Providing method, providing device, and machine learning program

Abstract

To make a proposal for improving an organization effectively, on the basis of a summary result of a PLE questionnaire.SOLUTION: A providing method causes a computer to execute: processing of acquiring a result of a questionnaire regarding a learning environment of an organization; and processing of presenting at least one of an analysis result obtained through the questionnaire result and a trained model and a proposal comment for developing the organization. The trained model outputs, on receipt of input of data corresponding to the questionnaire result, data corresponding to at least one of the analysis result and the proposal comment.SELECTED DRAWING: Figure 5

Description

[Technical field]

[0001] The present invention relates to a providing method, a providing device, and a machine learning program. [Background technology]

[0002] In order to effectively develop personnel that will lead to organizational performance, a method has been known in which a Positive Learning Environment (PLE) questionnaire is administered to organizational personnel, and organizational improvement measures are formulated based on the results of the questionnaire. Here, when a Positive Learning Environment (PLE) is established in an organization, and individuals are able to think and act independently, it is expected that individual awareness of learning will be strengthened, behavioral changes will occur, and a positive spiral of organizational growth will be formed. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2019-133225 A [Patent Document 2] JP 2002-328590 A [Patent Document 3] JP 2002-196658 A [Patent Document 4] JP 2002-072848 A [Patent Document 5] JP 2005-092492 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, it is difficult to make effective proposals for improving an organization based on the results of a PLE questionnaire. This is because each item in the PLE questionnaire has a complex relationship with the others, and it is difficult to find areas for improvement in an organization based on the results of each individual questionnaire item alone. Therefore, there is a need for a method to use the results of a PLE questionnaire to properly understand the current problems in an organization and to determine effective educational policies and human resource development strategies.

[0005] The present invention aims to make proposals for effectively improving organizations based on the collected results of a PLE questionnaire. [Means for solving the problem]

[0006] The present invention provides a provision method in which a computer executes a process of acquiring survey results regarding an organization's learning environment, and a process of presenting at least one of an analysis result of the organization and a suggested comment for growing the organization obtained using the survey results and a trained model, wherein the trained model outputs data corresponding to at least one of the analysis result and the suggested comment when data corresponding to the survey results is input.

[0007] The present invention also provides a provision device including an acquisition unit that acquires survey results regarding an organization's learning environment, and a presentation unit that presents at least one of an analysis result of the organization and a proposed comment for growing the organization obtained using the survey results and a trained model, wherein the trained model outputs data corresponding to at least one of the analysis result and the proposed comment when data corresponding to the survey results is input.

[0008] The present invention also provides a machine learning program that causes a computer to execute a process of acquiring survey results regarding an organization's learning environment, and a process of presenting at least one of an analysis result of the organization and a suggested comment for growing the organization obtained using the survey results and a trained model, wherein the trained model outputs data corresponding to at least one of the analysis result and the suggested comment when data corresponding to the survey results is input. Effect of the Invention

[0009] According to the present invention, it is possible to effectively make proposals for improving an organization based on the collected results of a PLE questionnaire. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 is a diagram illustrating an example of the configuration of a provision system according to a first embodiment. [Diagram 2] FIG. 2 is a diagram showing an example of a PLE questionnaire according to the first embodiment. [Diagram 3] FIG. 2 is a diagram showing an example of presented information according to the first embodiment. [Figure 4] FIG. 2 is a diagram showing an example of a logic tree according to the first embodiment. [Diagram 5] 2 is a block diagram showing an example of the configuration of a providing device according to the first embodiment. FIG. [Figure 6] FIG. 2 is a diagram showing an example of presented information according to the first embodiment. [Figure 7] FIG. 2 is a diagram showing an example of presented information according to the first embodiment. [Figure 8] FIG. 2 is a diagram showing an example of presented information according to the first embodiment. [Figure 9] FIG. 2 is a diagram showing an example of presented information according to the first embodiment. [Figure 10] FIG. 2 is a diagram showing an example of presented information according to the first embodiment. [Figure 11] 1 is a flowchart showing the overall flow of a provision process according to the first embodiment. [Figure 12] FIG. 11 is a diagram illustrating an example of the configuration of a provision system according to a second embodiment. [Figure 13] FIG. 11 is a diagram illustrating an example of a learning process according to the second embodiment. [Figure 14] FIG. 11 is a diagram illustrating an example of a learning process according to the second embodiment. [Figure 15] FIG. 11 is a diagram illustrating an example of a learning process according to the second embodiment. [Figure 16] FIG. 11 is a diagram illustrating an example of a learning process according to the second embodiment. [Figure 17] FIG. 11 is a block diagram showing an example of the configuration of a providing device according to a second embodiment. [Figure 18] 10 is a flowchart showing the overall flow of a provision process according to the second embodiment. [Figure 19] FIG. 2 is a diagram illustrating an example of a hardware configuration. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, a providing method, a providing device, and a machine learning program according to an embodiment of the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to the embodiment described below.

[0012] [Embodiment 1] The configuration of a providing system 100-1 and a providing device 10-1 according to the first embodiment and the flow of each process will be described below in order, and finally, the effects of the first embodiment will be described.

[0013] [1. Configuration of the provision system 100-1] The configuration of a provision system 100-1 according to the first embodiment (hereinafter, also referred to as the present system as appropriate) will be described in detail with reference to Fig. 1. Fig. 1 is a diagram showing an example of the configuration of the provision system according to the embodiment. Below, an example of the overall configuration of the present system 100-1 will be shown, and then the processing of the present system 100-1 will be described. Note that in this embodiment, a system for providing organizational improvement measures in business development will be described as an example, but this is not intended to be limited to this, and the present invention can be applied to various systems.

[0014] (1-1. Example of overall system configuration) This system 100-1 includes a providing device 10-1. Note that this system 100-1 may include a plurality of providing devices 10-1. In this system 100-1, the data input to the providing device 10-1 includes a PLE questionnaire result 21, and the data output by the providing device 10-1 includes presented information 30 including an analysis result 31 and a proposed comment 32.

[0015] (1-2. Overall system processing) The providing device 10-1 in such a system is an example of a computer device that has a trained machine learning model 14d and uses the machine learning model 14d to generate and output presentation information 30 from the PLE questionnaire result 21. In the following, the machine learning model 14d held by the providing device 10-1 will be described, followed by a description of the input process and output process executed by the providing device 10-1, in that order.

[0016] First, the machine learning model 14d possessed by the providing device 10-1 is a model trained to generate the presented information 30 in response to the input of the PLE questionnaire result 21, and a neural network or the like can be adopted. In addition, various algorithms can be adopted for the learning of the machine learning model 14d. For example, the providing device 10-1 inputs "PLE questionnaire result, presented information" as learning data "input data, correct answer information" to the machine learning model 14d, and executes learning of the machine learning model 14d so that the output result of the machine learning model 14d matches the correct answer information. Then, the providing device 10-1 uses the machine learning model 14d trained in this way to obtain information to be presented from the aggregated result of the PLE questionnaire, and executes various proposals for effectively improving the organization. Below, an example of executing various proposals will be described.

[0017] (1-2-1. Input processing) First, the providing device 10-1 accepts an input of the PLE survey result 21 (step S1). At this time, the PLE survey result 21 accepted by the providing device 10-1 may be data compiled for each item of the PLE survey, or may be data before compilation transmitted by a terminal device owned by the survey respondent.

[0018] (1-2-2. Output processing) Secondly, the providing device 10-1 outputs the presented information 30 including the analysis result 31 and the proposed comment 32 acquired from the machine learning model 14d (step S2). At this time, the providing device 10-1 inputs the PLE questionnaire result 21 to the machine learning model 14d and acquires the presented information 30. For example, the machine learning model 14d performs the following operations on the PLE questionnaire result 21: tallying up the results by item, comparing the results with those of other departments / divisions and other companies, evaluating the results by analysis category, identifying elements that require improvement, and the like. Then, the presented information 30 including the outputted analysis result 31 and the proposed comment 32 is presented to the companies, etc. that responded to the pre-training PLE questionnaire result 21-1, and is used as an organizational improvement measure in corporate development.

