Information processing method, program, and information processing device

Abstract

To provide an information processing method, and the like, for properly presenting various types of advice on actions to each of employees, in accordance with answers from the employees to a survey.SOLUTION: A computer acquires scores for a plurality of categories calculated based on answers from employees to a survey regarding a workplace. The computer receives a question regarding actions from an employee. The computer generates advice regarding actions of the employee using a language model, based on the received question and the scores corresponding to the categories.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to an information processing method, a program, and an information processing device. [Background technology]

[0002] In workplaces, engagement surveys are conducted to understand the psychological state of employees, with the aim of reducing turnover and improving productivity. Patent Document 1 discloses a technology that takes into consideration the results of an employee's personality test in addition to the employee's responses to the survey, and provides advice on actions that each employee should take, as well as actions that each employee's supervisor or a person in charge of the human resources department (hereinafter collectively referred to as supervisor) should take. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 7229494 Summary of the Invention [Problem to be solved by the invention]

[0004] When a supervisor takes action, it is important to take into consideration the personality and psychological state of the employee, even if the action is the same. Taking the wrong action could lead to a decrease in the employee's motivation or engagement. Therefore, it is necessary to understand the employee's condition as revealed from the survey results before taking appropriate action. Therefore, it is conceivable that the supervisor will request not only the action to be taken for the employee, but also explanations on points to note when taking the action and the employee's condition as revealed from the survey results. However, the technology disclosed in Patent Document 1 makes it difficult to meet such requests.

[0005] In one aspect, an object of the present invention is to provide an information processing method etc. that can appropriately present various pieces of advice regarding how to deal with each employee in accordance with the employee's responses to a survey. [Means for solving the problem]

[0006] In one aspect of the information processing method, a computer acquires scores for each of a plurality of categories calculated based on an employee's responses to a workplace survey, accepts questions regarding the employee's response, and generates advice regarding the employee's response using a language model based on the accepted questions and the scores for each category. [Effects of the Invention]

[0007] In one aspect, various pieces of advice regarding how to respond to each employee can be appropriately presented depending on the employee's responses to the survey. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is an explanatory diagram illustrating an example of the configuration of an information processing system. [Figure 2] FIG. 2 is a block diagram showing an example of the configuration of a server and a superior terminal; [Figure 3] FIG. 2 is an explanatory diagram showing an example of the record layout of a personality test DB and a survey DB. [Figure 4] FIG. 10 is an explanatory diagram illustrating an example of a record layout of an employee DB. [Figure 5] FIG. 10 is an explanatory diagram illustrating an example of a prompt creation table. [Figure 6] FIG. 10 is an explanatory diagram illustrating an example of a prompt creation table. [Figure 7] FIG. 10 is an explanatory diagram illustrating an example of a prompt creation table. [Figure 8] 10 is a flowchart illustrating an example of a procedure for providing advice. [Figure 9] FIG. 10 is an explanatory diagram showing an example of a screen. [Figure 10] FIG. 10 is an explanatory diagram showing an example of a screen. [Figure 11] FIG. 10 is an explanatory diagram showing an example of a screen. [Figure 12] FIG. 10 is an explanatory diagram illustrating an example of the configuration of a prompt creation table for an organization. [Figure 13] FIG. 10 is an explanatory diagram illustrating an example of the configuration of a prompt creation table for an organization. [Figure 14] 10 is a flowchart showing an example of a procedure for providing advice according to the second embodiment. [Figure 15] FIG. 10 is an explanatory diagram showing an example of a screen. [Figure 16] FIG. 10 is an explanatory diagram showing an example of a screen. [Figure 17] FIG. 10 is an explanatory diagram showing an example of a screen. DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, an information processing method, a program, and an information processing device according to the present disclosure will be described in detail with reference to the drawings showing embodiments thereof.

[0010] (Embodiment 1) FIG. 1 is an explanatory diagram showing an example of the configuration of an information processing system. In this embodiment, an information processing system is described that determines the status of an employee belonging to a company or the like based on the test results of a personality test and the response results of a workplace survey (questionnaire), and provides advice and commentary (explanation) for a supervisor (the employee's supervisor or a person in the human resources department, etc.) to respond (take action, support) to the employee. The information processing system of this embodiment generates various pieces of advice and commentary (explanation) using general-purpose large language models (LLMs). In this embodiment, an example is described in which Japanese is used as the language input to the large-scale language model, but other languages ​​may also be used.

[0011] The information processing system of this embodiment includes an information processing device 1 and multiple terminals 2, 2..., and each device 1, 2 is communicatively connected via a network N. The network N may be the Internet or a public telephone line network, or a local area network (LAN) established within the facility where the information processing system is installed. The information processing device 1 is a computer capable of various information processing and sending and receiving information, such as a server computer or a personal computer. In this embodiment, the information processing device 1 is assumed to be a server computer, and for simplicity, will be referred to as server 1 below. The terminal 2 is a terminal device used by the supervisor of an employee who has taken a personality test and a survey. The terminal 2 is a computer capable of various information processing and sending and receiving information, such as a personal computer, a tablet terminal, or a smartphone. It will be referred to as supervisor terminal 2 below. In addition to the supervisor terminal 2, the information processing system of this embodiment may also include an employee terminal having a configuration similar to that of the supervisor terminal 2 and used by the employee who has taken the personality test and survey.

[0012] In the information processing system of this embodiment, the server 1 transmits information on personality tests and surveys to employee terminals, collects responses to the personality tests and surveys via the employee terminals, and generates personality test and survey results based on the collected responses. The server 1 also determines the psychological state of each employee based on the results of the personality tests and surveys, and presents the results to the supervisor via the supervisor terminal 2. In this case, the server 1 receives various questions from the supervisor terminal 2 and transmits answers (advice, etc.) to the received questions to the supervisor terminal 2.

