Training support system, training support method, and training support program

Abstract

To provide a training support system that can eliminate complexity of shared tasks in discussions during training and promote activation in the discussions.SOLUTION: A training support system 1 includes an information processing device 10 and a database DB. The information processing device 10 comprises: a receiving unit 101 that receives an opinion considered by a trainee regarding a theme of the training from a trainee terminal 2; a transmitting unit 102 that transmits a result obtained by performing each processing to the outside; an identification unit 103 that identifies a group of the trainee who considered the received opinion; an aggregation processing unit 104 that performs processing of aggregating the opinions of the trainees in a specific unit; a summarization processing unit 105 that performs summarization processing by using a machine learning model for each aggregation unit; a discussion reflection unit 106 that reflects the results of discussions within the group by the trainees on the basis of the summarization result on the corresponding summarization result; and a display processing unit 107 that performs display processing of data to be output to an external output unit 55.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a training support system, a training support method, and a training support program for supporting training by making discussions in group work more efficient and encouraging more active discussions. [Background technology]

[0002] Conventionally, there are known training support systems that improve the efficiency of discussions in group work in which multiple people discuss a specific topic and help to stimulate the discussion. For example, the education support system described in Patent Document 1 (hereinafter, referred to as the "conventional system") uses a robot to provide appropriate advice and guidance according to the situation of the group work discussion. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 6637917 Summary of the Invention [Problem to be solved by the invention]

[0004] According to the conventional system, debaters can receive advice from outside the group to optimize the discussion, which allows them to have a lively discussion. However, in the conventional system, debaters must first share their individual opinions with others in sequence (hereinafter referred to as "sharing work"). Therefore, in conventional discussions, debaters must spend time and effort trying to understand the opinions of others, which hinders the essential activity of discussion, which is to state their own views and discuss the shared opinions.

[0005] The present invention has been made to address the above problems. That is, an object of the present invention is to provide a training support system (hereinafter, also referred to as "the present training system." The same applies to the present training support method and the present training support program.) that can eliminate the cumbersomeness of sharing work and promote active discussion. [Means for solving the problem]

[0006] In order to solve the above problems, the present invention provides A training support system that supports a training including an individual work in which a trainee taking the training summarizes his / her personal opinion on a theme of the training, and a group work in which the trainees discuss in a group consisting of a plurality of the trainees based on the opinions of each of the trainees, The training support system includes: a summarizing unit that summarises the opinions of at least two or more of the students who make up the group; and a summarization processing unit that summarizes the collected opinions using a machine learning model.

[0007] In order to solve the above problems, the present invention provides a method for producing a liquid crystal display comprising: A training support method implemented by a training support system that supports a training including an individual work in which a trainee taking the training summarizes his / her personal opinion on a theme of the training, and a group work in which a group of a plurality of the trainees discusses the training based on the opinions of each of the trainees, a summarizing step of summarizing the opinions of at least two or more of the students who make up the group; and a summarization step of summarizing the aggregated opinions using a machine learning model.

[0008] In order to solve the above problems, the present invention provides a method for producing a liquid crystal display comprising: A training support program that causes a computer device to execute support for the training, the training including an individual work in which a trainee taking the training summarizes his / her personal opinion on a theme of the training, and a group work in which a group of a plurality of the trainees discusses the training based on the opinion of each of the trainees, The computer device includes: a summarizing step of summarizing the opinions of at least two or more of the students who make up the group; The method further comprises executing a summarization process step of summarizing the collected opinions using a machine learning model.

[0009] This training support system, this training support method, and this training support program (hereinafter referred to as "this training system, etc.") are used for training that includes individual work in which trainees think individually about a theme, and group work in which a group of multiple trainees discuss a theme based on the opinions of each member of the group.

[0010] By configuring in this way, the training system etc. can share within the group the results of summarizing and summarizing the opinions of individuals generated in the individual work. Therefore, the trainees can understand the results of summarizing the opinions and quickly complete the sharing task without having to share each opinion with other group members one by one. Therefore, the time and effort that trainees spent on sharing tasks in conventional discussions is reduced, and trainees can focus on essential discussions on the topic. In addition, this training system etc. allows participants to deepen their discussions through substantive discussions, which can increase participants' sense of satisfaction, accomplishment, understanding, and other feelings of excitement about the discussions, making it easier for them to remember what they learned through the discussions.

[0011] In a preferred embodiment of the present invention, the training support system further comprises: a storage unit for storing in advance student identification information for identifying the student and group identification information for identifying a group to which the student belongs; a receiving unit for receiving, from a student terminal used by the student, opinion information including the student's opinion and the student identification information of the student; The information processing device is characterized by comprising a group identification unit that identifies the group to which the student belongs by comparing the group identification information with the student identification information.

[0012] With this configuration, the training system can read out the opinion information received by the receiving unit and the student identification information and group identification information stored in the storage unit, and identify the group related to each opinion by the group identification unit. Therefore, the training system can identify the group for each student's opinion using a computer, and can smoothly separate the opinions of each student into groups without the need for manual collection of opinions and sorting into groups.

[0013] In a preferred embodiment of the present invention, the training support system further comprises: The present invention is characterized in that it further comprises a discussion reflection unit that reflects the results of the discussion among the participants based on the summary result obtained by summarizing the collected opinions.

[0014] By configuring it in this way, the trainees can carry out group work based on the summarized opinions, and then check the results (the results of the discussion) reflected in the summarized opinions. Therefore, the trainees can recognize how the opinions of the members have changed through group work, from their opinions before the discussion to their opinions after the discussion. As a result, this training system can more easily allow trainees to recognize the degree to which the discussion has deepened and developed, and can further increase the trainees' excitement about the discussion, efficiently encouraging more active discussion.

[0015] In a preferred embodiment of the present invention, the summary processing unit The opinions in the individual work are summarized using the machine learning model.

[0016] With this configuration, trainees can check the summary results of their own personal opinions before the group work discussion. This allows trainees to easily organize their thoughts before the discussion. Trainees can also check the summary results to see how the machine learning model understood their opinions. This training system can therefore improve trainees' objectivity regarding their personal opinions, and can efficiently stimulate discussion by having trainees discuss based on the summary results.

[0017] In a preferred embodiment of the present invention, the summary processing unit The collected opinions are summarized using the machine learning model with a view to using them to evaluate the group work.

[0018] With this configuration, the present invention can evaluate group work using a specific viewpoint, and by machine learning the evaluation, predict the relationship between the evaluation of the group work and the viewpoint using a machine learning model. Therefore, the present training system can predict a summarization technique that is likely to result in a high evaluation of the group work by machine learning, and can summarize personal opinions based on this technique. Therefore, the present training system can have participants discuss based on the results of a summary process that is likely to encourage discussion that can be highly evaluated, that is, lively discussion.

[0019] In a preferred embodiment of the present invention, the discussion reflection unit includes: Further opinions that the student has further considered with respect to the answer that is the result of summarizing the opinions in the individual work are accepted from the student terminal, and the further opinions are reflected in the answer.