[0019] In the example of FIG. 1, the analysis result 31 is displayed as a "corporate development status analysis" that compares "Company A" and "Company B" for each specified analysis category using a radar chart, but the analysis result 31 may be displayed as a line graph or bar graph showing the achievement level for each question, a pie chart or band graph showing the achievement level for each analysis category, or various numerical values ​​may be displayed in a table, and is not particularly limited. Also, in the example of FIG. 1, the suggested comment 32 shows advice on an area (category) or an element within the area, such as "There seems to be an issue in the XX area. In particular, there seems to be room for improvement in △△ and □□." However, the suggested comment 32 may be advice related to multiple analysis categories or the priority order thereof, and is not particularly limited.

[0020] By executing the above-mentioned steps S1 to S2, the PLE questionnaire results can be used to properly grasp the current problems in the organization and to determine effective education policies and human resource development strategies.

[0021] (1-3. Examples of data involved in each process) 2 to 4, the data involved in the processing in steps S1 to S2 described above will be described. The data involved in the input processing in step S1, the data involved in the output processing in step S2, and the data involved in the input / output processing in steps S1 and S2 will be described in that order.

[0022] (1-3-1.PLE Survey 20) A specific example of the PLE questionnaire 20 before answering, which is involved in the input process of step S1, will be described with reference to FIG. 2. FIG. 2 is a diagram showing an example of the PLE questionnaire according to the first embodiment. As shown in FIG. 2, the PLE questionnaire 20 has 30 questions, from "Question 1" to "Question 30", and respondents answer each question anonymously with "Yes" or "No". For example, the answers of members belonging to each department or division of a company are collected, and it is found that items with a small number of "Yes" answers require improvement. Note that the PLE questionnaire 20 shown in FIG. 2 is only an example, and it is possible to reduce the number of items in the questionnaire or to add other questions. In addition, instead of the format of answering with "Yes" or "No", it is also possible to answer each question in a format of answering the organization's achievement level on a five-point scale, and is not particularly limited.

[0023] (1-3-2. Presentation information 30) A specific example of the presentation information 30 involved in the output process of step S2 will be described with reference to FIG. 3. FIG. 3 is a diagram showing an example of the presentation information according to the first embodiment. As shown in FIG. 3, the presentation information 30 includes analysis results 31 (31-1, 31-2). The analysis results 31 are presented as a radar chart or the like that can be compared for each analysis category, such as the analysis result 31-1. Note that the presentation information 30 including the analysis results 31 and the proposed comments 32 output by the providing device 10-1 differs depending on the learning content of the machine learning model 14d. In the following, the presentation information 30 of the machine learning model 14d different from the example of FIG. 1 will be described.

[0024] Here, each analysis category displayed on the radar chart is a collection of related elements classified by a logic tree described later. Here, in the example of analysis result 31-1 in Figure 3, the analysis categories shown are "Goals required in the workplace are indicated," "There is a learning strategy linked to achieving the goals," "What is learned is utilized in the workplace," "It is a positive learning environment," and "Learning outcomes are evaluated." The questions in the 20 PLE questionnaires mentioned above are classified into analysis categories taking into consideration their relevance, and the score of the analysis category is determined not simply by the score of each question, but also by the degree of achievement of questions classified in the same analysis category.

[0025] In addition, the analysis result 31 presents a benchmark for each organization based on the overall score, for example, analysis result 31-2. Here, the example of analysis result 31-2 in Figure 3 shows the score of the comparison target, Company A, at 23 points and ranking at 11th, and the score of Company B, at 4 points and ranking at 85th.

[0026] Meanwhile, as shown in Fig. 3, the presented information 30 includes suggested comments 32 (32-1, 32-2). The suggested comments 32 are, for example, suggested comment 32-1, and are presented as "executive summaries" that are comments for improving the training situation at Company B. In the example of Fig. 3, the executive summary shows that the training situation at Company B is lower than the industry average, and in order to improve it, a message is displayed recommending that "it is a positive learning environment" and "learning results are evaluated," which are values ​​below the threshold in the radar chart of the analysis result 31-1, be improved.

[0027] The content output in the executive summary can be changed depending on the content of machine learning. For example, when the machine learning model 14d is trained, the machine learning model 14d can automatically generate the content of the executive summary according to the input of the answer to the PLE questionnaire by including the evaluation value of each analysis category in the explanatory variable of the training data and the message content in the correct answer information (label) of the training data. In this case, the providing device 10-1 displays and outputs the result output from the machine learning model 14d as it is.

[0028] Furthermore, during learning of the machine learning model 14d, by including the evaluation value of each analysis category in the explanatory variables of the learning data and the keyword in the correct answer information (label) of the learning data, the machine learning model 14d can output the keyword in response to the input of the answer to the PLE questionnaire. In this case, the providing device 10-1 inputs the keyword output from the machine learning model 14d into a fixed phrase prepared in advance and displays and outputs it.

[0029] In addition, the proposed comment 32 presents "potential" or the like, which is a comment on the educational policy and effects of Company B, such as proposed comment 32-2. In the example of Figure 3, a message evaluating "What has been learned is being put to use in the workplace," which has a value above the threshold in the radar chart of analysis result 31-1, is displayed as potential.

[0030] The contents displayed in the potential can be arbitrarily changed according to the learning contents, similar to the executive summary. For example, during learning of the machine learning model 14d, the machine learning model 14d can automatically generate the contents of the potential according to the input of the PLE questionnaire by including the evaluation value of each analysis category in the explanatory variables of the learning data and the message content in the correct answer information (label) of the learning data. Furthermore, during learning of the machine learning model 14d, the machine learning model 14d can output a keyword according to the input of the PLE questionnaire by including the evaluation value of each analysis category in the explanatory variables of the learning data and the keyword in the correct answer information (label) of the learning data.

[0031] (1-3-3. Logic tree) A specific example of a logic tree involved in the input / output processing of steps S1 and S2 will be described with reference to FIG. 4. FIG. 4 is a diagram showing an example of a logic tree according to the first embodiment. As shown in FIG. 4, the logic tree classifies elements, which are questions or groups of questions in a questionnaire, into analysis categories that are related to each other, and indicates causal relationships and priorities. For example, the analysis categories classify PLE questionnaire items into "encouragement of innovation," "learning," "system," "place," "management organizational climate," and the like. In addition, for example, "management organizational climate" corresponds to the above "goals required in the workplace are indicated," "system" corresponds to the above "there is a learning strategy linked to achieving the goals," "place" corresponds to the above "learning is utilized in the workplace," "learning" corresponds to the above "it is a positive learning environment," and "encouragement of innovation" corresponds to the above "the results of learning are evaluated."

[0032] Here, the positional relationships of "Category A" to "Category E" and "Element a" to "Element r" shown in Figure 4 indicate that the higher you go, the more difficult it is to achieve. For example, an organization that has not achieved "Category A," "Category B," "Category C," and "Category E" cannot achieve "Category D." It also shows that, for example, to achieve "Category A," improvements must be made in the order of "Element о" → "Element j" → "Element a." Also, even elements classified into different categories are related to each other; for example, "element q" classified into "Category C" has a causal relationship with "element a" in "Category A," "element p" in "Category B," and "element r" in "Category E."

[0033] In the input / output processing of steps S1 to S2 described above, the providing device 10-1 can calculate the achievement level of each analysis category as a score using a pre-set logic tree based on the PLE questionnaire result 21. In addition, the machine learning model 14d possessed by the providing device 10-1 is a trained model based on the causal relationships and priorities indicated by the above logic tree.

[0034] [2. Configuration of the providing device 10-1] 5 to 9, a functional configuration of the providing device 10-1 included in the providing system 100-1 shown in Fig. 1 will be described. The configuration of the providing device 10-1 according to the first embodiment and the presented information according to the first embodiment will be described in detail below.