[0013] FIG. 2 is a block diagram showing an example configuration of the server 1 and the boss terminal 2. The server 1 includes a control unit 11, a main memory unit 12, a communication unit 13, and an auxiliary memory unit 14. The control unit 11 has one or more processors, such as a central processing unit (CPU), a microprocessing unit (MPU), a graphics processing unit (GPU), a tensor processing unit (TPU), or an AI chip (AI semiconductor), and executes various information processing and control processes to be performed by the server 1 by reading and executing a program P1 stored in the auxiliary memory unit 14. Note that if the control unit 11 includes multiple processors, each process may be executed by a different processor. The main memory unit 12 is a temporary storage area, such as a static random access memory (SRAM), a dynamic random access memory (DRAM), or a flash memory, and temporarily stores data generated when the control unit 11 executes arithmetic processing. The communication unit 13 is a communication module for performing communication-related processes and transmits and receives information to and from other devices via the network N.

[0014] The auxiliary storage unit 14 is a non-volatile storage area such as a large-capacity memory, a hard disk, or an SSD (Solid State Drive), and stores a program P1 (program product) and other data required for the control unit 11 to execute processing. The server 1 also has web server functionality, and the auxiliary storage unit 14 stores a survey site S for conducting personality tests and surveys and providing feedback on the personality test and survey results to employees or their superiors. The auxiliary storage unit 14 also stores a language model M. The language model M is a general-purpose large-scale language model constructed by pre-training using a large set of sentences, etc. The language model M is trained to, when input data including, for example, text data written in a natural language, perform calculations based on the input data to generate output data corresponding to the content of the input data and output the generated output data. The language model M can be configured using algorithms such as GPT (Generative Pre-trained Transformer)-3, GPT-3.5, and GPT-4, or may be configured by combining multiple algorithms. Note that the language model M is not limited to the Transformer-based model described above. The language model M is expected to be used as a program module constituting artificial intelligence software. The language model M performs a predetermined calculation on input data and outputs the calculation result. The auxiliary storage unit 14 stores data such as coefficients and thresholds of functions that define this calculation as the language model M. Instead of storing the language model M in the auxiliary storage unit 14, the server 1 may access a language processing server that stores the language model M and read it out. The auxiliary storage unit 14 also stores a personality test DB 141, a survey DB 142, an employee DB 143, and a prompt creation table 144, which will be described later. The auxiliary storage unit 14 may be an external storage device connected to the server 1, and any or all of the DBs 141 to 143 and the prompt creation table 144 may be stored in another server.

[0015] In this embodiment, the server 1 may be a multi-computer consisting of multiple computers, or may be a virtual machine virtually constructed by software within a single device. The server 1 may also be a local server installed in a facility where the server 1 (information processing system) is installed, or a cloud server connected to the facility via a network N. The server 1 is not limited to the above configuration and may include, for example, an input unit for accepting operation inputs and a display unit for displaying images. The server 1 may also include a reading unit for reading a non-transitory computer-readable recording medium 1a and read the program P1, etc., from the recording medium 1a. The program P1 may be located on a single computer or at a single site, or may be distributed across multiple sites and executed in a distributed manner on multiple computers interconnected via the network N.

[0016] The boss terminal 2 includes a control unit 21, a main memory unit 22, a communication unit 23, a display unit 24, an input unit 25, and an auxiliary memory unit 26. The control unit 21 has one or more processors such as CPUs, and executes various information processing to be performed by the boss terminal 2 by reading and executing a program P2 stored in the auxiliary memory unit 26. The main memory unit 22 is a temporary storage area such as RAM, and temporarily stores data generated when the control unit 21 executes arithmetic processing. The communication unit 23 is a communication module for performing processing related to communication, and transmits and receives information to and from other devices via the network N.

[0017] The display unit 24 is a liquid crystal display, an organic EL display, or the like, and displays various information in accordance with instructions from the control unit 21. The input unit 25 is an operation interface such as a keyboard or a mouse, and receives operation input from a user (supervisor), and sends a control signal according to the operation content to the control unit 21. The display unit 24 and the input unit 25 may be a touch panel configured as an integrated unit.

[0018] The auxiliary storage unit 26 is a non-volatile storage area such as a large-capacity memory, a hard disk, or an SSD, and stores the program P2 and other data required for the control unit 21 to execute processing. The auxiliary storage unit 26 also stores a web browser AP for accessing a web server. The boss terminal 2 may include a reading unit that reads the non-transitory computer-readable recording medium 2a, and may read the program P2 and the like from the recording medium 2a and store them in the auxiliary storage unit 26.

[0019] FIG. 3 is an explanatory diagram showing an example of the record layout of the personality test DB 141 and the survey DB 142. The personality test DB 141 is a database that stores information on questions (questionnaires) used in personality tests in this system. The personality test DB 141 includes a question ID column, an item column, and a question content column. The question ID column stores identification information (question IDs) for identifying each question used in the personality test. The item column stores the item names of items related to personality traits identified by the personality test. In this embodiment, the items related to personality traits have major items and minor items. The major items include interpersonal relationship characteristics, leadership characteristics, and situational awareness, and each major item includes multiple minor items. For example, minor items related to interpersonal relationship characteristics include attentive listening tendency and assertiveness, minor items related to leadership characteristics include problem-solving orientation, and minor items related to situational awareness include optimism. The question content column stores the content of questions to be asked to employees in association with each minor item. Each question is designed to analyze the respondent's personality and extract characteristic personality traits, and one or more questions are prepared for each item. The personality test DB 141 is not limited to the configuration shown in Fig. 3, and the number and types of items (major items and minor items), and the number and contents of questions corresponding to each item can be set arbitrarily.