[0020] With this configuration, the trainees can check the results of the summary in their individual work, reconsider the contents, and reflect their thoughts in the summary results. Therefore, this training system gives trainees an opportunity to reconsider their personal opinions in their individual work and add or revise the contents of the summary, allowing them to put together more refined personal opinions before the discussion. Therefore, this training system can perform summary processing for group work based on more refined opinions, and can more efficiently stimulate discussion.

[0021] In a preferred embodiment of the present invention, the discussion reflection unit includes: The present invention is characterized in that a process is carried out in which the results of the discussion are immediately reflected in the summary result in response to operations of at least two or more of the students constituting the group in the group work.

[0022] With this configuration, the participants can reflect the content of the discussion that took place after the summary results in real time in the summary results. Therefore, this training system allows the participants to discuss in groups while reflecting the opinions that arise in the discussion in real time in the summary results, allowing the participants to discuss while sorting or confirming the content of the discussion one by one. Therefore, this training system allows the participants to discuss while sorting their thoughts, and can efficiently stimulate discussion.

[0023] In a preferred embodiment of the present invention, the training support system further comprises: The display processing unit is characterized by having a display processing unit that displays the summary result and a discussion reflection result in which the result of the discussion is reflected in the same area.

[0024] By configuring it in this way, the training system displays the results of summaries that have been consolidated and the opinions that reflect the results of the discussion in the same area, so the contents of both can be confirmed at a glance. Therefore, the training system allows trainees to easily check at a glance the changes in the thoughts of trainees in their group and the progress of the discussion. This allows trainees to quickly advance discussions and efficiently encourage lively discussions.

[0025] In a preferred embodiment of the present invention, the training support system further comprises: The training system is characterized in that it has a display processing unit that displays in the same area either the opinions, answers which are the results of summarizing the opinions in the individual work, opinion reflection results in which further opinions of the participants regarding the answers in the individual work are reflected in the answers, the summary results, or discussion reflection results in which the results of the discussion are reflected, or a combination of these, and the information to be communicated to the participants by the instructor of the training.

[0026] With this configuration, the trainees can check the contents of each of the opinions and the messages from the instructor at a glance, since they are displayed in the same area. Therefore, trainees can consider the topic in individual or group work without having to spend additional time or effort listening to the messages from the instructor. Furthermore, by displaying the messages, this training system can reduce interruptions to the trainees' work due to the instructor's speech, and trainees can check the displayed messages at their own convenience. Therefore, this training system can prevent trainees' thinking from being interrupted and promote efficient progress or activation of discussion. Effect of the Invention

[0027] According to the present invention, a novel technology related to a training support system can be provided by aggregating the opinions of multiple participants in a training session and summarizing them using a machine learning model. [Brief description of the drawings]

[0028] [Figure 1] 1 is a block diagram illustrating an outline of a system configuration according to an embodiment of the present invention. [Diagram 2] 1 is a schematic diagram illustrating an example of a hardware configuration of an information processing device and a terminal according to an embodiment of the present invention. [Diagram 3] 1 is a flowchart showing a flow of training in which the training support system according to one embodiment of the present invention is used. [Figure 4] 4 is a flowchart showing a process flow of the training support system according to one embodiment of the present invention. [Diagram 5] FIG. 2 is a diagram showing an example of each piece of information used in the training support system according to one embodiment of the present invention. [Figure 6] FIG. 2 is a diagram showing an example of each piece of information used in the training support system according to one embodiment of the present invention. [Figure 7] FIG. 2 is a diagram showing an example of a screen displayed in the training support system according to one embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0029] Hereinafter, an embodiment of a training support system, a training support method, and a training support program according to the present invention will be described with reference to the drawings.

[0030] In this embodiment, the configuration and operation of the training support system will be described, but the same effects can be obtained with respect to the executed method (steps), device, computer program, etc. Also, the program in this embodiment may be provided as a non-transitory recording medium readable by a computer (computer device), or may be provided so as to be downloadable from an external server.

[0031] Furthermore, in this embodiment, the term "unit" may include a combination of hardware resources implemented by circuits in a broad sense, and software information processing that can be specifically realized by these hardware resources.

[0032] In this embodiment, information is represented by, for example, the physical value of a signal representing voltage or current, the high or low value of a signal as a collection of binary bits consisting of 0 or 1, or quantum superposition, so-called quantum bits, and communication and calculations can be performed on a circuit in a broad sense.

[0033] The broad definition of a circuit is a circuit realized by appropriately combining a circuit, circuitry, a processor, and a memory, etc. In other words, it includes a central processing unit (CPU), a graphics processing unit (GPU), a large scale integration (LSI), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), etc.

[0034] <About the training> Generally, depending on the format of training, it can be divided into face-to-face training where the instructor and the trainees receive instruction from the instructor, and online training where the instructor connects a personal computer, smartphone, or other mobile device to a network and the trainees are connected online. Training can also be divided into hierarchical training, which classifies trainees according to the organizational hierarchy and aims to have trainees acquire the mindset, knowledge, and skills required for trainees belonging to each hierarchy, and purpose-specific training, which sets a theme based on a specific purpose and conducts training for those related to that purpose.

[0035] In addition, the training methods used include individual work, in which participants work individually on a theme assigned to them during the training, role-playing, in which participants learn ways of thinking, knowledge, and skills by setting specific scenarios and playing set roles, and group work, in which participants are divided into groups made up of multiple people to discuss, solve problems, and make presentations on a theme assigned during the training.

[0036] Here, in this specification, individual work refers to the work of a trainee in which the trainee puts together his or her personal opinions on the training theme. Therefore, in individual work, the trainee only needs to think about the training theme personally, and even if the trainee is given this work as homework before the training and there is no time set aside for individual work on the day of the training, it can be said that individual work is being done. Also, in this specification, group work refers to the work of a group of multiple trainees in which trainees discuss the topic based on each trainee's personal opinion.

[0037] Therefore, even if there is no time set aside for individual work during training and only group work is conducted, if each participant considers and communicates their personal opinions to others during group work, and also expresses their views on the opinions of others and holds a discussion, it can be said that both individual and group work are taking place.

[0038] In this regard, the training in this specification may be any course that includes the individual work and group work, regardless of the implementation format or the rank of the participants.

[0039] Training does not only refer to education aimed at improving necessary professional knowledge and skills; it also includes internships where job-seekers or people seeking to improve their practical skills gain work experience for a certain period of time at a company, organization or other organisation, as well as workshops where participants discuss and share ideas about specific issues.

[0040] <System configuration> 1 is used in training attended by a plurality of people. In this embodiment, the training is assumed to be for new members of society who will be newly employed in the workplace. As shown in FIG. 1, a training support system 1 includes an information processing device 10 and a database DB.

[0041] The information processing device 10 is a server managed by a provider of the training support system 1, and provides the results of executing the training support system 1 to those who use the training, such as trainees and instructors. A general-purpose server computer or a personal computer (computer device) can be adopted as the information processing device 10, and the training support system 1 can also be configured by implementing the functional components described below in multiple computers. The information processing device 10 is also connected to multiple trainee terminals 2(a)-2(c) (hereinafter referred to as "trainee terminals 2"), an instructor terminal 3, and a person in charge terminal 4 via a network NW so that they can communicate with each other.

