. In the case of Sn = 4, 1000 people are randomly arranged according to the composition ratio of attributes and arranged in four rows. S
In the case of n = 5, when the number of persons does not match, the number is turned back to the insufficient step and arranged in five steps. When Sn = 6, 1,002
People are rearranged at random according to the composition ratio of attributes, and 20 dummy persons are placed at the end of each row and arranged in 6 rows.

Thus, Sn stages having the same length based on the attribute 1 of the participant are created. The number of people in each row is a multiple of the processing unit K.

(3) Calculation of Combination by Option For the stage created as described above,
The determination of the offset will be described. When Sn = 2
Will be described. The first (2 rows x K people =) 20 people in each row
Take out and make a combination of all possibilities, each
The average of all differences d for the possibility of_total(Learner dissimilarity
Average d ₂Average over all 10 groups of (si, sj)
Value). This is, for example, for each trial:
Let's say that it is expressed as a numerical example of. Trial 1 = Male 1: Female 1, Male 2: Female 2, ... d_total= 0.0 Trial 2 = Male 1: Female 2, Male 2: Female 3, ... d_total= 0.67 ··· Trial n = male 1: female 10, male 2: female 9, ... d_total= 0.5 Optimal as the number of pairs of different ones increases (for example,
For example, group study course 1) is d_totalThe larger
Group formation. Conversely, the number of similar pairs is large.
When it is very optimal (for example, group learning course 2)
Is d_totalThe smaller the is, the better the group organization. Obedience
Therefore, the quality of group formation based on the degree of difference is d_totalLarge
It can be thought of as a small one.

Here, a case will be described in which the larger d _total is, the better the group organization is. First, d _total
There are 3,628,800 possible combinations (10!) Of combinations of participants in order to determine the group formation that maximizes, and it is expected that it will take an enormous amount of time to try all. Therefore, in the sequential determination method, K
The portion consisting of human Sn stage, attempts "random combination more than 100 times, if extracts d _total is greater combination of trial results of d _total distribution tries all 2
The combination of participants is calculated based on the assumption that a so-called central limit theorem holds, which is well known in statistics that a good combination of σ or more can be determined.

FIG. 15 is a flowchart for the sequential determination method. However, here, 3 questions for 2 rows x 10 people
This shows a case where multiple choice questions are presented. In step S201, answers obtained by giving three questions and three-choice questions to two rows × 10 persons are taken out, and the stack (storage area) is initialized. Here, a storage area (stack) for 10 candidates for storing the order of K persons in the first row and the K persons in the second row combined therewith and the value of d _total is prepared in advance.

FIG. 16A shows the state at the start of the calculation (initial state).
FIG. 4 is an explanatory diagram of a state of a stack. Stack 2
20 is, for example, divided into groups G1 to G10, d _total of each assumed combination is 0, and the minimum value of d ₂ (si, sj) and attribute 2 are not calculated. In step S203, the number of trials is counted up. In step S205, the upper 1
Zero is fixed, and 10 random numbers are further generated and assigned to the lower 10 persons in the order of generation. In step S207, it is assumed that the lower ten people are rearranged in descending order of random numbers, and that the upper ten people and two people are combined. Step S209
Then, as shown in step S107 and its modification, d _total of the assumed combination and d ₂ (si, s
The minimum value of j) is calculated.

The order of K persons in the first row is fixed, and K random numbers are generated for the K persons in the second row, and each person in the lower row (female i, female i + 1,... ) Is a group, and d _total is calculated for each group at that time.

[0105] At step S211, the d _total of the assumed combination determines whether greater than d _total minimum value of the stack (G10). When d _total is larger than d _total at the bottom of the stack, the order of (female i, female i + 1,...) is stored as a possible combination.
The calculated values of d _total and other combinations (e.g., d ₂ (si, sj), etc.), and the values of the personal attribute 2 and later are also stored in the same row.

In step S 213, if it is larger, the minimum value of the stack and the current d _{total and} others are replaced, and
Sort the rows of the stack in descending order of d _total . Here, the stacks are rearranged in descending order of d _total . It is rejected when the calculated dtotal is smaller than the lowermost d _total stack. In the second trial, similarly, a different random number (a sequence different from the previous one is obtained) is generated,
The comparison processing is performed in the same manner as the previous time. This number of times (about 500 times or more 100 times is preferable.) To obtain a _{d-to tal} larger 10 candidates repeatedly.

In order to determine a combination from the ten candidates, the combination is determined by looking at the minimum value of the difference between learners d ₂ (si, sj). If the value of K is too large for the number of questions or options, d ₂ (si, sj) becomes small, so d
_{Excluded even} if _total is large. When there are a plurality of combinations in which d _total has the same value, a combination that maximizes d _total ′ including the degree of difference according to attribute 2 is selected.