[0020] The survey DB 142 is a database that stores information about questions (questionnaires) used in surveys in this system. The survey DB 142 includes a question ID column, a category column, and a question content column. The question ID column stores identification information (question ID) for identifying each question used in the survey. The category column stores the category into which the content of each question is classified. In the survey of this embodiment, questions are asked in six categories: motivation for work, attachment to the company, physical and mental state, human relationships, work content, and workload, so the category column stores any of these. The question content column stores the content of the questions to be asked to employees. Each question includes a question that can derive an employee's mental health state (stress state, engagement, mental health) related to the workplace (work environment) and work content, and a question that can derive the items (issues) that cause the derived state, and one or more questions are prepared for each category. The survey DB 142 is not limited to the configuration shown in FIG. 3, and the number and types of categories, and the number and contents of questions corresponding to each category can be set arbitrarily.

[0021] The server 1 of this embodiment generates a personality test screen based on the contents stored in the personality test DB 141 and transmits it to the employee terminal. The server 1 then acquires (collects) from the employee terminal the answers to each question entered via the personality test screen. The personality test of this embodiment is configured to allow the employee to select one of five options for each question: "1: Not at all applicable," "2: Not very applicable," "3: Neither applicable nor unrelated," "4: Somewhat applicable," and "5: Very applicable." The server 1 acquires the question ID and answer content (any of 1 to 5) for each question from the employee terminal. When the server 1 acquires the answers to the personality test from the employee terminal, it calculates a score for each item related to personality traits based on the answers to each question. The score for each item in the personality test can be a score assigned based on the answer to each question, or a standard deviation that can be calculated based on the test results (scores) of test-takers who have previously taken the personality test. The stronger the personality trait characteristics of each item, the higher the score for that item.

[0022] The server 1 also generates a survey screen based on the contents stored in the survey DB 142, transmits it to the employee terminal, and acquires (collects) answers to each question entered via the survey screen from the employee terminal. For example, the server 1 asks questions to derive the employee's mental health state regarding the workplace (work environment) and work content, and then presents questions to derive the items (issues) that are causing the derived state. The survey of this embodiment is configured to allow responses to questions from six options, for example, "1: Not applicable at all," "2: Not applicable," "3: Somewhat not applicable," "4: Somewhat applicable," "5: Applicable," and "6: Very applicable." The server 1 acquires the question ID and answer content (any of 1 to 6) of each question from the employee terminal. When the server 1 acquires the survey answers from the employee terminal, it calculates a score for each category in the survey based on the answers to each question. For example, if a score of 1 to 6 is assigned to each answer in the order of "not applicable at all," "not applicable," "somewhat not applicable," "slightly applicable," "applicable," and "very applicable," the server 1 calculates the sum of the scores assigned to the answers to each question for each category. Note that five questions are asked for each category, and each answer is assigned a score of 1 to 6, so a total score between 5 and 30 is calculated for each category. The server 1 then converts the calculated total score for each category into a score out of 10 using the following formula: The better the employee's condition in each category, the higher the score for that category.

[0023] (Total score - lowest total score) / (Highest total score - lowest total score) x 10 …(formula) If five questions are asked in one category and each answer is assigned a score from 1 to 6, the minimum total score is 5 and the maximum total score is 30.

[0024] The scores assigned to each question are not limited to 1 to 6, and weighting may be applied to the scores for each question when calculating the total score for each category. The formula for converting the total scores for each category into a score out of 10 is not limited to the above formula. After calculating a score out of 10 for each category in the survey, the server 1 calculates a well-being score out of 10 based on the scores for each category. The well-being score is an overall score for the respondent that is comprehensively assessed from the scores for each category. For example, the server 1 may calculate the average of the scores for the six categories as the well-being score, or may weight each category and calculate the weighted average as the well-being score. For example, the server 1 may weight mental and physical state, vitality for work, or attachment to the company more heavily than other categories.

[0025] Furthermore, the server 1 generates an analysis result of the respondent based on the score of each category as the survey result. For example, a database (not shown) is provided that stores comments on the respondent's status, etc., analyzed from the score of each category (survey result). By referencing this database, the server 1 can obtain an analysis result (analysis comment) corresponding to the respondent's survey result. The server 1 may obtain an analysis result that takes into account not only the score of each category but also the increase or decrease from the previous (past) score of each category (difference from the previous score, change over time). In this case, a database can be used that stores comments on the respondent's status, etc., analyzed from the score of each category and the change from the previous score of each category (survey result). The server 1 also generates advice to be presented to the respondent and the respondent's supervisor based on the score of each category as the survey result. For example, a database (not shown) is provided that stores advice to be presented to the respondent and the supervisor based on the score of each category (survey result) and the respondent's personality test result. By referencing this database, the server 1 can obtain advice to be presented to the respondent and the supervisor based on the respondent's survey result and personality test result. Here too, the server 1 may acquire advice that takes into account not only the score for each category but also the increase or decrease in each category from the previous (past) score. Advice for the respondent and his / her supervisor is generated by identifying a predetermined number of categories that require attention based on the score for each category and the increase or decrease from the previous score (change in score over time), and generating advice for each identified category. As described above, the personality test results and survey results acquired by the server 1 are stored in the employee DB 143 in association with the employee ID assigned to each employee.