[0042] The student terminal 2 is a terminal (computer device) operated by a student receiving training. Specifically, a stationary personal computer, a portable personal computer, a smartphone, a tablet terminal, a wearable device, or other mobile terminal device can be adopted. In this embodiment, it is assumed that the training system operates by a web browser application of a terminal 50 including the student terminal 2, but the operation of the training system may be realized by installing a dedicated application for implementing the training system, etc., in the terminal 50.

[0043] The instructor terminal 3 is a terminal operated by an instructor who instructs the trainees in the training. The person in charge terminal 4 is a terminal used by a person in charge who has requested the instructor to instruct the trainees. The instructor terminal 3 and the person in charge terminal 4 can be the same device as the trainee terminal 2.

[0044] The training support system 1 can be connected to any number of student terminals 2, instructor terminals 3, and staff terminals 4, or may be connected to only student terminals 2.

[0045] In this embodiment, the network NW is an IP (Internet Protocol) network, but there is no restriction on the type of communication protocol, and there is also no restriction on the type and scale of the network.

[0046] The database DB stores various types of information for executing the training support system 1. Detailed information to be stored will be explained in the functional configuration described later.

[0047] <Hardware configuration> Fig. 2(a) is a diagram showing an example of a hardware configuration of the information processing device 10. As shown in Fig. 2, the information processing device 10 includes a control unit 11, a storage unit 12, and a communication unit 13. However, in addition to the above configuration, the information processing device 10 may also include functional units that perform various functions provided in a known computer, such as an audio input device and an audio output device.

[0048] The control unit 11 includes one or more processors such as a CPU or an MPU, and controls the overall operation and processing of the information processing device 10 by executing the training support system according to the present invention, the OS, and other applications.

[0049] The memory unit 12 is a storage device such as an HDD or optical disk, or a semiconductor memory element such as an SSD, ROM, or RAM, and stores various data such as the training support program of the present invention and data and software used when the control unit 11 executes processing based on the program.

[0050] The communication unit 13 is an interface for connecting the information processing device 10 to a network NW. Specifically, a LAN (Local Area Network) or a WAN (Wide Area Network) can be adopted.

[0051] 2(b) is a diagram showing an example of a hardware configuration of a terminal 50 (the student terminal 2, the lecturer terminal 3, and the person in charge terminal 4 in FIG. 1). The terminal 50 includes a control unit 51, a storage unit 52, a communication unit 53, an input unit 54, and an output unit 55.

[0052] The control unit 51 of the terminal 50 includes one or more processors such as a CPU or an MPU, and controls the overall operation and processing of the terminal 50 .

[0053] The memory unit 52 of the terminal 50 is a storage device such as an HDD or an optical disk, or a semiconductor memory element such as an SSD, ROM, or RAM, and stores the web browser application and data used when the control unit 51 executes processing based on a program.

[0054] The communication unit 53 of the terminal 50 controls communication with the network NW.

[0055] The input unit 54 of the terminal 50 is an input device such as a keyboard, a mouse, a touch panel, an OCR (Optical Character Reader), or a microphone, and inputs an operation request from a user to the control unit 51.

[0056] The output unit 55 of the terminal 50 is an output device such as a display, a projector, or a speaker, and displays the results of processing by the control unit 51, etc.

[0057] <Functional configuration> 2(a), the information processing device 10 includes, as its functional configuration, a receiving unit 101, a transmitting unit 102, a group identifying unit 103, an aggregation processing unit 104, a summary processing unit 105, a discussion reflection unit 106, and a display processing unit 107. These are information processing performed by software stored in the storage unit 12, which is specifically realized by hardware such as the control unit 11 and the storage unit 12.

[0058] The receiving unit 101 receives various types of information from the outside. The receiving unit 101 in this embodiment receives personal opinions (hereinafter, referred to as “personal opinions”) and student identification information from the student terminal 2.

[0059] The transmission unit 102 transmits information necessary for the operation of this training system. The transmission unit 102 in this embodiment transmits to the storage unit 12 the results of summarizing the individual opinions.

[0060] The group identification unit 103 identifies which group the student who has thought of the individual opinion belongs to. In this embodiment, the group identification unit 103 collates the student identification information included in the received opinion information with the student identification information linked to the group identification information to identify the group to which the student who has thought of the individual opinion belongs.

[0061] The aggregation processing unit 104 aggregates the individual opinions of each of the students. In this embodiment, the aggregation processing unit 104 aggregates the individual opinions of at least two students who make up a group as an aggregation unit.

[0062] The summary processing unit 105 performs a process of summarizing the individual opinions. In this embodiment, the summary processing unit 105 performs a process of summarizing the individual opinions aggregated for each aggregation unit, using a trained machine learning model read out from the storage unit 12.

[0063] The discussion reflection unit 106 performs a process of reflecting the results of the discussion among the students (hereinafter referred to as "discussion results") in the results of the summary processing (hereinafter referred to as "summary results"). In this embodiment, the discussion reflection unit 106 performs a reflection process of reflecting the results of the discussion that the students discussed in the group based on the summary results in the corresponding summary results.

[0064] The display processing unit 107 displays information required when using this training system, such as individual opinions and summary results. In this embodiment, the display processing unit 107 displays a group work screen (hereinafter referred to as a "GW screen") based on the summary result and the discussion reflection result obtained by reflecting the discussion result in the corresponding summary result, and transmits the result to the student terminal 2.

[0065] <Database> The database DB in Figure 1 stores student information about the students shown in Figure 5(a), group information about the group to which the students belong shown in Figure 5(b), organization information about the organization such as a company or school to which the students belong shown in Figure 5(c), member information for matching the students with members of the organization shown in Figure 5(d) (e.g., information about employees or information about students), information about the instructors shown in Figure 6(a) (hereinafter referred to as "instructor information"), and information about training shown in Figure 6(b) (hereinafter referred to as "training information").

[0066] The database DB may also store the following information (not shown): information on the training text, participant identification information for identifying participants, group identification information for identifying a group to which the participants belong, information on the person in charge (hereinafter referred to as "person in charge information"), information on individual opinions, information on further opinions, information on the identification result, information on the aggregation unit (hereinafter referred to as "aggregation unit information"), the results of summarizing each individual opinion (hereinafter referred to as "answers") and information on the summary results, information on the discussion results, the results of reflecting further opinions in the answers (hereinafter referred to as "opinion reflection results") and information on the discussion reflection results, information on the viewpoint for evaluating group work, information on the evaluation of the group work discussions, information on the machine learning model, information on the results of display processing, and other information such as data and programs required for this training system, etc. are stored.

[0067] A part or all of the various information may be stored in the memory unit 12 or the memory unit 52 of the student terminal 2, the lecturer terminal 3, or the person in charge terminal 4, or a part of the information may be stored in a separate storage device.