In this case, since the operation is performed on the above-mentioned ten items, the result can be obtained by ten calculations. Also,
If d _total ′ in which attribute 2 is added is the same, a combination that maximizes / minimizes d _total ″ in which the degree of difference due to attribute 3 is added is selected. In this case, the result is obtained by ten calculations. If the attribute 3 is the same, the start of the processing order is selected to determine the group organization.

The next (Sn stage × K people =) 20 people are taken out from each stage, and all combinations are made. The same processing is performed for the second and subsequent times, and the processing is further performed 50 times (500 persons / K
Person) determines all combinations.

Here, the calculation mainly based on the weighted addition / subtraction is 20
Since the processing is performed 500 times × 50 times = 25,000 times per person, the calculation time by a normal small computer is about 10 minutes. This does not feel like waiting time for learners expecting the combination result. Further, if the processing is started when a certain number of answer results (K groups or more) are returned, the processing can be actually performed without waiting time.

In step S215, step S211 is executed.
In, the assumption was a combination of d _total stack d _total
If it is smaller than the minimum value (G10), it is determined whether the number of trials has reached a specified value. FIG. 16B is an explanatory diagram of the state of the stack at the end of the operation.

The stack 230 is divided into, for example, groups G1 to G7 as combination candidates, and each of the groups G1 to G7 includes d _{total of the} assumed combination.
Are arranged in ascending order. Also, groups G1 to G7
, The minimum value of d (s) and attribute 2 are calculated as shown in the figure.

Next, the case where Sn = 3 will be described. However, overlapping description will be omitted. First,
The first (3 rows × K people) 30 persons are taken out from each row to create a combination of all possibilities, and d _total is calculated for each possibility. Trial 1 = Male 1: Male 2: Male 3, d _total = 0.5 Trial 2 = Female 1: Male 2: Male 3, d _total = 0.6 Trial n = Male x: male 2: male 3, ... d _total = 0.4

Here, since it is a three-stage combination calculation, d
_totalThere are three ways to find₃C ₂= 3) Combination
(In trial 1, male 1-male 2, male 2-male 3, male 3-male 1)
It is necessary to find the sum of the differences for each of the. one
The number of trials is 200 for each of the two combinations and 600
The number of times / 30 people. Maximum d_totalFind 10
You. Furthermore, according to exclusion by d (s), attribute 2 and attribute 3,
Selection processing is performed in the same manner. The same applies to the second and subsequent times.
Processing, and 34 unions
To decide.

Further, the calculation mainly of addition / subtraction is performed 600 times / 30 times.
The processing is performed by a general small-sized computer for about 9 minutes because the processing is performed for 20 times 400 times a person × 34 times. This calculation time is
There is no waiting time for learners expecting a combination result. Finally, one super is added to the first group, and only the first group has four members.

The case where Sn = 4 will be described. In the same manner as above, take out 40 persons (4 rows × K persons) and make all combinations. The number of trials is set to 200, and exclusion by ds and selection processing by attribute 2 and attribute 3 are performed in the same manner. 4 people of the combination calculation _{(4 C} 2 = 6) So, 1, 2
00 times / 40 people × 25 times = 30,000 times.

The case where Sn = 5 will be described. In the same manner as above, 50 persons are taken out (5 rows × K persons) and all combinations are made. The number of trials is 100, and d (s)
Exclusion and selection processing by attribute 2 and attribute 3 are performed in the same manner. So five of the combination calculation _{(5 C} 2 = 10),
1,000 times / 50 people × 20 times = 20,000 times.

The case where Sn = 6 will be described. In the same manner as described above, 60 persons are taken out (6 rows × K persons) and all combinations are made. The number of trials is 100, and exclusion by ds and selection processing by attribute 2 and attribute 3 are performed in the same manner. Since six combination calculation _{(6 C 2 = 15),} 1,
500 times / 60 people × 17 times = 25,500 times. Lastly, the last set has four members except for two dummy members from the last set.

According to the present embodiment, even if the number of participants is large, in the case of a group learning course, the groups are organized so that opinions on the theme are as different or close as possible. It is easy to get a chance for group discussion. In the case of a management game course, a simulation close to an actual company organization can be performed by forming a group based on the opinions and attributes of the learner and determining the role assignment. In addition, since the network is used, advanced (university-level) learning can be performed without leaving the area and staying at home. The above information exchange method can be applied to a so-called Internet university.

Further, since the results obtained by various group learning, discussion, simulation, and the like can be accumulated as educational know-how, the burden on specialized educators can be reduced. In addition, since group formation and role sharing can be performed according to the characteristics of the participants, the motivation of the participants to learn increases, and the learning effect can be enhanced.