[0026] FIG. 4 is an explanatory diagram showing an example of a record layout of the employee DB 143. The employee DB 143 is a database that stores information about employees belonging to companies or the like that use this system. The employee DB 143 includes an employee ID column, a name column, a department column, a personality test result column, and a survey result column. The employee ID column stores identification information (employee ID) for identifying each employee. The name column and department column store the employee's name and department in association with the employee ID. The personality test result column stores the score for each item representing the employee's personality traits as the test result of the employee's personality test in association with the employee ID. The personality test result column may store the employee's answer to each question or information indicating a high or low tendency for each item. The personality test is conducted, for example, after the employee joins the company, and the determined test results are stored in the employee DB 143. The personality test may also be conducted appropriately, such as once every few months to several years, when the employee is transferred to another department, and the test results stored in the employee DB 143 may be updated. The survey result column stores information about the results of a survey conducted by an employee, in association with the employee ID. The information about the survey results includes the survey implementation date and time, scores for each category (vitality for work, attachment to the company, mental and physical state, job content, interpersonal relationships, and workload), a well-being score, analysis results (analysis comments), advice for the employee, and advice for the supervisor. Surveys are conducted approximately once a week to once every few months, and the survey results along with the implementation date and time are stored in the employee DB 143. The employee DB 143 is not limited to the configuration shown in FIG. 4; each piece of information may be divided into multiple DBs and registered, or some information may be registered in a DB on another server. For example, the employee's personality test results or survey results may be stored in another server that manages employee information, in association with the employee ID, etc.

[0027] As described above, the server 1 stores the personality test results and survey results of each employee in the employee DB 143. In this embodiment, the server 1 performs a process of providing feedback on the survey results of each employee (hereinafter referred to as subordinates) to each employee's supervisor. Specifically, the server 1 generates a feedback screen displaying the survey results of the subordinate employees and transmits it to the supervisor terminal 2. Upon receiving the feedback screen from the server 1, the supervisor terminal 2 displays it on the display unit 24, allowing the supervisor to check the survey results of each subordinate. On the feedback screen, the supervisor can ask questions about how to treat their subordinates, and the server 1 can provide answers (advice) to the questions. In this case, the server 1 performs a process of generating answers (advice) to the received questions using a language model M. The language model M may be a general-purpose large-scale language model as is, or may be fine-tuned using information on the respondent's (subordinate's) personality traits obtained from the personality test results and information on the respondent's psychological state and challenges obtained from the survey results. The fine tuning of the language model M may be performed by the server 1 or by another learning device. The language model M fine-tuned by the other learning device is downloaded from the learning device to the server 1 via the network N or the recording medium 1a, for example, and stored in the auxiliary storage unit 14.

[0028] The following describes prompts that the server 1 inputs into the language model M when generating an answer (advice) to a question using the language model M. FIGS. 5 to 7 are explanatory diagrams showing an example of the prompt creation table 144. The prompt creation table 144 includes a prompt template shown in FIG. 5, a pitch wording numerical value reference table, a previous difference wording numerical value reference table, and a personality description sentence table shown in FIG. 6, and an imperative sentence table shown in FIG. 7. When generating an answer (advice) to a question about a subordinate or a question about how to deal with (treat) a subordinate, the server 1 of this embodiment creates a prompt that includes not only the content of the question but also information based on a survey result of the subordinate and information based on a personality test result, and inputs this into the language model M. The server 1 generates the prompt using the prompt creation table 144.

[0029] The prompt template shown in Figure 5 includes sentences related to the premise (background) for generating advice (answers to questions), sentences related to the subordinate's psychological state based on survey results, sentences related to the subordinate's challenges based on survey results, sentences related to the subordinate's personality based on personality test results, and sentences related to commands. The sentences related to the premise are pre-prepared default sentences. The sentences related to the psychological state are sentences in which the survey results of the subject (subordinate) are inserted into pre-prepared sentence templates for each category in the survey (energy for work, attachment to the company, physical and mental state, work content, interpersonal relationships, and workload). In the example shown in Figure 5, w1_hl, w2_hl, etc. in the sentences related to the psychological state are inserted with sentences corresponding to the scores of each category in the survey. The sentences corresponding to the scores of each category are identified from the high and low wording numerical standard table shown in Figure 6. The high and low wording numerical standard table stores sentences corresponding to each score range. The sentences regarding psychological state may use phrases corresponding to the scores of each category, or may use sentences in which the scores of each category are inserted directly. In the sentences regarding psychological state, w1_change, w2_change, etc., phrases corresponding to the difference in the score of each category from the previous survey (current score - previous score) are inserted. The phrases corresponding to the difference in score from the previous survey (change in score over time) are identified from the previous difference phrase numerical value reference table shown in Figure 6. The previous difference phrase numerical value reference table stores phrases corresponding to each range of difference from the previous survey. The sentences regarding the issues may use, for example, one or more of the analysis results included in the survey results, advice for the individual, and advice for the supervisor. In the example of Figure 5, the content of the issues (specifically, psychological state) and solutions (for example, advice to the supervisor, or other measures the supervisor should take) are included for three categories identified from the survey results (for example, the three categories with the lowest scores). For example, a predetermined number of characteristic items (five in the example of FIG. 5) are identified based on the scores of each item of personality traits included in the personality test results, and an explanation corresponding to the score of each identified item is used for the personality-related sentences. The personality-related sentences are identified from the personality explanation table shown in FIG.The personality description table stores, in association with each item of the personality trait, a description for when the score for each item is less than 50 and a description for when the score for each item is 50 or greater, when the score for each item is a standard deviation. For example, when the score for each item is a standard deviation, the characteristic items can be items with a large difference between the score and 50. In this case, a predetermined number of items are identified in descending order of the difference between the score and 50. The command sentence is identified from the command sentence table shown in FIG. 7. In this embodiment, multiple options (questions) are prepared in advance for questions accepted via the feedback screen, and the command sentence is registered in the command sentence table in association with the question item and the option for each item. As described above, a prompt to be input to the language model M is created by creating and inserting into the prompt template sentences related to the subject's psychological state, task, and personality, as well as command sentences corresponding to the question to be answered.

[0030] The prompt template is not limited to the example shown in Figure 5. For example, the prompt may include at least information based on the survey results, but may not include information based on the personality test results. The prompt may also include a message for controlling the output result from the language model M. For example, a message such as "When presenting solutions to the following problems in the future, please base your proposal on the solutions described below" may be included in the problem-related sentence in Figure 5. This allows the output result from the language model M to be controlled and the desired content to be output from the language model M to be guided. Furthermore, a message regarding constraints on the output result from the language model M may be included in the statement in the statement table in Figure 7, such as "Please write the output in a Markdown table format with two columns, 'Concerns and Risks' and 'Explanation,' and approximately 400 characters long."