[0035] (2-1. Configuration of the providing device 10-1) First, the configuration of the providing device 10-1 shown in Fig. 1 will be described with reference to Fig. 5. Fig. 5 is a block diagram showing an example of the configuration of the providing device according to embodiment 1. The providing device 10-1 has an input unit 11, an output unit 12, a communication unit 13, a storage unit 14, and a control unit 15-1.

[0036] (2-1-1. Input section 11) The input unit 11 controls input of various information to the providing device 10-1. For example, the input unit 11 is realized by a mouse, a keyboard, etc., and accepts input of setting information and the like to the providing device 10-1.

[0037] (2-1-2. Output section 12) The output unit 12 controls the output of various information from the providing device 10-1. For example, the output unit 12 is realized by a display or the like, and outputs setting information stored in the providing device 10-1.

[0038] (2-1-3. Communications Department 13) The communication unit 13 manages data communication with other devices. For example, the communication unit 13 performs data communication with each communication device via a router, etc. The communication unit 13 can also perform data communication with an operator's terminal (not shown).

[0039] (2-1-4. Storage section 14) The storage unit 14 stores various information referenced when the control unit 15-1 operates and various information acquired when the control unit 15-1 operates. The storage unit 14 has a learning data storage unit 14a, a questionnaire result storage unit 14b, a presentation information storage unit 14c, and a machine learning model 14d. Here, the storage unit 14 can be realized by, for example, a semiconductor memory element such as a random access memory (RAM) or a flash memory, or a storage device such as a hard disk or an optical disk. In the example of FIG. 5, the storage unit 14 is installed inside the providing device 10-1, but may be installed outside the providing device 10-1, or multiple storage units may be installed.

[0040] (Learning data storage unit 14a) The learning data storage unit 14a stores learning data for learning the machine learning model 14d. For example, the learning data storage unit 14a stores data to be input to the machine learning model 14d, such as “PLE questionnaire results, presented information” as learning data “input data, correct answer information.”

[0041] (Survey result storage unit 14b) The survey result storage unit 14b stores the PLE survey results 21 to be input to the machine learning model 14d. For example, the survey result storage unit 14b stores responses to the PLE survey 20 shown in Fig. 2 for each company, department, or division. The survey result storage unit 14b may also store data obtained by tallying the responses to the PLE survey 20 shown in Fig. 2.

[0042] (Presentation information storage unit 14c) The presented information storage unit 14c stores the analysis results 31 and the proposed comments 32 acquired from the machine learning model 14d by the presentation unit 15c of the control unit 15-1 described below. For example, the presented information storage unit 14c stores, as the analysis results 31, a radar chart showing the scores for each analysis category of "Company A" and "Company B" shown in Fig. 1, and, as the proposed comments 32, comments showing areas where "Company B" has issues shown in Fig. 1.

[0043] (Machine learning model 14d) The machine learning model 14d is a model trained to generate presentation information 30 in response to input of the PLE questionnaire result 21. For example, the machine learning model 14d is a trained model generated by a generation unit 15a of a control unit 15-1 (described later) using training data stored in a training data storage unit 14a.

[0044] (2-1-5. Control unit 15-1) The control unit 15-1 controls the entire providing device 10-1. The control unit 15-1 includes a generating unit 15a, an acquiring unit 15b, and a presenting unit 15c. Here, the control unit 15-1 can be realized by, for example, an electronic circuit such as a central processing unit (CPU) or a micro processing unit (MPU), or an integrated circuit such as an application specific integrated circuit (ASIC) or a field programmable gate array (FPGA).

[0045] (Generation unit 15a) The generation unit 15a inputs learning data to the machine learning model 14d to generate a trained model. For example, the generation unit 15a inputs "PLE questionnaire result, presented information" as learning data "input data, correct answer information" to the machine learning model 14d, and executes learning of the machine learning model 14d so that the output result of the machine learning model 14d matches the correct answer information. The generation unit 15a acquires the learning data from the training data storage unit 14a. Also, as described above, the analysis content by the machine learning model 14d can be changed by changing the correct answer information.

[0046] (Acquisition part 15b) The acquiring unit 15b acquires responses to a questionnaire regarding the learning environment of an organization. For example, the acquiring unit 15b acquires PLE questionnaire results 21 answered by members of each predetermined organization such as a company, a department, or a division. At this time, the acquiring unit 15b may acquire data of the PLE questionnaire results 21 answered anonymously by the members via the communication unit 13. The acquiring unit 15b may also acquire data that compiles the PLE questionnaire results answered by the members of each organization.

[0047] (Presentation part 15c) The presentation unit 15c presents at least one of an analysis result 31 of the organization and a proposed comment 32 for growing the organization, which are obtained using the questionnaire result 21 and the machine learning model 14d. Here, the machine learning model 14d is a machine learning model trained by the generation unit 15a so as to output data corresponding to at least one of the analysis result 31 and the proposed comment 32 when data corresponding to the questionnaire result 21 is input.

[0048] For example, the presentation unit 15c presents a result including the scores of each of a plurality of analysis categories as the analysis result 31. Here, the survey result 21 includes answers to a plurality of survey items each belonging to one of the analysis categories. In addition, the machine learning model 14d outputs data corresponding to the analysis result 31 so that the answer to the survey item belonging to a certain analysis category is reflected not only in the score of the one analysis category but also in the scores of other analysis categories. To explain using the example of FIG. 4, for example, in the PLE survey result 21, "element a", which is a survey item included in "category A", is also related to "element b", which is a survey item included in "category B". Therefore, when outputting the score of "category B", the machine learning model 14d outputs the analysis result 31 so that the answer to the related "element a" is reflected in addition to the answer to "element b".

[0049] Moreover, the presenting unit 15c presents a comment proposing measures to improve the score of the analysis category as the proposed comment 32. Moreover, the machine learning model 14d outputs data corresponding to the proposed comment 32 so that the answer to the questionnaire item belonging to a certain analysis category is reflected not only in the proposed measures to improve the score of the certain analysis category but also in the proposed measures to improve the scores of other analysis categories. To explain using the example of FIG. 4 described above, when outputting a proposal to improve the score of "Category B", the machine learning model 14d outputs the proposed comment 32 so that it is reflected not only in the "element b" of "Category B" and the "element k" and "element p" in the lower layer, but also in the proposed measures to improve the answers of "element a" of "Category A" having a relation thereto and the "element j" and "element o" in the lower layer.

[0050] In addition, the presenting unit 15c presents, as the proposed comment 32, a comment proposing a measure to improve the score of the analysis category based on the graph showing the analysis result 31. For example, the presenting unit 15c presents a comment proposing a measure to increase the area of ​​a radar chart showing the scores of a plurality of analysis categories. To explain using the example of the analysis result 31 in FIG. 1 described above, the machine learning model 14d presents a comment proposing a measure to maximize the area of ​​the pentagon in the radar chart of the "corporate development status analysis" of "Company B". In addition, the presenting unit 15c can also present, as the proposed comment 32, a comment proposing a measure to increase the height or area of ​​a line graph or bar graph showing the achievement level by question, or to increase the area of ​​a specific category in a pie chart or bar graph showing the achievement level by analysis category.

[0051] In addition, the presentation unit 15c presents comments proposing measures to improve the scores of the analysis categories, including at least one of rubric development, human resource development strategy formulation, workshops, training design, practical support, and coaching for management.

[0052] Here, the multiple analysis categories include an analysis category relating to innovation encouragement, an analysis category relating to learning, an analysis category relating to systems, an analysis category relating to the field, and an analysis category relating to management organizational climate.

[0053] (2-2. Specific examples of presented information 30) Next, the presented information 30 acquired by the presenting unit 15c will be described in detail with reference to Fig. 6 to Fig. 10. Fig. 6 to Fig. 10 are diagrams showing an example of the presented information according to the first embodiment. In the following, the presented information 30 will be described in the order of the analysis result 31 and the proposed comment 32.