[0068] <Training flow> The flow of the training in this embodiment will be described with reference to FIG. The training in this embodiment will be described assuming that it is conducted face-to-face. When the training begins, in step 601 (hereinafter referred to as S601; similarly, other steps are referred to by reference numbers only), the instructor explains the contents of the training to the trainees. In this embodiment, the instructor informs the trainees of the flow and objectives of the training, the contents of the tasks to be performed in the training, the time allocation for each task, important points to note, and the instructor's self-introduction. Next, in S602, the instructor introduces a theme that he wants the trainees to think about in the training, and then specifically explains the contents of the theme (S603) and gives the trainees time to do individual work (S604).

[0069] In the individual work in S604, each student thinks about the theme and inputs their own personal opinion into the input unit 54 of the student terminal 2. Although not shown, the training system can display the results (answers) of summarizing each of the input individual opinions on the output unit 55. Furthermore, the training system can also display on the output unit 55 the contents (opinion reflection results) of the answers that reflect further opinions (further opinions) that the student has thought about in the individual work based on the results (answers) of the summarization process.

[0070] Next, in S605, the instructor instructs the students to split into groups, and the students perform group work on the theme in each group. In the group work in S605, the results of summarizing the individual opinions entered by each student in the individual work are displayed on the output unit 55 of the student terminal 2, and the students express their opinions and hold discussions based on the results. Although not shown, this training system can also display on the output unit 55 the contents (discussion reflection results) of the summary results that reflect the opinions (discussion results) that have emerged as a result of the discussion during the group work.

[0071] Next, in S606, the participants present the results of their group work discussions for each group. After that, in S607, the instructor provides an explanation of the theme. In this embodiment, the training repeats the above flow until the end of the training (S608) or until all the themes prepared by the instructor for the training have been completed. However, the form of the training is not limited to the above embodiment, and multiple individual work sessions may be repeated before the group work, or individual work may be performed again after the group work. Also, the overall presentation may be omitted as appropriate.

[0072] The flow of training by the training support system in this embodiment is roughly as described above. Specific processing steps of the training support system in this embodiment will be described in detail below with reference to FIGS.

[0073] <Various registrations> In S701 of FIG. 4, users such as trainees, instructors, personnel in charge, and those managing the training system who wish to use the training system register trainee information in advance in the database DB.

[0074] As shown in Fig. 5(a), the student information includes information such as the student's name and contact information, and is managed by the student ID. The student information may also include a number for identifying the student's terminal (student terminal ID), organizational information about the organization to which the student belongs (company ID, described later), member information for matching the student with members of the organization (employee ID, described later), information about the job, information about the results of tests administered to the student, such as ability tests and personality tests, registration date, update date, joining date, language, nickname, gender, personality, face photo, birthday, and other information.

[0075] The ability test and personality test are not limited to tests conducted in the organization to which the trainee belongs, and may be any test that diagnoses the trainee's ability, personality, or character. Specifically, the abilities investigated in the ability test refer to the talents and academic abilities required for the trainee to carry out their work, such as the ability to smoothly communicate, the ability to ask questions, the ability to listen, the ability to think logically, the ability to understand, the ability to lead others, the ability to control stress, the ability to explain, the ability to effectively conduct sales, the ability to adjust opinions, and the ability to express themselves in writing. In this manner, by registering information that can determine the abilities and personalities of trainees, the machine learning model used in this training system can analyze and learn about the tendency of trainees' thinking on the same training topic in relation to the trainees' abilities, etc. Furthermore, by analyzing the abilities, etc. of trainees belonging to a group, the machine learning model can also learn about the tendency of thinking in each group (hereinafter referred to as "group personality"). Therefore, with the above configuration, this training system can predict and implement a method of summarization that is likely to stimulate discussion, taking into account the influence of the abilities, etc. of trainees and the group's personality on the trainees' opinions.

[0076] Examples of participants include people new to the workforce, young staff with three years or less work experience, staff from each department of an organization, a new boss who has had subordinates for the first time, a new executive or manager who has taken on a position for the first time, and other working adults who belong to an organization and work for that organization, contract employees who have an open-ended labor contract with their employer, temporary workers, part-time workers, casual workers, and other non-regular workers, students attending educational institutions such as high school or university, and volunteers who engage in community activities unpaid.

[0077] In addition, in S702, the user registers group information in advance in the database DB.

[0078] As shown in Fig. 5(b), the group information includes information such as the number of people and the IDs of the participants in the group, and is managed by the group ID. It can also include information such as the company ID, group name, and participant role names. Examples of role names include group representative, facilitator, presenter, secretary, and timekeeper.

[0079] Furthermore, in S703, the user registers information such as organization information, member information, instructor information, training information, and aggregation unit information in advance in the database DB.

[0080] As shown in Fig. 5(c), the organizational information includes information such as the organization name (company name) such as A Co., Ltd. or B Industry, and the organization's contact information (company contact information), and is managed by company ID. It can also include the organization's industry or activity type, organization's size, organization's logo, and organization's address. Examples of the size of a company include the number of employees, amount of capital, sales, amount of profit such as operating profit and net profit, total assets, and ranking in the field of activity. In this way, by configuring the system to register the industry, activity, and scale of an organization, the tendency of the opinions of trainees on the same training topic can be analyzed for each organization and machine-learned. Therefore, with the above configuration, the training system can predict and implement a method of summarization that is likely to stimulate discussion, taking into account the influence of the characteristics of the organization on the opinions of trainees.

[0081] As shown in FIG. 5(d), member information includes information such as name and company ID, and is managed by employee ID. Member information can also include information such as gender, the location (department) of the role to which the member is assigned, the content of the work the member engages in, the period of affiliation, and the location of the work. If the person is a student, the content of the work the member engages in can also include the content of the academic subject in which the person mainly specializes and the academic field in which the person excels. Furthermore, employee IDs can be used in place of student IDs.

[0082] As shown in Fig. 6(a), the instructor information includes the instructor's name, contact information, etc., and is managed by the instructor ID. The instructor information can also include information such as gender, the level of the trainees, the training field classified by training theme that the instructor has been in charge of most (training field), and the training method that the instructor is good at. In this way, by configuring the system to register the gender and training field of the instructor, it is possible to perform machine learning by analyzing the relationship between the method of proceeding with the training based on the instructor's attributes and past experience and the tendency of the opinions of the trainees on the same training topic. Therefore, with the above configuration, the training system can predict and implement a summary processing method that is likely to stimulate discussion, taking into account the influence of the instructor's attributes and experience on the trainees' opinions.

[0083] As shown in Fig. 6(b), the training information includes information such as the name of the training, the contents of the training, or the training field (training field) classified by the trainee's rank or training theme, and is managed by the training ID. The training information can also include information such as the type of training indicating the method used in the training, the instructor ID, the company ID, the date the training was held, the start time of the training, the end time of the training, and the opinions collected during the training. In this way, by configuring the system to register the training contents and opinions for each training session, it is possible to analyze and perform machine learning on the tendency of the opinions of the trainees for each training content (content or training field, type, season as determined from the training date, length of training as determined from the start time and end time). Therefore, with the above configuration, the training system can predict and implement a summary processing method that is likely to stimulate discussion, taking into account the influence of the training contents, etc. on the trainees' opinions.