[0031] The following describes a process in the information processing system of this embodiment in which the server 1 generates and presents answers (advice) to questions received via a feedback screen. FIG. 8 is a flowchart showing an example of the advice presentation process, and FIGS. 9 to 11 are explanatory diagrams showing example screens. In the following process, the server 1 transmits a feedback screen to the boss terminal 2, which then receives an operation from the boss via the feedback screen displayed on the boss terminal 2 and transmits data corresponding to the received operation to the server 1. The server 1 performs processing based on the data received from the boss terminal 2 and transmits the processing results to the boss terminal 2, which then updates the feedback screen based on the received processing results. This allows the boss to view the feedback screen using the boss terminal 2 and receive advice on various questions via the feedback screen. In this way, in this embodiment, data is transmitted and received (exchanged) between the server 1 and the boss terminal 2. However, for simplicity, FIG. 8 does not illustrate the transmission and reception of data between the server 1 and the boss terminal 2.

[0032] In the information processing system of this embodiment, when a supervisor (an employee with subordinates, a human resources department staff member, etc.) wants to understand the status of a subordinate, the supervisor accesses the server 1 (survey site S) using a predetermined URL (Uniform Resource Locator) and requests a feedback page for the supervisor. When the supervisor terminal 2 requests a feedback page for the supervisor, the control unit 11 of the server 1 generates a feedback screen displaying the requested survey results of the subordinate and displays it on the supervisor terminal 2 (S11). Specifically, the control unit 11 transmits the feedback screen to the supervisor terminal 2, and the control unit 21 of the supervisor terminal 2 displays the received feedback screen on the display unit 24. FIG. 9 shows an example of the feedback screen. The screen in FIG. 9 displays the survey results for one subordinate, including the scores for each survey category and well-being score, advice for the supervisor, and analysis results (analysis comments). This information displayed on the feedback screen can be read from the employee DB 143. Note that since the feedback screen here is viewed by the supervisor, advice for the supervisor from the survey results is displayed, but advice for the employee from the survey results may also be displayed. Furthermore, analytical comments on the survey results may be modified for the supervisor and displayed on the feedback screen.

[0033] The server 1 may, for example, compile the survey results of each employee by department, generate a department feedback screen displaying the survey results of employees (members) in each department, display it on the supervisor's terminal 2, and accept a request for a feedback screen for any subordinate by selecting any one subordinate via the department feedback screen. The server 1 may also compile the survey results of all employees belonging to the company, generate a company feedback screen displaying the survey results for the entire company, and display it on the supervisor's terminal 2. In this case, the server 1 may accept the selection of any department via the company feedback screen, display the department feedback screen for the selected department on the supervisor's terminal 2, and accept a request for a feedback screen for any subordinate via the department feedback screen.

[0034] The screen in FIG. 9 is provided with a "Result Interpretation and Advice" button (operation button, hereinafter referred to as the "Mikiwame AI" button A1) for asking questions about the displayed survey results and responses (actions, support, management) to the subordinate. When the supervisor wants to receive an explanation about the survey results or advice on how to interact with the subordinate, the supervisor operates the "Mikiwame AI" button A1 via the input unit 25 of the supervisor terminal 2. The control unit 11 of the server 1 determines whether the "Mikiwame AI" button A1 has been operated (S12), and if it determines that it has not been operated (S12: NO), it waits while performing other processing until it is operated. When the "Mikiwame AI" button A1 has been operated, the control unit 21 of the supervisor terminal 2 notifies the server 1 that the button has been operated, so that the control unit 11 of the server 1 can know that the "Mikiwame AI" button A1 has been operated.

[0035] When the control unit 11 determines that the "Mikiwame AI" button A1 has been operated (S12: YES), it displays a question menu field A2 on the feedback screen being displayed on the supervisor terminal 2 (S13), as shown in FIG. 10. The question menu field A2 displays a plurality of questions (question menu) related to each of a plurality of items prepared in advance, and the display area of ​​each question is a button that allows the supervisor to select one of the questions (options). In the question menu field A2, the supervisor selects the button of a question (option) that is closest to the content of the question the supervisor wishes to ask, or selects the "I want to consult via free text" button. The control unit 11 accepts the input of a question (selection of an option) via the question menu being displayed on the supervisor terminal 2 (S14). In the example of FIG. 10, the option (question) of "Actions that should be considered by the supervisor or human resources" in the item "Responses to improve the condition" is selected.

[0036] When the control unit 11 receives a question, it reads out the survey results and personality test results of the subordinate (employee) displayed on the feedback screen from the employee DB 143 (S15). Based on the read survey results and personality test results, the control unit 11 creates a sentence to be used as a prompt to instruct the language model M to generate an answer (advice) to the question. Specifically, the control unit 11 creates a sentence regarding the psychological state of the subordinate from the survey results (S16). As described above, the control unit 11 inserts words and sentences corresponding to the survey results of the subordinate into a sentence template prepared for each survey category, referring to the pitch and low wording numerical value standard table and the previous difference wording numerical value standard table, to create a sentence regarding the psychological state. The control unit 11 also creates a sentence regarding the subordinate's challenges from the survey results (S17). For example, the control unit 11 extracts advice for the supervisor from the survey results and uses it as a sentence regarding the challenges. The control unit 11 also creates a sentence regarding the subordinate's personality from the personality test results (S18). For example, the control unit 11 identifies a predetermined number of characteristic items based on the scores of each item (personality trait) included in the personality test results, and for each identified item, obtains an explanation corresponding to the score from the personality explanation table and uses it as a sentence related to personality.