[0054] (2-2-1. Analysis result 31) An example of the analysis result 31 will be described with reference to Fig. 6. As shown in Fig. 6, the presenting unit 15c acquires, as the analysis result 31, the number and percentage of "Yes" responses to each question in the PLE questionnaire result 21. The example in Fig. 6 shows the number and percentage of responses that answered "Yes" to each of "Question 1" to "Question 30" in the PLE questionnaire 20 shown in Fig. 2, out of the total number of responses "n=18".

[0055] An example of the analysis result 31 will be described with reference to Fig. 7. As shown in Fig. 7, the presenting unit 15c acquires, as the analysis result 31, a table of evaluation values ​​for each analysis category output by the machine learning model 14d to which the PLE questionnaire result 21 has been input. In the example of Fig. 7, for each of "Category A" to "Category E", a "current evaluation value (%)" which is the evaluation value of the organization to be evaluated and an "overall result (%)" which is the overall evaluation value of the organization to be evaluated are shown.

[0056] At this time, the machine learning model 14d outputs an analysis result 31 including scores for the above analysis categories, such as an analysis category related to innovation encouragement, an analysis category related to learning, an analysis category related to systems, an analysis category related to the environment, and an analysis category related to management organizational culture.

[0057] An example of the analysis result 31 will be described with reference to Fig. 8. As shown in Fig. 8, the presenting unit 15c acquires the answer results of "Yes" to each question of the PLE questionnaire result 21 as the analysis result 31. The example in Fig. 8 shows the "evaluation result" (shown by a solid line) which is the proportion of the organization to be evaluated that answered "Yes" to each of "Question 1" to "Question 30" in the PLE questionnaire 20 shown in Fig. 2, and the "overall result" (shown by a dashed line) which is the overall proportion of the organization to be evaluated.

[0058] An example of the analysis result 31 will be described with reference to Fig. 9. As shown in Fig. 9, the presenting unit 15c acquires, as the analysis result 31, a radar chart of the evaluation value for each analysis category output by the machine learning model 14d to which the PLE questionnaire result 21 has been input. In the example of Fig. 9, for each of "Category A" to "Category E", a radar chart is used to show the "Current Evaluation Value (%)" (shown by a solid line) which is the evaluation value of the organization to be evaluated, and the "Overall Result (%)" (shown by a dashed line) which is the overall evaluation value of the organization to be evaluated.

[0059] (2-2-2. Proposal Comment 32) An example of the proposed comment 32 will be described with reference to FIG. 10. As shown in FIG. 10, the presenting unit 15c acquires, as the proposed comment 32, a comment for growing the organization output by the machine learning model 14d to which the PLE questionnaire result 21 has been input. In the example of FIG. 10, the first comment is that there is an issue in the "XX area (category)", and that there is room for improvement in particular in "△△" and "□□". The second comment is that there is an issue in the "XX area (category)", and that there is room for improvement in particular in "※※".

[0060] At this time, the machine learning model 14d outputs, for example, a proposed comment 32 proposing a measure to increase the area of ​​the radar chart showing the score of the above analysis category. In addition, the machine learning model 14d outputs a comment including, for example, rubric development, human resource development strategy formulation, workshop, training design, practical support, coaching for management, etc. as the above measure.

[0061] [3. Flow of each process] The flow of each process according to the first embodiment will be described with reference to Fig. 11. Fig. 11 is a flowchart showing the flow of the entire provision process according to the first embodiment. The following describes the flow of the survey result acquisition process, presentation information generation process, and information presentation process according to the first embodiment in that order.

[0062] (3-1. Survey result acquisition process flow) The acquiring unit 15b acquires the PLE survey result 21 (step S101), and stores the PLE survey result 21 in the storage unit 14 (step S102).

[0063] (3-2. Presentation information generation process flow) The presentation unit 15c acquires the analysis result 31 output by the machine learning model 14d (step S103), acquires the proposed comment 32 output by the machine learning model 14d (step S104), and stores the acquired analysis result 31 and proposed comment 32 in the memory unit 14 (step S105).

[0064] (3-3. Information presentation process flow) The presenting unit 15c selects information to be presented from the acquired analysis results 31 and proposed comments 32 (step S106), presents the selected information (step S107), and ends the process.

[0065] The order and execution timing of the above-mentioned steps S101 to S107 can be changed dynamically or statically. Also, some of the steps S101 to S107 may be omitted.

[0066] [4. Effects of the embodiment] First, in the process according to the above-described first embodiment, a survey result 21 regarding the learning environment of an organization is acquired, and at least one of an analysis result 31 of the organization and a proposed comment 32 for growing the organization obtained using the survey result 21 and the machine learning model 14d is presented. At this time, when data corresponding to the survey result 21 is input, the machine learning model 14d outputs data corresponding to at least one of the analysis result 31 and the proposed comment 32. Therefore, in this process, it is possible to effectively make a proposal for improving the organization based on the aggregated results of the PLE survey.

[0067] Secondly, in the process according to the above-mentioned first embodiment, the analysis result 31 includes scores for each of a plurality of analysis categories, the questionnaire result 21 includes responses to a plurality of questionnaire items each belonging to one of the analysis categories, and the machine learning model 14d outputs data corresponding to the analysis result 31 such that a response to a questionnaire item belonging to one analysis category is reflected not only in the score for that one analysis category but also in the scores for other analysis categories. Therefore, in this process, the scores for the analysis categories are calculated based on the aggregated results of the PLE questionnaire, and a proposal for effectively improving the organization can be made.

[0068] Thirdly, in the process according to the above-mentioned first embodiment, the analysis result 31 includes scores for each of a plurality of analysis categories, the questionnaire result 21 includes responses to a plurality of questionnaire items each belonging to one of the analysis categories, the proposed comment 32 proposes measures for improving the scores of the analysis categories, and the machine learning model 14d outputs data corresponding to the proposed comment 32 so that the response to the questionnaire item belonging to a certain one analysis category is reflected not only in the proposed measures for improving the score of the one analysis category but also in the proposed measures for improving the scores of other analysis categories. Therefore, in this process, the scores of the analysis categories can be calculated more effectively based on the aggregated results of the PLE questionnaire, and a proposal for improving the organization can be made effectively.

[0069] Fourth, in the process according to the above-described first embodiment, the proposed comments 32 propose measures to improve the scores of the analysis categories based on the graph showing the analysis results 31. Therefore, in this process, the scores of the analysis categories can be calculated more quantitatively based on the aggregated results of the PLE questionnaire, and proposals can be made to effectively improve the organization.

[0070] Fifth, in the process according to the above-mentioned first embodiment, the measures for improving the scores of the analysis categories include at least one of rubric development, human resource development strategy formulation, workshops, training design, practical support, and coaching for management. Therefore, in this process, it is possible to make more specific proposals for improving the organization based on the aggregated results of the PLE questionnaire.

[0071] Sixth, in the process according to the first embodiment described above, the multiple analysis categories include an analysis category related to innovation encouragement, an analysis category related to learning, an analysis category related to systems, an analysis category related to places, and an analysis category related to management organizational culture. Therefore, in this process, based on the aggregated results of the PLE questionnaire, it is possible to calculate the scores of the analysis categories more specifically and make proposals for effectively improving the organization.

[0072] Seventh, in the process according to the above-described embodiment 1, the survey results are responses to a PLE survey. Therefore, in this process, it is possible to effectively make proposals for improving an organization based on the collected results of a widely used PLE survey.

[0073] [Embodiment 2] Meanwhile, the machine learning model 14d described in the first embodiment can maintain the accuracy of the presentation information 30 output by the machine learning model 14d at a high level by training it through reinforcement learning. Therefore, in the second embodiment, an example will be described in which the providing device 10-2 maintains the accuracy of the machine learning model 14d through reinforcement learning using a PLE questionnaire result after education is provided according to the presentation information 30 generated by the machine learning model 14d.