[0084] The aggregation unit information, not shown, is information about the unit (hereinafter referred to as "aggregation unit") when aggregating individual opinions into a certain group. The aggregation unit information includes information such as the number of people, the IDs of the students who make up the aggregation unit, and the group ID, and is managed by the aggregation ID. The aggregation unit information may also include information such as the company ID and the aggregation unit name.

[0085] <Processing in individual work> Next, the student proceeds with the individual work through S600 to S604 in Fig. 3, and inputs the individual opinion that the student has thought up in the individual work into the input unit 54 of the student terminal 2. The control unit 51 of the student terminal 2 that receives this input transmits opinion information linking the individual opinion with student identification information (student ID) to the communication unit 13 of the information processing device 10. The communication unit 13 transmits the received opinion information to the receiving unit 101. In S704 in Fig. 4, when the receiving unit 101 receives the opinion information from the student terminal 2 via the communication unit 13, it transmits the opinion information to the memory unit 12. The memory unit 12 records the received opinion information.

[0086] In S705, the group identification unit 103 reads out the opinion information stored in the storage unit 12 and the group information registered in advance, and compares the opinion information with the group information to identify a group.

[0087] In this embodiment, if the student ID included in the opinion information matches the student ID included in the group information, the group identification unit 103 determines that the student associated with the student ID belongs to the matched group, and if they do not match, it determines that the student does not belong to the group and repeats the same matching process until they match. The group identification unit 103 performs this process on all opinion information and identifies which group the student belongs to for each individual opinion. The group identification unit 103 also links information about the individual opinions included in the opinion information with student identification information (student ID) and transmits the identification result that summarizes the individual opinions by group to the storage unit 12. The storage unit 12 records the received identification result.

[0088] In S706, the aggregation processing unit 104 reads out the opinion information stored in the storage unit 12 and the aggregation unit information registered in advance, and aggregates the individual opinions for each aggregation unit.

[0089] Specifically, the aggregation processing unit 104 determines whether one group ID is linked to one aggregation ID in the aggregation unit information, and identifies whether the aggregation unit and the group are groups composed of the same elements. The elements include, for example, a student ID included in the aggregation unit information and a student ID included in the group information. If the aggregation unit and the group are the same, the aggregation processing unit 104 reads out the identification results stored in the storage unit 12 and acquires individual opinions for each group (aggregation unit). On the other hand, if the aggregation unit and the group are different, the aggregation processing unit 104 compares the student ID included in the opinion information with the student ID included in the aggregation unit information to acquire individual opinions for each aggregation unit.

[0090] Next, the aggregation processing unit 104 creates an aggregated file consisting of a single sentence that lists the acquired individual opinions without changing the content of each individual opinion. The aggregation processing unit 104 performs this aggregation process on all groups (aggregation units) and creates one aggregated file for each aggregation unit. The aggregation processing unit 104 transmits the aggregation result that links the created aggregated file with the identification result to the storage unit 12. The storage unit 12 records the received aggregation result.

[0091] In S707, the summarization processing unit 105 reads out the aggregation results stored in the storage unit 12 and the pre-registered machine learning model, and performs summarization for each aggregated file using the summarization method predicted by the machine learning model based on the personal opinions included in the aggregation results. The summarization processing unit 105 transmits summary information linking the summary results obtained by the summarization processing with the corresponding aggregation results to the storage unit 12. The storage unit 12 records the received summary information.

[0092] <Machine learning model 1> Here, the creation of machine learning model 1 and machine learning model 2 (not shown) will be described.

[0093] First, in the machine learning model 1, the user collects documents in which the opinions of the trainees are written during the training (hereinafter referred to as "opinion documents") as learning data and stores them in the storage unit 12. This collection may be done automatically from the terminals where the trainees input their opinions, or may be done manually.

[0094] The opinion documents used as learning data are preferably of the same type as the texts input by the trainee to the trainee terminal 2 when using this training system or displayed at least on the trainee terminal by processing of the information processing device 10. In this embodiment, the same type refers to personal opinions, answers, further opinions, opinion reflection results, summary results, discussion results, and discussion reflection results.

[0095] Furthermore, the information processing device 10 creates learning data using the collected opinion documents. In this regard, the present invention may use the opinion documents as they are as learning data, but in this embodiment, in order to improve learning accuracy, learning data is created that has been preprocessed by standardizing full-width and half-width notations and correcting typos and other typos.

[0096] Next, the information processing device 10 estimates the liveliness of the discussion in the subsequent group work (hereinafter also referred to as "activity level") from the degree of difference between the results of the summary processing and the results of the discussion held by the participants based on that result.

[0097] Specifically, the information processing device 10 acquires the discussion results, summary results corresponding to the discussion results, and individual opinions, etc. corresponding to the summary results (individual opinions, answers, or opinion reflection results) from the storage unit 12. Based on these acquired discussion results, etc., the information processing device 10 analyzes the parts of the summary results where changes have been made due to the discussion results and the contents of the changes (hereinafter referred to as "change contents").

[0098] For example, if the changes are grammatical corrections such as adding punctuation marks, correcting words, and changing particles, or adjusting the format to make the text more natural (hereinafter referred to as "formal changes"), the liveliness of the discussion corresponding to that discussion result is estimated to be low. On the other hand, if the changes are changes to the gist of the text in the summary result or the addition of new opinions to the summary result, the liveliness of the discussion related to that discussion result is estimated to be high, and the estimated result of the liveliness expressed as a numerical value is transmitted to the storage unit 12.

[0099] In this case, the analysis to understand the meaning of the document can use any method used in machine learning (semantic analysis) to analyze the relationships between words in the sentence and analyze the dependency structure between words and phrases (syntactic analysis), and then link proper nouns in the sentence from a prior list based on this structure and recognize their meaning.

[0100] Next, the information processing device 10 evaluates the discussion by assigning a numerical value (hereinafter referred to as "perspective points") to each discussion on a viewpoint that is considered to liven up the discussion in the group work. In this embodiment, the viewpoint includes information indicating at least one of advice from the instructor during the discussion, advice by machine learning that created the summary result (hereinafter referred to as "advice by AI"), the combination of students constituting the group, the quality of the theme, the quality of the summary, the quality of the instructor, the motivation of the person in charge and the students for the training, and the impression of the venue where the training is held.

[0101] Specifically, advice from the instructor during the discussion includes indicators for assigning a numerical value, such as the timing and expression of the advice given by the instructor during the discussion, the students who gave the advice, and the method of conveying the advice. Advice from the AI ​​conveys supplementary information about the summary results to students and personal opinions that were not reflected in the summary results (personal opinions, answers, or results of reflecting opinions), and includes the same indicators as advice from the instructor.

[0102] In addition, the combination of students that make up a group includes as indicators the abilities, personalities, number of students, etc. In addition, the abilities and personalities of students may be analyzed for each group, and the tendency of each group may be used as an indicator.

[0103] The quality of the topic includes indicators such as the way the topic was introduced, the time spent on the introduction, the timing of when the topic was communicated to the students, the ease of discussing the topic (whether it was a topic the students had previously thought about, whether the topic was designed with the students' familiarity in mind, etc.), the way the topic was expressed, etc. The quality of the summary includes indicators such as the changes made, the number of formal changes, the number of words in the changes, and the parts that were changed.