[0037] The control unit 11 uses the sentences created in steps S16 to S18 to create a prompt corresponding to the question received in step S14 (S19). Here, the control unit 11 acquires an imperative sentence corresponding to the selected question from the imperative sentence table, and creates a prompt by inserting the sentence related to the psychological state created in step S16, the sentence related to the task created in step S17, the sentence related to personality created in step S18, and the imperative sentence acquired from the imperative sentence table into the prompt template shown in Fig. 5. Note that the prompt is not limited to a configuration including the above-mentioned sentences, and may, for example, be a configuration that does not include a sentence related to personality.

[0038] The control unit 11 inputs the created prompt into the language model M and acquires output information according to the content of the prompt, which is generated and output by the language model M (S20). Then, based on the acquired output information from the language model M, the control unit 11 displays an answer (advice) to the question received in step S14 in the question menu field A2 currently displayed on the supervisor terminal 2, as shown in FIG. 11A (S21). For example, the control unit 11 acquires advice written in natural language from the language model M and displays the acquired advice in the question menu field A2. In the example of FIG. 10, the question "Actions that should be considered by the supervisor or human resources" is selected. As a result, the question menu field A2 in FIG. 11A displays the actions that should be considered and explanations including points to note when taking the action. The supervisor can improve the subordinate's motivation and engagement by taking actions such as interviewing and providing appropriate support based on the advice displayed in the question menu field A2. The control unit 11 may be configured to display the advice generated using the language model M in the question menu field A2, or may be configured to display it in a display field separate from the question menu field A2.

[0039] The question menu field A2 is provided with a message input field A3, and if the superior has any questions about the advice he or she viewed, he or she inputs a further question (message) in the input field A3. The control unit 21 of the superior terminal 2 accepts a message input into the input field A3 in the question menu field A2 via the input unit 25 and transmits the accepted message to the server 1, whereby the control unit 11 of the server 1 accepts the message. The control unit 11 determines whether or not the message has been accepted (S22). If it determines that the message has been accepted (S22: YES), the control unit 11 returns to step S19 and creates a prompt corresponding to the accepted message (further question) (S19). For example, the control unit 11 adds the output information acquired from the language model M in step S20 or the answer information displayed in the question menu field A2 in step S21, and the accepted message to the prompt created in the most recent step S19, to create a new prompt. Note that the control unit 11 is not limited to a configuration in which a new prompt is created from the prompt created in the most recent step S19, but may create a new prompt that includes the sentence created in steps S16 to S18, the previous question and its answer, and a new question.

[0040] Then, the control unit 11 performs the processes of steps S20 to S21 using the newly created prompt. When the question "Is there anything in particular I should be careful about?" is entered in the input field A3 on the screen of FIG. 11A, the answer to the entered question is displayed as shown in the question menu field A2 of FIG. 11B. By entering a question in the input field A3, the supervisor can receive advice on questions regarding the survey results of the subordinates and questions regarding how to deal with the subordinates. Therefore, by taking appropriate action in accordance with the advice presented via the question menu field A2, the supervisor can improve the subordinates' motivation and engagement and support the implementation of optimal actions in managing the subordinates. If the control unit 11 determines that a message has not been received via the input field A3 in the question menu field A2 (S22: NO), the control unit 11 terminates the process.

[0041] According to the above-described process, in this embodiment, the survey results of employees (subordinates) are fed back to their superiors, allowing the superiors to ask questions about points of uncertainty and receive detailed advice on how to deal with their subordinates. Therefore, specific advice is given on the actions that superiors should take, taking into account the psychological state and personality of the employees, and explanations and further advice are provided regarding the advice, allowing the superiors to accurately understand the content of the advice and take appropriate action in accordance with the advice. Furthermore, in this embodiment, multiple questions are prepared in advance, and questions and advice can be received in chatbot format, allowing superiors to easily find out what they want to know by using advice verbalized by the language model M.

[0042] In this embodiment, advice is presented that takes into account not only the survey results but also the personality test results, so that optimal actions are proposed for each employee, enabling optimal support for each employee. Furthermore, in this embodiment, employees regularly take surveys, so that the employee's condition can be understood by taking into account not only the results at the time of taking the survey but also changes in the survey results over time (chronological changes), allowing the employee to receive appropriate advice. By understanding the stress and mental health status of their subordinates and their changes, and taking action based on the presented advice, supervisors can help their subordinates reduce stress from work and the workplace, create a comfortable working environment, and contribute to improving each employee's motivation and engagement. As a result, it is possible to reduce absences and resignations due to mental health problems and to expect improved performance throughout the organization.

[0043] (Embodiment 2) In the first embodiment described above, the system is configured to provide advice and commentary when a supervisor responds (takes action, provides support) to an employee (subordinate). In this embodiment, an information processing system is described that provides advice and commentary on responses that should be taken by each organization, such as a department, established in a company or the like. Specifically, the system of this embodiment provides advice and commentary on responses that should be taken by each organization's supervisor (manager, department head, human resources department staff, management, etc.) to employees of each organization, based on the results of a survey of employees belonging to each organization. The organization in this embodiment can be any unit, such as a department, division, section, section, team, etc., such as a human resources department, planning department, or development department, or it can be the entire company or corporation.

[0044] The information processing system of this embodiment can be realized using devices similar to those of the information processing system of embodiment 1 shown in Figures 1 and 2, so a description of the configuration of each device will be omitted. The auxiliary storage unit 14 of the server 1 of this embodiment stores a prompt creation table for an organization in addition to or instead of the prompt creation table 144 shown in Figures 5 to 7. The prompt creation table for an organization includes a prompt template for an organization, a pitch wording numerical value reference table and a previous difference wording numerical value reference table shown in Figure 6, and an organization command statement table. The pitch wording numerical value reference table and previous difference wording numerical value reference table are the same as the pitch wording numerical value reference table and previous difference wording numerical value reference table of embodiment 1 shown in Figure 6, but tables in which different ranges and / or wordings are registered may also be used.