[0074] The configuration of a providing system 100-2 and a providing device 10-2 according to the second embodiment, and the flow of each process will be described below in order, and finally, the effect of the second embodiment will be described. Note that the description of the same configuration and process as those of the first embodiment will be omitted.

[0075] [1. Configuration of the provision system 100-2] The configuration of a provision system (hereinafter, also referred to as the present system, as appropriate) 100-2 according to the second embodiment will be described in detail with reference to Fig. 12. Fig. 12 is a diagram showing an example of the configuration of the provision system according to the second embodiment. Below, an example of the overall configuration of the present system 100-2 will be shown, and then the processing of the present system 100-2 will be described.

[0076] (1-1. Example of overall system configuration) The system 100-2 includes a providing device 10-2. The system 100-2 may include a plurality of providing devices 10-2. The system 100-2 includes a pre-education PLE questionnaire result 21-1 and a post-education PLE questionnaire result 21-2 as data to be input to the providing device 10-2, and presentation information 30 including an analysis result 31 and a proposed comment 32 as data to be output by the providing device 10-2.

[0077] (1-2. Overall system processing) The providing device 10-2 in such a system is an example of a computer device that has a trained machine learning model 14d and generates and outputs presentation information 30 from PLE questionnaire results 21 (21-1, 21-2) using the machine learning model 14d. The following describes the pre-training input process, output process, post-training input process, and learning process executed by the providing device 10-2 in that order. Note that the machine learning model 14d possessed by the providing device 10-2 is common to the machine learning model 14d possessed by the providing device 10-1 of embodiment 1, and therefore description thereof will be omitted.

[0078] (1-2-1. Pre-education input processing) First, the providing device 10-2 accepts input of the pre-education PLE questionnaire result 21-1 (step S11). At this time, the pre-education PLE questionnaire result 21-1 accepted by the providing device 10-2 may be data compiled for each item of the PLE questionnaire, or may be data before compilation transmitted by a terminal device owned by the survey respondent.

[0079] (1-2-2. Output processing) Secondly, the providing device 10-2 outputs the presentation information 30 including the analysis result 31 and the proposed comment 32 acquired from the machine learning model 14d (step S12). At this time, the providing device 10-2 inputs the pre-training PLE questionnaire result 21-1 to the machine learning model 14d to acquire the presentation information 30. For example, the machine learning model 14d performs the following operations for the PLE questionnaire result 21: tallying up responses for each questionnaire item, comparing results with other departments and divisions and other companies, evaluating each analysis category, and identifying elements that need improvement. Then, the presentation information 30 including the output analysis result 31 and the proposed comment 32 is presented to the companies that responded to the pre-training PLE questionnaire result 21-1, and is used as an organizational improvement measure in corporate development. Note that the presentation information 30 output by the providing device 10-2 shown in FIG. 12 is the same as the presentation information 30 output by the providing device 10-1 shown in FIG. 1, so a description thereof will be omitted.

[0080] (1-2-3. Post-training input processing) Thirdly, the providing device 10-2 accepts input of the post-education PLE questionnaire result 21-2 (step S13). For example, the providing device 10-2 accepts input of the post-education PLE questionnaire result 21-2 from a company or the like after education has been provided based on the presented information 30 corresponding to the pre-education PLE questionnaire result 21-1. In addition, at this time, the providing device 10-2 accepts input of the achievement level of the analysis category targeted by the organization such as the company after education has been provided based on the presented information 30.

[0081] (1-2-4. Learning process) Fourth, the providing device 10-2 trains the machine learning model 14d based on the post-education PLE questionnaire result 21-2 (step S14). For example, the providing device 10-2 sets a reward based on the difference between the evaluation value for each analysis category targeted by an organization such as a company and the post-education PLE questionnaire result 21-2, and performs reinforcement learning on the machine learning model 14d.

[0082] By executing the above-mentioned steps S11 to S14, the PLE questionnaire results can be used to properly grasp the current problems of the organization and to determine effective education policies and human resource development strategies. In particular, by executing the above-mentioned steps S13 to S14, it is expected that the machine learning model based on the PLE questionnaire results will evolve, and more effective education policies and human resource development strategies can be determined.

[0083] (1-3. Specific example of the learning process of machine learning model 14d) A specific example of the learning process in step S14 described above will be described with reference to Fig. 13 to Fig. 16. Fig. 13 to Fig. 16 are diagrams showing an example of the learning process according to embodiment 2. The learning processes related to the questionnaire items, the radar chart, and the logic tree will be described in this order below.

[0084] (1-3-1. Learning process based on questionnaire items) A specific example of the learning process in step S14 described above will be described with reference to Fig. 13. As shown in Fig. 13, the providing device 10-2 compares a pre-education PLE questionnaire result 21-1 with a post-education PLE questionnaire result 21-2, sets rewards so that answers to the questionnaire items are improved, and performs reinforcement learning.

[0085] FIG. 13 shows pre-training survey results 21-1 (shown by shading) and post-training survey results 21-2 (shown by diagonal lines) for each question of the PLE questionnaire 20 by the same respondents (n=31). For example, for "Question 1," 25 out of 31 people answered "Yes" in the pre-training survey result 21-1, and 30 out of 31 people answered "Yes" in the post-training survey result 21-2, and it can be seen that 5 people improved before and after the training. In this way, when organizational improvement is achieved by presenting presented information 30 such as suggested comments 32 and providing education based on the suggested comments 32, it is preferable that the number of "Yes" responses for each questionnaire item increases more.

[0086] Here, the providing device 10-2 compares the pre-education PLE questionnaire result 21-1 with the post-education PLE questionnaire result 21-2, sets a reward so as to increase the number of respondents who selected "Yes" for the questionnaire item belonging to the analysis category that the company or the like aims to improve, and performs reinforcement learning. That is, the providing device 10-2 performs reinforcement learning by giving a higher reward as the increase in the number of respondents is larger, and a lower reward as the increase in the number of respondents is smaller. In addition, the providing device 10-2 can also perform reinforcement learning by giving a negative reward when the number of respondents is decreasing. That is, the providing device 10-2 trains the machine learning model 14d so as to output presentation information 30 such as the suggestion comment 32 for increasing the number of "Yes" responses to each questionnaire item.

[0087] (1-3-2. Learning process based on radar chart) A specific example of the learning process in step S14 described above will be described with reference to Fig. 14. As shown in Fig. 14, the providing device 10-2 compares a pre-education PLE questionnaire result 21-1 with a post-education PLE questionnaire result 21-2, sets a reward so as to increase the area of ​​a radar chart showing the score of the analysis category, and performs reinforcement learning.

[0088] FIG. 14 shows a radar chart of the scores (shown by dashed lines) before and after training (shown by solid lines) for each analysis category of the PLE questionnaire 20 by the same respondent. For example, in "Category A," the evaluation value was about 65% before the education and was 80% or more after the education, which shows an improvement of about 15% before and after the education. Similarly, the evaluation values ​​also increased in "Category B" to "Category E," and it can be seen that the area of ​​the pentagon in the radar chart increases as the scores of the analysis categories improve. In this way, it is preferable that the area of ​​the radar chart showing the scores of the analysis categories is further increased by presenting the presented information 30 such as the proposed comments 32. Note that the radar chart in FIG. 14 is an example, and may be a rectangle or a hexagon depending on the number of analysis categories, and the shape of the radar chart is not particularly limited.

[0089] Here, the providing device 10-2 compares the pre-training PLE questionnaire result 21-1 with the post-training PLE questionnaire result 21-2, sets a reward so that the area of ​​the radar chart indicated by the score of the analysis category that the company or the like aims to improve increases, and performs reinforcement learning. That is, the providing device 10-2 performs reinforcement learning by giving a higher reward as the increased area of ​​the radar chart increases, and a lower reward as the increased area of ​​the radar chart decreases. In addition, the providing device 10-2 can also perform reinforcement learning by giving a negative reward when the area of ​​the radar chart decreases. That is, the providing device 10-2 trains the machine learning model 14d so as to output presentation information 30 such as a proposed comment 32 for increasing the area of ​​the radar chart of the analysis category.