[0104] Indicators for the quality of the instructor include the instructor's field of expertise, the number of times he or she has taught training, the instructor's career history, the impression the instructor gives to others based on the instructor's speaking style, the instructor's speaking speed, and the level of ability to smoothly proceed with the training (facilitation ability).Indicators for the motivation of the instructor and participants for the training include the training budget and number of instructors, the number of requests made to the instructor by the instructor, the clarity of the training objectives for the instructor, follow-up surveys of participants by the instructor, whether or not there was progress observation or re-education, the attendance rate of participants, whether or not participants were late, the level of response to questions posed by the instructor to the participants, and the response rate to questionnaires.

[0105] Indicators for the impression of the venue where the training will be held include the venue's facilities (indoor or outdoor, brightness of the lighting, whether the temperature can be adjusted to an appropriate level, size of the venue, whether it has windows) and location.

[0106] The perspectives explained above may be converted into numerical values ​​based on the results of a survey conducted on participants, instructors, and staff after the training has been conducted, or criteria may be determined in advance and used to convert the values ​​into numerical values.

[0107] Finally, the information processing device 10 creates a machine learning model 1 using a neural network, using the estimation results and learning data stored in the storage unit 12. In the machine learning of this embodiment, the task is to predict what values ​​the estimation results for the discussion and the viewpoint points of the discussion will take when a certain summary process is executed based on a plurality of input individual opinions, etc. (individual opinions, answers, or opinion reflection results), and the machine learning model 1 is created by learning using the learning data.

[0108] Therefore, when the machine learning model 1, which has been trained based on the training data, is further given an opinion text as additional training data, it can estimate a summarization method that will increase the estimated result and viewpoint points from the relationship between the summary result, the estimated result, and the viewpoint points. Furthermore, the machine learning model 1 is updated by verifying whether the activation of discussion has actually been promoted based on at least the individual opinions and the summary result created using the machine learning model 1 and the result of reflecting the discussion related to the summary result.

[0109] <Machine learning model 2> Next, the creation of machine learning model 2 (not shown) will be described. First, the user collects as learning data personal opinions, further opinions, opinion reflection results, discussion results, or discussion reflection results, a processing method output by the model based on input data (hereinafter referred to as "input data"), and an answer or summary result created as a result of the processing, and stores them in storage unit 12. The collection and storage of learning data, creation of learning data, and updating of the machine learning model described in machine learning model 1 are also performed in machine learning model 2 in the same way.

[0110] Here, the processing method output by the model refers to the method of processing the individual opinions, etc. predicted by the model when the individual opinions, etc. (individual opinions, answers, or results of reflecting opinions) are input into the model in order to obtain results from the individual opinions, etc. that are likely to stimulate discussion; in other words, the instructions given by the model to the information processing device 10 for processing the input data.

[0111] Therefore, the processing method output by the model (hereinafter referred to as the "processing method") includes not only a summary processing method, but also a series of processing methods for aggregating and summarizing individual opinions, etc. after the identification processing has been performed in the group identification unit.

[0112] Next, the information processing device 10 estimates the level of activity of the discussion in the subsequent group work from the degree of difference between the result of processing according to the processing method and the result of the discussion by the students based on that result. As a specific method of estimation, a method similar to that of the machine learning model 1 can be adopted.

[0113] Next, the information processing device 10 evaluates the discussion by giving viewpoint points to each discussion from a viewpoint that is considered to make the discussion in the group work more lively. In this embodiment, the viewpoint may be the same as that of the machine learning model 1.

[0114] Finally, the information processing device 10 creates a machine learning model 2 using a neural network, using the estimation results and learning data stored in the storage unit 12. In the machine learning of this embodiment, the machine learning model 2 is created by learning using the learning data, with the task being to predict what values ​​the estimation results for the discussion and the viewpoint points of the discussion will take when a processing method is executed based on a plurality of input individual opinions, etc. (individual opinions, answers, or opinion reflection results).

[0115] Specifically, when machine learning model 2 is instructed to aggregate and summarize the personal opinions of all members of a specific group based on input data, the discussion based on the resulting summary may not obtain high estimation results and viewpoint points. Machine learning model 2 further learns from such cases as learning data, and when it receives the same input data again, it can change the processing method to output instructions to analyze the content of each member's personal opinion and perform summary processing for opinions with similar meanings.

[0116] More specifically, if a group has four participants, the training system will aggregate the opinions of all four participants and perform a summary that reflects the opinions of all four participants. It is desirable for the results of this summary process to stimulate discussion in group work, but there may be situations in which this is not possible. In such cases, the summary process may be performed based on the opinions of only two participants in the group, resulting in a summary that stimulates discussion. In this way, machine learning model 2 learns from the learning data a series of events that shows that a particular processing method is not likely to stimulate discussion, but that a new processing method presented subsequently is likely to stimulate discussion.

[0117] In addition, machine learning model 2 may output as processing methods a summary of the opinions of four people in the specific group, a summary of the opinions of four people divided into arbitrary aggregation units (for example, differences in the attributes of the participants or the meaning of the opinions), i.e., a summary method that learns an aggregation method that will result in a high evaluation of the inference results or perspective points and then summarizing each aggregation unit aggregated under the learned conditions, or a summary method that incorporates or adds the contents of opinions classified as minority opinions within the group or on the same topic, and these output processing methods may be repeatedly learned as training data to estimate processing methods that will result in higher evaluations of the inference results or perspective points.

[0118] By creating the machine learning model 2 as described above, the machine learning model 2 can learn a series of processing methods that can be expected to stimulate discussion based on the relationship between the estimation results and viewpoint points, and can make accurate estimations.

[0119] Note that the machine learning models in the present invention are not limited to the above machine learning model 1 and machine learning model 2, and machine learning models obtained by other learning may also be used.

[0120] <Group work processing> Next, the students hold a discussion while referring to the summary result in S605 of Fig. 3, and input the result of the discussion (discussion result) into the input unit 54 of the student terminal 2. The control unit 51 of the student terminal 2 that receives this input transmits discussion information linking the discussion result with the student identification information (student ID) to the communication unit 13 of the information processing device 10. The communication unit 13 transmits the received discussion information to the receiving unit 101. In S708 of Fig. 4, when the receiving unit 101 receives the discussion information from the student terminal 2 via the communication unit 13, it transmits the discussion information to the memory unit 12. The memory unit 12 records the received discussion information.

[0121] In S709, the discussion reflection section 106 reads out the discussion information and summary information stored in the storage section 12, identifies a summary result corresponding to each discussion result, and performs processing to reflect the discussion results in the identified summary result.

[0122] Specifically, the discussion reflection unit 106 compares the participant identification information (participant ID) included in the discussion information with the participant identification information (participant ID) included in the summary information, and if the two IDs match, obtains the summary result included in the summary information, and if they do not match, repeats the same comparison process until they match, and obtains the summary result at the point when they match. The discussion reflection unit 106 performs this process for all discussion information, and identifies the summary result that corresponds to the aggregation unit to which the participant who entered each discussion result belongs. That is, the discussion reflection unit 106 identifies the summary result onto which the results of the discussion are to be reflected by the above identification.