[0045] 12 and 13 are explanatory diagrams showing an example of the configuration of a prompt creation table for an organization. FIG. 12 shows an example of a prompt template for an organization, and FIG. 13 shows an example of an instruction statement table for an organization. When generating answers (advice) to questions about the status of employees at the organizational level and questions about responses at the organizational level, the server 1 of this embodiment creates prompts that include not only the content of the questions but also information based on survey results of employees belonging to the organization, and inputs these into the language model M. The organizational prompt template shown in FIG. 12 includes sentences related to the premise (background) for generating advice (answers to questions), sentences related to survey results at the organizational level based on the survey results of employees within the organization, sentences related to consideration steps for interpreting the organizational results, and sentences related to commands. The sentences related to the premise are default sentences prepared in advance. The sentences related to the organizational survey results are sentences prepared in advance for each category of the survey (vitality for work, attachment to the company, physical and mental state, work content, human relationships, and workload) with the survey results for the organizational level inserted into them. The organizational survey results are scores calculated from the survey results of each employee belonging to the organization. For example, the average (mean score) of the survey results (scores for each category) for each employee can be used. Note that the survey results for each organization can also be calculated as a weighted average weighted based on the employee's job title, team, years of service, etc. The median of the survey results for each employee can also be used as the organizational survey results. In the example of Figure 12, for hl11 to hl13, hl21, and hl31 in the text regarding the organizational survey results, wording corresponding to the score for each category in the survey at the organizational level is identified from the high / low wording numerical standard table and inserted. Note that in addition to text using wording corresponding to the score for each category, text in which the score for each category is inserted directly can also be used. In the text about the organizational survey results, changes 11 to 13, change 21, and change 31 are inserted with wording corresponding to the difference in the organizational category score from the previous survey (current score - previous score), which is identified from the previous difference wording numerical standard table shown in Figure 6.Although the sentences related to the step of interpreting the organizational results are default sentences prepared in advance, different sentences may be prepared depending on the size of the organization (company unit, department unit, division unit, etc.), the content of the organization (human resources department, planning department, development department, etc.), and the content of the question. The sentences related to commands are identified from the command statement table shown in FIG. 13 and used based on the question entered via the feedback screen. The command statement table in FIG. 13 has a configuration similar to the command statement table of the first embodiment shown in FIG. 7, and command statements are registered in association with question items and options for each item. As described above, a prompt to be input to the language model M is created by creating and inserting sentences related to the organizational survey results and command statements related to the question into the prompt template for the organization.

[0046] The prompt template for an organization is not limited to the example shown in Fig. 12. For example, the prompt may include information based on the results of a personality test of employees belonging to the organization. The prompt may also include a message for controlling the output result from the language model M.

[0047] The following describes a process in the information processing system of this embodiment in which the server 1 generates and presents an answer (advice) to a question received via a feedback screen. Fig. 14 is a flowchart showing an example of the advice presentation process procedure of the second embodiment, and Figs. 15 to 17 are explanatory diagrams showing example screens. The process shown in Fig. 14 is the process shown in Fig. 8 in which steps S31 to S43 are added before step S11. Explanations of the same steps as in Fig. 8 will be omitted. Also, Fig. 14 does not show steps S12 to S22 in Fig. 8. For simplification, Fig. 14 also omits illustration of data transmission and reception between the server 1 and the boss's terminal 2.

[0048] In the information processing system of this embodiment, when a superior (such as a department head, manager, personnel department staff member, or executive team member) wants to understand the status of the organization he or she manages, he or she accesses the server 1 (survey site S) using a predetermined URL and requests a feedback page for the superior. At that time, the superior selects an organization he or she manages and requests a feedback page for the selected organization. The control unit 11 of the server 1 determines whether or not the selection of the organization has been accepted from the superior's terminal 2 (S31). If it is determined that the selection has not been accepted (S31: NO), the process proceeds to step S11. In this case, the control unit 11 executes the same process as in the first embodiment shown in FIG. 8. Specifically, the control unit 11 of the server 1 generates a feedback screen displaying the results of the survey of subordinates requested by the superior's terminal 2, displays it on the superior's terminal 2 (S11), and executes the processes from step S12 onward.

[0049] When the control unit 11 determines that the selection of an organization has been accepted (S31: YES), it reads out the survey results of each employee belonging to the selected organization from the employee DB 143 (S32). The control unit 11 calculates the average score for each category based on the scores for each category, which are the survey results for each employee (S33). The control unit 11 may calculate a weighted average score using weightings set for each employee, or may determine the median score of each employee for each category. The control unit 11 generates a feedback screen displaying the survey results of the selected organization and displays it on the supervisor terminal 2 (S34). Here, the control unit 11 transmits the feedback screen to the supervisor terminal 2, and the control unit 21 of the supervisor terminal 2 displays the received feedback screen on the display unit 24. FIG. 15 shows an example of the organizational feedback screen. The screen of FIG. 15 displays the average scores for each survey category and the average value of the well-being scores as the survey results for the organization. The screen of FIG. 15 also has an employee area displaying the survey results (well-being scores) of employees belonging to the selected organization. The employee area is provided with a top priority care tab, a priority care tab, and an all member tab. By switching between tabs, information on employees identified as top priority care recipients, employees identified as priority care recipients, or all employees belonging to the organization is displayed, depending on the selected tab. The employee area displays a list of employee names and well-being scores according to the selected tab. Note that top priority care recipients are a predetermined number (10 in FIG. 15) of employees with the lowest well-being scores among the organization's employees, and priority care recipients are a predetermined number (10 in FIG. 15) of employees with the second lowest well-being scores after the top priority care recipients.