[0090] (1-3-3. Learning process based on logic tree) A specific example of the learning process in step S14 described above will be described with reference to Fig. 15 and Fig. 16. As shown in Fig. 15 and Fig. 16, the providing device 10-2 performs learning so as to change the priority order of improvement measures presented in the output presentation information 30, based on the relationship between questionnaire items belonging to an analysis category.

[0091] FIG. 15 and FIG. 16 show the relationship between the analysis categories "Category A" to "Category E". As explained in FIG. 4, the logic tree classifies the elements of the questionnaire questions or question groups into related analysis categories and shows the causal relationships and priorities. For example, the analysis categories are the PLE questionnaire items classified into "innovation encouragement", "learning", "system", "place", "management organizational climate", etc. Also, the positional relationship between "Category A" to "Category E" shown in FIG. 15 and FIG. 16 shows that the higher the hierarchy, the more difficult it is for an organization to achieve. For example, FIG. 15 shows that "Category D" cannot be achieved unless "Category C", which is lower in the hierarchy, is achieved. Also, it can be seen that in order to achieve "Category C", it is necessary to achieve "Category A", then "Category B", and then "Category E". In this way, it is preferable to present the presented information 30 such as the proposed comments 32 that take into account the relationships (causal relationships and priorities) between the above analysis categories.

[0092] Here, the providing device 10-2 can perform machine learning to compare the pre-education PLE questionnaire result 21-1 with the post-education PLE questionnaire result 21-2, and when the score of the analysis category that the company or the like requests to improve has not improved, present a proposal to improve a different analysis category. For example, using the example of FIG. 15, when the analysis category that the company or the like aims to improve is "category C" and an improvement measure for "category C" is presented, the providing device 10-2 compares the pre-education PLE questionnaire result 21-1 with the post-education PLE questionnaire result 21-2, and when the score of the analysis category of "category C" has not improved, the providing device 10-2 performs learning of the machine learning model 14d to present improvement measures for "category B" and "category E" that are lower layers that have a causal relationship with "category C". At this time, the providing device 10-2 can also perform learning of the machine learning model 14d to change the relationship of the logic tree shown in FIG. 15. That is, the providing device 10-2 performs learning of the machine learning model 14d to change the priority order of the improvement measures and output presentation information 30 such as a more preferable proposed comment 32.

[0093] The providing device 10-2 can also perform the learning process shown in the example of FIG. 15 by reinforcement learning using a reward. For example, in the example of FIG. 16, when the providing device 10-2 presents improvement measures for "category B" or "category E" in order to improve "category C" but does not improve, the providing device 10-2 performs reinforcement learning by giving a low reward. On the other hand, when the providing device 10-2 presents an improvement measure for "category A" in order to improve "category C" and improves it, the providing device 10-2 performs reinforcement learning by giving a high reward. Specifically, the providing device 10-2 sets a lower reward as the difference between the target value of the score of the analysis category that the company or the like requests to improve and the score of the analysis category based on the post-education PLE questionnaire result 21-2 is larger. In other words, when the providing device 10-2 presents an improvement measure that is closer to the achievement level required by the company or the like, the providing device 10-2 performs reinforcement learning by giving a higher reward. At this time, the providing device 10-2 can also perform reinforcement learning so as to change the relationship from the logic tree shown in FIG. 15 to the logic tree shown in FIG. 16. That is, the providing device 10-2 changes the priority order of improvement measures through reinforcement learning, and trains the machine learning model 14d so as to output more preferable presented information 30 such as the proposed comment 32.

[0094] [2. Configuration of the providing device 10-2] The functional configuration of the providing device 10-2 included in the providing system 100-2 shown in FIG. 12 will be described with reference to FIG. 17. FIG. 17 is a block diagram showing a configuration example of the providing device according to the second embodiment. The configuration of the providing device 10-2 according to the second embodiment will be described in detail below. The providing device 10-2 has an input unit 11, an output unit 12, a communication unit 13, a storage unit 14, and a control unit 15-2. Note that the input unit 11, the output unit 12, the communication unit 13, and the storage unit 14 have the same configuration and process as those of the first embodiment described above, and therefore will not be described here.

[0095] (Control unit 15-2) The control unit 15-2 controls the entire providing device 10-2. The control unit 15-2 includes a generating unit 15a, an acquiring unit 15b, a presenting unit 15c, and a learning unit 15d. Here, the control unit 15-2 can be realized by, for example, an electronic circuit such as a CPU or an MPU, or an integrated circuit such as an ASIC or an FPGA.

[0096] (Generation unit 15a) The generation unit 15a inputs learning data to the machine learning model 14d to generate a trained model. For example, the generation unit 15a inputs "PLE questionnaire result, presented information" as learning data "input data, correct answer information" to the machine learning model 14d, and executes learning of the machine learning model 14d so that the output result of the machine learning model 14d matches the correct answer information. The generation unit 15a acquires the learning data from the training data storage unit 14a. Also, as described above, the analysis content by the machine learning model 14d can be changed by changing the correct answer information.

[0097] (Acquisition part 15b) The acquiring unit 15b acquires a first survey result (pre-training PLE survey result) 21-1, which is a survey result on the learning environment conducted before the implementation of an improvement measure for improving the organization. For example, the acquiring unit 15b acquires the pre-training PLE survey result 21-1 answered by members of each predetermined organization such as a company, a department, a division, etc. Also, the acquiring unit 15b acquires a second survey result (post-training PLE survey result) 21-2, which is a survey result conducted after the implementation of an improvement measure based on the output result of the machine learning model 14d in response to the input of the first survey result. For example, the acquiring unit 15b acquires the post-training PLE survey result 21-2 answered by members of each predetermined organization such as a company, a department, a division, etc.

[0098] At this time, the acquiring unit 15b may acquire data on the PLE survey results 21 (21-1, 21-2) answered anonymously by the members via the communication unit 13. In addition, the acquiring unit 15b may acquire data obtained by aggregating the PLE survey results 21 (21-1, 21-2) answered by the members of each organization.

[0099] (Presentation part 15c) The presentation unit 15c presents at least one of the analysis result 31 of the organization and the proposed comment 32 for growing the organization, which are obtained by using the PLE questionnaire result 21 (21-1, 21-2) and the machine learning model 14d. For example, the presentation unit 15c presents a result including the scores of each of a plurality of analysis categories as the analysis result 31. Here, the questionnaire result 21 includes answers to a plurality of questionnaire items each belonging to one of the analysis categories. In addition, the machine learning model 14d outputs data corresponding to the analysis result 31 so that the answers to the questionnaire items belonging to a certain one analysis category are reflected not only in the score of the one analysis category but also in the scores of other analysis categories. Note that the process of the presentation unit 15c and the presentation information 30 presented by the presentation unit 15c are the same as those in the above-mentioned embodiment 1, and therefore description thereof will be omitted.

[0100] (Learning section 15d) The learning unit 15d executes machine learning of the machine learning model 14d by reinforcement learning using a reward based on the post-education PLE questionnaire result 21-2 and a target state. For example, the learning unit 15d sets a reward based on the difference between the score (evaluation value) of an analysis category targeted by an organization such as a company and the score of the analysis category indicated by the post-education PLE questionnaire result 21-2, and performs reinforcement learning. To explain using a specific example, when the target score of "Category A" is 90% and the scores of "Category A" indicated by the post-education PLE questionnaire result 21-2 are 85% and 80%, the learning unit 15d performs reinforcement learning so that the score of 85%, which has a smaller difference from 90%, is deemed to have achieved the target score and is given a higher reward.

[0101] Here, the PLE questionnaire 20 includes a plurality of questionnaire items each belonging to one of a plurality of analysis categories as shown in the logic tree of Fig. 4. Similarly, the plurality of analysis categories are set with causal relationships (relationships between causes and the results that result from them) based on the relationships between the questionnaire items that affect each other.