[0123] Next, the discussion reflection unit 106 reflects the discussion result corresponding to the identified summary result. Specifically, the discussion reflection unit 106 compares the identified summary result with the corresponding discussion result, and performs a process of reflecting the discussion result in the summary result by making no changes to the matching wording and updating the wording of the summary result to the wording of the discussion result for the mismatched wording. The discussion reflection unit 106 links the discussion reflection result created based on the reflection process to information on the discussion result contained in the discussion information and summary information including the corresponding summary result, and transmits it to the memory unit 12. The memory unit 12 records the received information (discussion reflection result, etc.).

[0124] <Display processing> In S710, the display processing unit 107 reads out the discussion reflection results and the like stored in the storage unit 12, and acquires information regarding the correspondence between the discussion reflection results and each group from the group ID included in the aggregated result of the summary information. Based on the acquired information, the display processing unit 107 identifies the participants of each group corresponding to the discussion reflection results. In addition, the display processing unit 107 acquires the corresponding summary results and discussion reflection results from the discussion reflection results and the like, performs display processing on a GW screen (GW screen 90 described later) including these, and transmits the display processing result to the output unit 55 of the participant terminal 2 used by the identified participant. When the control unit 51 receives the display processing result from the display processing unit 107 via the transmission unit 102, it causes the output unit 55 to display the GW screen into which the discussion reflection results have been input. The display process is similar to the case where the summary results are displayed between S707 and S708 (not shown). In this case, the discussion reflection result can be read as the summary result, the summary result can be read as the individual opinion linked to the summary information aggregation result, i.e., the individual opinion before being summarized, and the GW screen can be read as the individual work screen.

[0125] <Individual work screen> As shown in FIG. 7(a), the individual work screen 80 displays the theme of the individual work and a field for inputting the student's thoughts about the theme of the individual work ("personal notes" shown in FIG. 7(a), hereafter referred to as the "personal notes field").

[0126] Also, a section for inputting further opinions (hereinafter, the "further opinion section") and a section for displaying replies ("Response" shown in FIG. 7(a), hereinafter, the "Response section") may be displayed, and the personal memo section, further opinion section, and response section may be configured in the same area. In this case, the display portion within the same area may be divided according to the content to be displayed, or the content of the display portion may be switched based on a specific operation.

[0127] Furthermore, the individual work screen 80 may include a column for displaying information to be communicated from the instructor to the student ("Comments from the instructor" shown in FIG. 7(a) hereinafter referred to as the "instructor column"), a column for displaying the time limit, end time, or remaining time for the individual work ("5 minutes remaining" shown in FIG. 7(a) hereinafter referred to as the "time display column"), a button for instructing the creation of an answer based on an individual opinion ("Create answer by AI" shown in FIG. 7(a) hereinafter referred to as the "answer creation button"), and a button for instructing the creation of a summary result from the student's opinion on the topic ("Submit" shown in FIG. 7(a) hereinafter referred to as the "Submit button"). Note that the arrangement and display form of each display column and each button on the individual work screen 80 and the GW screen are not limited to the configuration shown in FIG. 7(a) and may be selected arbitrarily.

[0128] <Group work screen> As shown in Figure 7(b), the GW screen 90 displays the group work theme, a column for displaying summary results (the box below "Summary" shown in Figure 7(b), hereafter referred to as the "Summary column"), and a column for displaying the results of the discussion reflection (the box below "Summary" shown in Figure 7(b), hereafter referred to as the "Summary column").

[0129] Also, a column for inputting opinions (discussion results) that participants have considered based on the summary results in group work (hereinafter referred to as the "discussion results column") and a column for displaying each member's personal opinions or responses (the boxes below "Mr. A," "Mr. B," and "Mr. C" in Figure 7(b)) may be displayed, and the summary column and discussion results column may be configured in the same area. In this case, the discussion results column may be configured so that the summary results are displayed in advance, and participants only need to input the minimum parts of the summary results they wish to change, and the discussion reflection results are displayed.

[0130] The discussion result column may display the content newly input by the students so that it can be distinguished from the summary result. This configuration allows the students to more clearly recognize the results of the discussion, and increases their excitement about the discussion, leading to more active discussion. In addition, the discussion result column may allow multiple students to simultaneously input changes to the summary result. In this case, the display processing unit 107 may be configured to provide a column for displaying the changes made by each student in the column for inputting the discussion result. This configuration allows the members to confirm and understand how each student perceives the content of the discussion. In addition, the discussion reflection result may be configured to be displayed in real time based on the changes input by the students, even if the summary creation button is not operated.

[0131] Furthermore, the GW screen 90 can be configured to include an instructor column ("Comments from the instructor" shown in FIG. 7(b)) and a time display column for group work ("8 minutes remaining" shown in FIG. 7(b)), similar to the individual work screen 80.

[0132] The answer creation button, send button, and summary creation button may perform their respective processes when the time limit for the individual work or group work has expired, and it is not necessary to operate each button.

[0133] In the training in this embodiment, the above flow (S601 to S607, S701 to S710) is repeated until the end of the training (S608). This allows the training system to end its operation.

[0134] As can be understood from the above operation description, according to the above embodiment, the training support system predicts a method of summarizing individual opinions that is likely to stimulate discussions using a machine learning model based on the estimation result that measures the liveliness of the discussion and the index of viewpoint points, and shares the results of summarization using that method with the participants in the group. As a result, the training support system according to the present invention can streamline the complicated sharing work by simply allowing each participant to understand the results of summarization of their individual opinions. Therefore, participants can smoothly start substantive discussions in which they discuss their individual opinions within the group, and the creation of a summary result that is likely to stimulate the discussion can not only stimulate but also promote the development of the discussion.

[0135] The present invention is not limited to the above-described embodiment, and various modifications are possible without departing from the object of the present invention.

[0136] For example, in the above embodiment, the group identification unit 103 is configured to perform the identification process of identifying which group a student belonging to each individual opinion has entered after receiving the opinion information and before the aggregation process. However, this identification process only needs to be performed before identifying the output unit 55 to which the summary result is to be sent after the summary process, and the timing of the process is not limited to this embodiment. Note that, if a group is specified as the aggregation unit, it is natural that the identification process must be performed before the aggregation process.

[0137] In the above embodiment, the group identification unit 103 is configured to identify a group by comparing the opinion information with the group information. However, when a group ID is registered in association with a student ID, the group identification unit 103 may identify a group from the group ID associated with the student ID of the opinion information without performing the comparison.

[0138] In the above embodiment, the student ID is used as the student identification information included in the opinion information and summary information. However, the student identification information may be any information that can identify the student, and may be, for example, information indicating the relationship between the number for identifying the terminal used in the training and each student. The same applies to the group identification information.