Note that the information exchange method of the present invention includes a computer-readable recording medium on which an information exchange program is recorded, a program product including the information exchange program and loadable into an internal memory of the computer, and a computer program including the information exchange program. It can be provided by a program or a program product stored in a possible recording medium.

[0122]

According to the present invention, as described above, appropriate group formation is performed for a large number of learners with different backgrounds who are participants on the network, and the effect of group learning (discussion) is enhanced. Can be. Further, according to the present invention, a learner's willingness to learn can be increased, and the burden on a specialized educator can be reduced.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of an information exchange apparatus 100 according to the present invention.

FIG. 2 is a schematic configuration diagram of a storage unit 40.

FIG. 3 is a flowchart illustrating a group formation process of the information exchange apparatus 100 according to the present invention.

FIG. 4 is an explanatory diagram of a question file 41.

FIG. 5 is an explanatory diagram of an answer file 42, a dissimilarity file 43, and an attribute file 44.

FIG. 6 is an explanatory diagram of a participant registration file 45.

FIG. 7 is an explanatory diagram of a group pattern division file.

FIG. 8 is an explanatory diagram of a group pattern dissimilarity file 47.

FIG. 9 shows d ₁ (q, si, sj) and d ₂ (si, s
j), d ₃ (g), d _total (C = 1) correlation diagram.

FIG. 10 is an explanatory diagram of a result output file.

FIG. 11 is an explanatory diagram of a gain matrix 67.

FIG. 12 is an explanatory diagram of a learner screen and a group distribution list in a question answering mode.

FIG. 13 is an explanatory diagram of a learner screen and a group distribution list in the group discussion mode.

FIG. 14 is an explanatory diagram showing a stage for each attribute 1 (2 or 5 types) created by the sequential determination method.

FIG. 15 is a flowchart of a sequential determination method.

FIG. 16 is a diagram illustrating a state of a stack at the start of an operation (initial state) or at the end of an operation.

[Explanation of symbols]

 REFERENCE SIGNS LIST 100 information exchange device 10 learner terminal 20 network 30 teacher monitor terminal 40 storage unit 41 question file 42 answer file 43 dissimilarity file 44 attribute file 45 participant registration file 46 group pattern division file 47 group pattern dissimilarity file 48 result output File 50 operation unit 55 transmission / reception server 60 simulator

Claims (11)