[0102] Furthermore, the learning unit 15d compares the pre-education PLE questionnaire result 21-1 with the post-education PLE questionnaire result 21-2, identifies changes in multiple questionnaire items belonging to the analysis category to be improved, and executes machine learning of the machine learning model 14d by reinforcement learning using rewards based on the changes. For example, the learning unit 15d sets a reward based on the difference between the score of the questionnaire item in the pre-education PLE questionnaire result 21-1 (the number of respondents who answered "Yes") and the score of the questionnaire item in the post-education PLE questionnaire result 21-2, and performs reinforcement learning. To explain using a specific example, when comparing the difference in the scores between the pre-education PLE questionnaire result 21-1 and the post-education PLE questionnaire result 21-2, reinforcement learning is performed so that a higher reward is given to the questionnaire item with a difference of "10" compared to the questionnaire item with a difference of "4".

[0103] The learning unit 15d identifies a difference between a graph showing scores of a plurality of analysis categories using a plurality of questionnaire items included in the pre-education PLE questionnaire result 21-1 and the corresponding graph in the post-education PLE questionnaire result 21-2, and executes machine learning of the machine learning model 14d by reinforcement learning using a reward based on the difference. For example, the learning unit 15d executes machine learning of the machine learning model 14d by reinforcement learning that gives a higher reward the larger the area of ​​a radar chart in the post-education PLE questionnaire result 21-2 is compared with the area of ​​a radar chart showing scores of a plurality of analysis categories using a plurality of questionnaire items included in the pre-education PLE questionnaire result 21-1. To explain using a specific example, when comparing the score difference between the pre-education PLE questionnaire result 21-1 and the post-education PLE questionnaire result 21-2, reinforcement learning is performed so that a higher reward is given to a radar chart with an area difference of "9" when compared at a certain size.

[0104] The learning unit 15d can also identify differences in graphs such as line graphs, bar graphs, pie charts, and band graphs before and after training as graphs showing scores of multiple analysis categories, and execute machine learning of the machine learning model 14d by reinforcement learning using rewards based on the differences. For example, the learning unit 15d can execute reinforcement learning so that a higher reward is given as the height or area of ​​a specific category is increased in a line graph, bar graph, pie chart, band graph, or the like showing the achievement level by analysis category.

[0105] The learning unit 15d executes machine learning of the machine learning model 14d so as to present an improvement measure for a second analysis category having a causal relationship with the first analysis category when the post-education PLE questionnaire result 21-2 obtained after the implementation of an improvement measure for improving a first analysis category among the multiple analysis categories has not improved from the pre-education PLE questionnaire result 21-1. For example, when the score of the analysis category of the post-education PLE questionnaire result 21-2 has not improved compared to the pre-education PLE questionnaire result 21-1, the learning unit 15d performs machine learning so as to present an improvement measure for an analysis category that is a different analysis category and has a causal relationship. To explain using a specific example, when it is desired to improve "category C" as an analysis category, and the proposed improvement measure for "category C" does not result in improvement, the learning unit 15d performs machine learning so as to present improvement measures for "category B" and "category E" having a causal relationship with "category C".

[0106] The learning unit 15d executes machine learning of the machine learning model 14d so that the priority of the analysis category to be improved based on the causal relationship is changed by reinforcement learning using a smaller reward as the difference between the target state and the post-education PLE questionnaire result 21-2 for the first analysis category is larger than the difference between the target state and the pre-education PLE questionnaire result 21-1. For example, the learning unit 15d calculates the difference between the score (evaluation value) of the analysis category targeted by an organization such as a company and the score of the analysis category indicated by the pre-education PLE questionnaire result 21-1, and similarly calculates the difference between the score (evaluation value) of the target analysis category and the score of the analysis category indicated by the post-education PLE questionnaire result 21-2, sets a reward based on the size of the two differences, and performs reinforcement learning. To explain using a specific example, if the target score for "Category A" is 90%, the pre-education PLE questionnaire result 21-1 indicates a score for "Category A" of 60%, and the post-education PLE questionnaire result 21-2 indicates a score for "Category A" of 50%, the latter has a lower evaluation value and a larger difference, so it can be determined to be inappropriate. In other words, since the priority order of the logic tree, etc. is inappropriate, the learning unit 15d performs reinforcement learning so as to give a lower reward and changes the priority order to obtain an appropriate logic tree.

[0107] [3. Flow of each process] The flow of each process according to the second embodiment will be described with reference to Fig. 18. Fig. 18 is a flowchart showing the flow of the entire provision process according to the second embodiment. The following describes the flow of the survey result acquisition process, presentation information generation process, information presentation process, and survey result learning process according to the second embodiment in that order.

[0108] (3-1. Survey result acquisition process flow) The acquiring unit 15b acquires the pre-training PLE questionnaire result 21-1 (step S201), and stores the pre-training PLE questionnaire result 21-1 in the storage unit 14 (step S202).

[0109] (3-2. Presentation information generation process flow) The presentation unit 15c acquires the analysis result 31 output by the machine learning model 14d (step S203), acquires the proposed comment 32 output by the machine learning model 14d (step S204), and stores the acquired analysis result 31 and proposed comment 32 in the memory unit 14 (step S205).

[0110] (3-3. Information presentation process flow) The presenting unit 15c selects information to be presented from the acquired analysis results 31 and proposed comments 32 (step S206), and presents the selected information (step S207).

[0111] (3-4. Survey result learning process flow) The acquisition unit 15b acquires the post-education PLE questionnaire result 21-2 (step S208). Next, the learning unit 15d compares the pre-education PLE questionnaire result 21-1 with the post-education PLE questionnaire result 21-2 (step S209), and performs reinforcement learning by providing a reward (step S210). Thereafter, if the reinforcement learning is to be ended (step S211: Yes), the learning unit 15d ends the processing. On the other hand, if the reinforcement learning is not to be ended (step S211: No), the learning unit 15d proceeds to the processing of step S201.

[0112] The order and execution timing of the above-mentioned steps S201 to S211 can be changed dynamically or statically. Also, some of the steps S201 to S211 may be omitted.

[0113] [4. Effects of the embodiment] First, in the process according to the above-mentioned second embodiment, a pre-education PLE questionnaire result 21-1, which is the result of a questionnaire about the learning environment conducted before the implementation of an improvement measure for improving the organization, is obtained, a post-education PLE questionnaire result 21-2, which is the result of a questionnaire conducted after the implementation of an improvement measure based on the output result of the machine learning model 14d in response to the input of the pre-education PLE questionnaire result 21-1, is obtained, and machine learning of the machine learning model 14d is performed based on the post-education PLE questionnaire result 21-2. Therefore, in this process, it is possible to make proposals for improving the organization more effectively based on the aggregated results of the PLE questionnaire.

[0114] Secondly, in the process according to the second embodiment described above, machine learning of the machine learning model 14d is performed by reinforcement learning using a reward based on the post-training PLE questionnaire result 21-2 and a target state. Therefore, in this process, it is possible to make a proposal for improving the organization more effectively by reinforcement learning using a reward based on the aggregated results of the PLE questionnaire.

[0115] Thirdly, in the process according to the second embodiment described above, the pre-training PLE questionnaire result 21-1 and the post-training PLE questionnaire result 21-2 are compared to identify changes in multiple questionnaire items belonging to the analysis category to be improved, and machine learning is performed on the machine learning model 14d by reinforcement learning using rewards based on the changes. Therefore, in this process, based on the aggregated results of the PLE questionnaire, it is possible to make proposals for more effectively improving the organization by reinforcement learning using rewards that take into account the changes in the questionnaire items.

[0116] Fourth, in the process according to the above-described second embodiment, machine learning of the machine learning model 14d is performed by reinforcement learning that gives a higher reward the larger the area of ​​a radar chart showing scores of a plurality of analysis categories using a plurality of questionnaire items included in the pre-training PLE questionnaire result 21-1 is compared to the area of ​​the radar chart. Therefore, in this process, based on the aggregated results of the PLE questionnaire, it is possible to make proposals for improving the organization more effectively by reinforcement learning using a reward that takes into account changes in the radar chart.