[0139] In the above embodiment, the summarization processing unit 105 is configured to perform summarization for each aggregate unit based on the summarization method predicted by the trained machine learning model. However, the summarization processing unit 105 may be configured to perform summarization for each individual opinion in the individual work, and the aggregation processing unit 104 may perform summarization again for each aggregate unit based on the summary result (answer). With this configuration, the students can organize their own thoughts by checking the summary result of their individual opinions before the discussion in the group work, and deepen the content of their opinions from a spontaneous state to a thoughtful result by themselves. In addition, the students can know how the machine learning model understood their opinions, and can smoothly proceed with the discussion by paying attention to expressions that may cause misunderstandings.

[0140] In the above embodiment, the discussion reflection unit 106 is configured to perform a process of reflecting the results of the discussion corresponding to the answers or summary results in these answers or summary results. However, the discussion reflection unit 106 may be configured to perform a process of reflecting further opinions that the students have considered in their individual work based on the answers in the answers. With this configuration, the students can revisit their personal opinions in their individual work and use the polished content to take part in the discussion.

[0141] In the above embodiment, the aggregation processing unit 104 is configured to aggregate individual opinions for each group, assuming that the aggregation unit and the group are the same. However, the aggregation processing unit 104 may be configured to aggregate individual opinions for each group of at least two or more students constituting a group, and does not necessarily have to use the group as the aggregation unit. For example, in the group ID 20001 shown in FIG. 5(b), multiple aggregation units may exist within the same group, with student ID 10001 and student ID 10002 as the first aggregation unit and other students as the second aggregation unit. In this case, the display processing unit 107 may perform display processing in a display form in which a display area is divided for each summary result, or may perform display processing in a display form in which multiple summary results are combined into one and each can be identified.

[0142] In the above embodiment, the aggregation processing unit 104 is configured to aggregate the individual opinions into one sentence (aggregated file) without changing the contents of the descriptions. However, the aggregation processing unit 104 may be configured to analyze the contents of multiple individual opinions (individual opinions, answers, and opinion reflection results) belonging to an aggregation unit, collect them into sentences or opinions containing similar meanings, and aggregate them into groups based on the meanings within the aggregation unit. With this configuration, the machine learning model of the summary processing unit 105 understands the opinions within the aggregation unit based on the groups in which opinions with similar meanings are summarized, so that it is possible to accurately grasp the contents of the opinions and predict a summary method that is likely to stimulate discussion. In addition, since the summary processing unit 105 can more accurately understand the opinions input by the students, it is possible to suppress the excessive omission of each meaning contained in the opinions in the summary processing, and to reflect claims classified as minority opinions in the summary result. Therefore, it is possible to encourage students who have been unable to actively discuss because they have minority opinions to participate in the discussion, and it is also possible to evaluate whether minority opinions, which tend to be buried in discussions, stimulate discussion.

[0143] In the above embodiment, the display processing unit 107 is configured to generate screen data in the display mode shown in Figures 7(a) and (b) regarding the arrangement and expression method of each display area of ​​the personal work screen 80 or the GW screen 90. However, the display processing unit 107 may be configured to perform display processing in a mode in which at least the personal opinions, summary results, and discussion reflection results are displayed in the same area.

[0144] Furthermore, the display processing unit 107 may be configured to display and process individual opinions, answers summarizing the individual opinions in individual work, opinion reflection results reflecting further opinions of the students based on the answers, summary results or discussion reflection results, or a combination of these, and information to be communicated from the instructor to the students in the same area.

[0145] In the above embodiment, the present training system etc. is described on the assumption that it is used in training. However, the present training system etc. can also exert the same effect as in the case of the above training in a test for selecting people to be employed as workers by an organization or in a discussion for evaluating members of an organization. [Explanation of symbols]

[0146] 1... training support system, 2... participant terminal, 3... instructor terminal, 4... person in charge terminal, 10... information processing device, NW... network, DB... database, 11: control unit, 12: storage unit, 13: communication unit, 101: receiving unit, 102: transmitting unit, 103: group identification unit, 104: aggregation processing unit, 105: summary processing unit, 106: discussion reflection unit, 107: display processing unit, 50: terminal, 51: control unit, 52: storage unit, 53: communication unit, 54: input unit, 55: output unit, 80...Individual work screen, 90...Group work screen (GW screen)

Claims

1. A training support system that supports a training including an individual work in which a trainee taking the training summarizes his / her personal opinion on a theme of the training, and a group work in which the trainees discuss in a group consisting of a plurality of the trainees based on the opinions of each of the trainees, The training support system includes: a summarizing unit that summarises the opinions of at least two or more of the students who make up the group; A training support system comprising: a summary processing unit that summarizes the collected opinions using a machine learning model.

2. The training support system according to claim 1, further comprising: a storage unit for storing in advance student identification information for identifying the student and group identification information for identifying a group to which the student belongs; a receiving unit for receiving, from a student terminal used by the student, opinion information including the student's opinion and the student identification information of the student; a group identification unit that identifies the group to which the student belongs by comparing the group identification information with the student identification information.

3. The training support system according to claim 1 or 2, further comprising: A training support system comprising: a discussion reflection unit that reflects the results of the discussion among the trainees based on the summary result obtained by summarizing the collected opinions.

4. 3. The training support system according to claim 1, The summary processing unit includes: A training support system characterized by summarizing the opinions in the individual work using the machine learning model.

5. 3. The training support system according to claim 1, The summary processing unit includes: A training support system characterized in that the collected opinions are summarized using the machine learning model with a view to using them to evaluate the group work.

6. 4. The training support system according to claim 3, The discussion reflection unit includes: A training support system characterized by accepting further opinions that the student has further considered regarding the answer that is the result of summarizing the opinions in the individual work from the student terminal used by the student, and reflecting these further opinions in the answer.

7. 4. The training support system according to claim 3, The discussion reflection unit includes: A training support system characterized in that a process is carried out in which the results of the discussion are immediately reflected in the summary result in response to operations of at least two or more of the students constituting the group in the group work.

8. The training support system according to claim 3, further comprising: A training support system further comprising a display processing unit for displaying the summary result in the same area.

9. The training support system according to claim 3, further comprising: A training support system characterized by having a display processing unit that displays in the same area either the opinions, answers that are the result of summarizing the opinions in the individual work, opinion reflection results in which further opinions of the participants regarding the answers in the individual work are reflected in the answers, the summary results, or discussion reflection results in which the results of the discussion are reflected, or a combination of these, and information to be communicated to the participants by the instructor of the training.

10. A training support method implemented by a training support system that supports a training including an individual work in which a trainee taking the training summarizes his / her personal opinion on a theme of the training, and a group work in which a group of a plurality of the trainees discusses the training based on the opinions of each of the trainees, a summarizing step of summarizing the opinions of at least two or more of the students who make up the group; A training support method comprising: a summarizing step of summarizing the collected opinions using a machine learning model.

11. A training support program that causes a computer device to execute support for the training, the training including an individual work in which a trainee taking the training summarizes his / her personal opinion on a theme of the training, and a group work in which a group of a plurality of the trainees discusses the training based on the opinion of each of the trainees, The computer device includes: a summarizing step of summarizing the opinions of at least two or more of the students who make up the group; A training support program that executes a summary processing step of summarizing the collected opinions using a machine learning model.

Images