[Claims] Claims: 1. A question corresponding to a learning or discussion theme is read from a question file, and a question presentation step for presenting to a participant in the learning or discussion, and an answer file is created based on an answer from the participant to the question. Based on the answer input step to be performed and the participant registration file storing the participant identifiers and the number of persons, the number of persons in the group, and the selection criteria, create a group pattern division file storing the group pattern numbers and the participant identifiers corresponding to the group numbers. A group pattern dividing step, the answer file, a dissimilarity file preliminarily storing the relevance between the answer choices of the question, and the group pattern dividing file are read, and for each group pattern based on each data, the Averaged dissimilarity, which represents the relevance between participants Calculating an average of all the dissimilarities, storing a group number corresponding to the group pattern number, the dissimilarity average, and the average of the dissimilarities to create a group pattern dissimilarity file; According to the selection criteria in the file, if a group organization with a large or small difference is required,
A group pattern selecting step of referring to the group pattern dissimilarity file, selecting a group pattern number having the maximum or minimum total dissimilarity, and storing the selected group pattern in a result output file; A group display step of reading out the stored group pattern and outputting a group number to each participant. 2. For each participant grouped according to the group number obtained in the group display step, the group pattern stored in the result output file is read, and the participants in the same group learn or read each other. 2. The information exchange method according to claim 1, further comprising an inter-group discussion step enabling discussion. 3. The total dissimilarity average calculating step includes: a total dissimilarity average d._totalIs the difference d between the options₁(Q,
si, sj), difference d between learners₂(Si, sj),
Difference d in group₃(G) and the following equation is used.
The information exchange according to claim 1 or 2, wherein the information exchange is requested.
Method. d₂(Si, sj) = Σd₁(Q, si, sj) / n_q  d₃(G) = Σd₂(Si, sj) / n_p  d_total= Σd₃(G) / n_g  However, question number q, learner (participant) number si, sj
(I ≠ j), total number of questions n_q, N_p=_SnC₂,group
Number of learners Sn, group number g, total number of groups n_gWhen
I do. 4. The total dissimilarity average calculating step further reads an attribute file storing an attribute corresponding to each participant, and further calculates a total dissimilarity average including a weight based on the attribute. 3. The method according to claim 1, wherein
Information exchange method described in any of the above. 5. The step of calculating the average of all dissimilarities, the method comprising: providing a learner (si, sj) with a record in the attribute file;
Weighting the remembered attributes w (si) and w (sj)
Case, the difference d between learners₂(Si, sj) is the choice
The difference d between₁(Q, si, sj), difference between learners
d₂(Si, sj), difference d in group₃(G) and
According to claim 4, which is obtained by using the following equation based on
Information exchange method described. d₂(Si, sj) = Σw (si) w (sj) d
₁(Q, si, sj) / n_q  d₃(G) = Σd₂(Si, sj) / n_p  d_total= Σd₃(G) / n_g  However, question number q, learner (participant) number si, sj
(I ≠ j), total number of questions n_q, N_p=_SnC₂,group
Number of learners Sn, group number g, total number of groups n_gWhen
I do. 6. The total dissimilarity average calculating step uses a similarity instead of the dissimilarity, and calculates a total similarity average r._total
Is the difference d between the options₁(Q, si, sj), learner
Similarity r between₂(Si, sj), similarity r in group
₃(G) based on the following equation.
The information exchange method according to claim 1. r₁(Q, si, sj) = 1-d₁(Q, si, sj) r₂(Si, sj) = Σr₁(Q, si, sj) / n_q  r₃(G) = Σr₂(Si, sj) / n_p  r_total= Σr₃(G) / n_g  However, question number q, learner (participant) number si, sj
(I ≠ j), total number of questions n_q, N_p=_SnC₂,group
Number of learners Sn, group number g, total number of groups n_gWhen
I do. 7. The group pattern dividing step extracts one group pattern, and the total dissimilarity average calculating step obtains a total dissimilarity average, and adds a group pattern and a total dissimilarity average to a stack file. Compare the average or the initial value of all the dissimilarities of the stored patterns with the one calculated this time, and store only the predetermined number of upper patterns corresponding to the selection criteria stored in the participant registration file in the stack file. The information exchange method according to any one of claims 1 to 6, wherein a predetermined number of group patterns are extracted in advance to calculate an average of all dissimilarities. 8. The information exchange method according to claim 7, wherein only a participant of a predetermined processing unit is replaced to form a new group pattern. 9. A question file corresponding to a learning or discussion theme is read from a question file, and a question presentation step for presenting to a participant in the learning or discussion, and an answer file is created based on an answer from the participant to the question. Based on the answer input step to be performed and the participant registration file storing the participant identifiers and the number of persons, the number of persons in the group, and the selection criteria, create a group pattern division file storing the group pattern numbers and the participant identifiers corresponding to the group numbers. A group pattern dividing step, the answer file, a dissimilarity file preliminarily storing the relevance between the answer choices of the question, and the group pattern dividing file are read, and each group pattern is read for each group pattern based on each data. Averaged dissimilarity, which represents the relevance between participants Calculating an average of all dissimilarities, storing a group number corresponding to the group pattern number, an average of dissimilarities, and an average of all dissimilarities to create a group pattern dissimilarity file; According to the selection criteria in the file, if a group organization with a large or small difference is required,
A group pattern selecting step of referring to the group pattern dissimilarity file, selecting a group pattern number having the maximum or minimum total dissimilarity, and storing the selected group pattern in a result output file; A computer-readable recording medium recording an information exchange program including: a group display step of reading out a stored group pattern and outputting a group number to each participant. 10. An answer file storing answers from participants to a question, a participant registration file storing participant identifiers and the number of persons, the number of persons in a group, and selection criteria, and a participant corresponding to a group pattern number and a group number. Group pattern file that stores the user identifiers, the difference file that stores the relevance between the alternatives to the questions, and the group that stores the group numbers corresponding to the group pattern numbers, the average of the differences, and the average of all the differences A storage unit for storing a pattern dissimilarity file and a result output file storing the selected group pattern; and an operation unit for reading, writing, or creating each of the files, wherein the operation unit is adapted to a learning or discussion theme. Read the corresponding question from the question file and present it to the participants of the study or discussion. The answer file is created based on the answer from the participant with respect to, the group pattern dividing file is created based on the participant registration file, the answer file, the dissimilarity file, and the group pattern dividing file are created. Read out, for each group pattern based on each data, obtain a dissimilarity average representing the relevance between participants in the group, calculate an average of all dissimilarities that is an average of all groups, and read the group pattern dissimilarity file. Create, according to the selection criteria in the participant registration file, if group formation with a large or small difference is required,
With reference to the group pattern dissimilarity file, a group pattern number having the maximum or minimum total dissimilarity is selected, the selected group pattern is stored in the result output file, and the group stored in the result output file is selected. An information exchange device for reading a pattern and outputting a group number to each participant. 11. The information according to claim 10, further comprising a simulator for reading the group pattern stored in the result output file and connecting participants in the same group so as to be able to learn or discuss. method of exchange.