[0117] Fifth, in the process according to the above-described second embodiment, when the post-education PLE questionnaire result 21-2 obtained after the implementation of an improvement measure for improving a first analysis category among the multiple analysis categories has not improved from the pre-education PLE questionnaire result 21-1, machine learning of the machine learning model 14d is executed so as to present an improvement measure for a second analysis category having a causal relationship with the first analysis category. Therefore, in this process, based on the aggregated results of the PLE questionnaire, machine learning considering the logic tree can be used to make a proposal for improving the organization more effectively.

[0118] Sixth, in the process according to the above-mentioned second embodiment, the machine learning of the machine learning model 14d is executed so that the greater the difference between the survey result of the target state for the first analysis category and the post-education PLE survey result 21-2 is between the survey result of the target state and the pre-education PLE survey result 21-1, the greater the priority order of the analysis category to be improved based on the causal relationship is changed by reinforcement learning using a smaller reward. Therefore, in this process, based on the aggregated results of the PLE survey, a proposal for improving the organization more effectively can be made by reinforcement learning using a reward that takes the logic tree into consideration.

[0119] Seventh, in the process according to the above-mentioned embodiment 2, the survey results are responses to a PLE survey. Therefore, in this process, it is possible to make more effective suggestions for improving an organization based on the collected results of a widely used PLE survey.

[0120] 〔system〕 The information including the processing procedures, control procedures, specific names, various data and parameters shown in the above documents and drawings can be changed arbitrarily unless otherwise specified.

[0121] In addition, each component of each device shown in the figure is a functional concept, and does not necessarily have to be physically configured as shown in the figure. In other words, the specific form of distribution and integration of each device is not limited to that shown in the figure. In other words, all or part of them can be functionally or physically distributed and integrated in any unit according to various loads, usage conditions, etc.

[0122] Furthermore, each processing function performed by each device may be realized, in whole or in part, by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware using wired logic.

[0123] [Hardware] Next, an example of the hardware configuration of the providing device 10 (10-1, 10-2) will be described. Note that the other devices may have a similar hardware configuration. FIG. 19 is a diagram for explaining an example of the hardware configuration. As shown in FIG. 19, the providing device 10 (10-1, 10-2) has a communication device 10a, a HDD (Hard Disk Drive) 10b, a memory 10c, and a processor 10d. In addition, each unit shown in FIG. 19 is connected to each other by a bus or the like.

[0124] The communication device 10a is a network interface card or the like, and communicates with other servers. The HDD 10b stores programs and DBs that operate the functions shown in Figs.

[0125] The processor 10d reads out a program that executes the same processes as the respective processing units shown in FIG. 5 and FIG. 17 from the HDD 10b or the like and loads it in the memory 10c, thereby operating a process that executes the respective functions described in FIG. 5, FIG. 17, or the like. For example, this process executes the same functions as the respective processing units possessed by the providing device 10-1 and the providing device 10-2. Specifically, the processor 10d reads out a program having the same functions as the generating unit 15a, the acquiring unit 15b, the presenting unit 15c, the learning unit 15d, or the like from the HDD 10b or the like. Then, the processor 10d executes a process that executes the same processes as the generating unit 15a, the acquiring unit 15b, the presenting unit 15c, the learning unit 15d, or the like.

[0126] In this way, the providing device 10-1 and the providing device 10-2 operate as devices that execute various processing methods by reading and executing the programs. The providing device 10-1 and the providing device 10-2 can also realize the same functions as the above-mentioned embodiment by reading the above-mentioned programs from a recording medium using a medium reading device and executing the read programs. Note that the programs in these other embodiments are not limited to being executed by the providing device 10-1 and the providing device 10-2. For example, the present invention can be similarly applied to cases where other computers or servers execute the programs, or where these execute the programs in cooperation with each other.

[0127] This program can be distributed via a network such as the Internet. This program can also be recorded on a computer-readable recording medium such as a hard disk, a flexible disk (FD), a CD-ROM, a magneto-optical disk (MO), or a digital versatile disk (DVD), and can be executed by being read out from the recording medium by a computer. [Explanation of symbols]

[0128] 10, 10-1, 10-2 Provided equipment 11 Input section 12 Output section 13. Communications Department 14 Storage section 15-1, 15-2 Control section 15a Generator 15b Acquisition part 15c Presentation section 15d Learning Section 20 PLE Survey 21, 21-1, 21-2 PLE Survey Results 30 Presentation information 31, 31-1, 31-2 Analysis results 32, 32-1, 32-2 Proposal Comments 100-1, 100-2 provision system

Claims

1. The computer A process of executing machine learning of a machine learning model so as to output at least one of an analysis result of the organization and a proposed comment for growing the organization in response to input of data on the questionnaire result regarding the learning environment of the organization; A process of acquiring the survey results; A process of presenting at least one of the analysis result and the proposed comment obtained by using the acquired survey result and the machine learning model; A method of providing When data corresponding to the acquired survey results is input, the machine learning model outputs a score corresponding to the survey results using a logic tree indicating causal relationships and priorities of the survey items, and outputs data corresponding to at least one of the analysis results and the proposed comments according to the output score. How it is provided.

2. the analysis results include a score for each of a plurality of analysis categories; the survey results include responses to a plurality of survey items each belonging to one of the analysis categories; The machine learning model uses the logic tree to output data corresponding to the analysis results so that a response to a questionnaire item belonging to a certain analysis category is reflected not only in the score of the certain analysis category but also in the scores of other analysis categories. The method of claim 1.

3. the analysis results include a score for each of a plurality of analysis categories; the survey results include responses to a plurality of survey items each belonging to one of the analysis categories; The suggested comments suggest measures to improve the analysis category score; The machine learning model uses the logic tree to output data corresponding to the proposed comments so that a response to a questionnaire item belonging to a certain analysis category is reflected not only in a proposal of measures to improve the score of the certain analysis category but also in a proposal of measures to improve the scores of other analysis categories. The method of providing according to claim 1 or 2.

4. The suggestion comment suggests measures to improve the score of the analysis category based on a graph showing the analysis result. The method of claim 3.

5. The measures include at least one of rubric development, human resource development strategy formulation, workshops, training design, implementation support, and management coaching; The method of providing according to claim 3 or 4.

6. The plurality of analysis categories include an analysis category related to innovation encouragement, an analysis category related to learning, an analysis category related to systems, an analysis category related to places, and an analysis category related to management organizational climate; The method according to any one of claims 2 to 5.

7. The survey results are responses to a Positive Learning Environment (PLE) survey. The method according to any one of claims 1 to 6.

8. A generation unit that executes machine learning of a machine learning model so as to output at least one of an analysis result of the organization and a proposed comment for growing the organization in response to input of data on a questionnaire result regarding the learning environment of the organization; An acquisition unit for acquiring the survey results; a presentation unit that presents at least one of the analysis result and the proposed comment obtained by using the acquired survey result and the machine learning model; A providing device comprising: When data corresponding to the acquired survey results is input, the machine learning model outputs a score corresponding to the survey results using a logic tree indicating causal relationships and priorities of the survey items, and outputs data corresponding to at least one of the analysis results and the proposed comments according to the output score. Providing device.

9. On the computer, A process of executing machine learning of a machine learning model so as to output at least one of an analysis result of the organization and a proposed comment for growing the organization in response to input of data on the questionnaire result regarding the learning environment of the organization; A process of acquiring the survey results; A process of presenting at least one of the analysis result and the proposed comment obtained by using the acquired survey result and the machine learning model; A machine learning program that executes When data corresponding to the acquired survey results is input, the machine learning model outputs a score corresponding to the survey results using a logic tree indicating causal relationships and priorities of the survey items, and outputs data corresponding to at least one of the analysis results and the proposed comments according to the output score. Machine learning programs.

Images