[0050] The screen in FIG. 15 has a "Perform organizational analysis with AI" button A4 for asking questions about the displayed organizational survey results and responses (actions, support, management) to the organization. When the supervisor wants to receive an explanation about the organizational survey results or advice on responses that should be made to the organization, the supervisor operates the "Perform organizational analysis with AI" button A4 via the input unit 25 of the supervisor's terminal 2. The control unit 11 of the server 1 determines whether the "Perform organizational analysis with AI" button A4 has been operated (S35), and if it determines that it has not been operated (S35: NO), the control unit 11 waits while performing other processes until it is operated. When the "Perform organizational analysis with AI" button A4 has been operated, the control unit 21 of the supervisor's terminal 2 notifies the server 1 that the button A4 has been operated, so that the control unit 11 of the server 1 can know that the "Perform organizational analysis with AI" button A4 has been operated.

[0051] When the control unit 11 determines that the "Analyze organization with AI" button A4 has been operated (S35: YES), it displays a question menu field A5 on the feedback screen displayed on the supervisor terminal 2 (S36), as shown in FIG. 16. The question menu field A5 displays a plurality of questions (question menu) related to each of a plurality of items prepared in advance, and the display area of ​​each question is a button that allows the supervisor to select one of the questions. In the question menu field A5, the supervisor selects the button for a question that is close to the content of the question the supervisor wants to ask, or selects the "I want to consult using free text" button. The control unit 11 accepts the input of a question via the question menu displayed on the supervisor terminal 2 (S37). In the example of FIG. 16, the question "Actions that managers should consider" under the item "Responses for organizational improvement" has been selected.

[0052] When the control unit 11 receives a question input, it creates a sentence to be used as a prompt to instruct the language model M to generate an answer (advice) to the question, based on the survey results of the organization indicated by the average scores of each category calculated in step S33. Specifically, the control unit 11 creates a sentence related to the survey results of the organization from the survey results (S38). Here, the control unit 11 creates a sentence according to the survey results of the organization by referring to the pitch wording numerical value standard table and the previous difference wording numerical value standard table in a sentence template prepared for each category of the survey.

[0053] The control unit 11 uses the sentence created in step S38 to create a prompt corresponding to the question accepted in step S37 (S39). Here, the control unit 11 acquires a statement corresponding to the selected question from the statement table, and creates a prompt by inserting the sentence related to the survey results of the organization created in step S38 and the statement acquired from the statement table into the prompt template shown in Fig. 12.

[0054] The control unit 11 inputs the created prompt into the language model M and acquires output information according to the content of the prompt generated and output by the language model M (S40). Based on the acquired output information from the language model M, the control unit 11 then displays an answer (advice) to the question received in step S37 in the question menu field A5 currently displayed on the supervisor terminal 2, as shown in FIG. 17A (S41). In the example of FIG. 16, the question "Actions the Manager Should Consider" is selected. As a result, the question menu field A5 in FIG. 17A displays the actions the manager should consider and explanations including important points to note. The supervisor can improve the motivation and engagement of each employee by taking actions such as interviewing and providing appropriate support to each employee based on the advice displayed in the question menu field A5. The control unit 11 may display the advice generated using the language model M in the question menu field A5 or in a display field separate from the question menu field A5.

[0055] The question menu field A5 is provided with a message input field A6, and if the superior has any questions about the advice he or she viewed, he or she inputs a further question (message) in the input field A6. The control unit 21 of the superior terminal 2 accepts a message input into the input field A6 in the question menu field A5 via the input unit 25 and transmits the accepted message to the server 1, whereby the control unit 11 of the server 1 accepts the message. The control unit 11 determines whether or not the message has been accepted (S42). If it determines that the message has been accepted (S42: YES), the control unit 11 returns to step S39 and creates a prompt corresponding to the accepted message (further question) (S39). For example, the control unit 11 adds the output information acquired from the language model M in step S40 or the answer information displayed in the question menu field A5 in step S41, and the accepted message to the prompt created in the most recent step S39, to create a new prompt. In addition, the control unit 11 is not limited to a configuration in which a new prompt is created from the prompt created in the most recent step S39, but may also create a new prompt that includes the sentence created in step S38, the previous question and its answer, and a new question.

[0056] Then, the control unit 11 performs the processes of steps S40 to S41 using the newly created prompt. When the question "Is there anything in particular I should be careful about?" is entered in the input field A6 on the screen of FIG. 17A, the answer to the entered question is displayed as shown in the question menu field A5 of FIG. 17B. By entering a question in the input field A6, the supervisor can obtain advice on questions regarding the results of the organizational survey and questions regarding how to respond to the organization. Therefore, by taking appropriate action in accordance with the advice presented via the question menu field A5, the supervisor can improve the motivation and engagement of each employee in the organization and support the achievement of organizational goals.

[0057] 15 to 17, the employee area is configured to allow the selection of any employee, and when any employee is selected, a feedback screen such as that shown in FIG. 9 is displayed for the selected employee. Therefore, when the control unit 11 determines that a message has not been received via the input field A6 in the question menu field A5 (S42: NO), it determines whether or not the selection of any employee has been received from the employee area (S43). Here, when the control unit 21 of the supervisor terminal 2 receives the selection of any employee via the employee area of ​​the displayed feedback screen, it notifies the server 1 of the information of the selected employee, thereby allowing the control unit 11 of the server 1 to identify the selected employee.

[0058] If the control unit 11 determines that the selection of any employee has not been accepted (S43: NO), it returns to step S42, and if it determines that the selection of any employee has been accepted (S43: YES), it proceeds to step S11. That is, if the selection of any employee has been accepted via the organization feedback screen, a feedback screen such as that shown in Fig. 9 regarding the selected employee is provided to the supervisor terminal 2.

[0059] According to the above-described process, in this embodiment, the employee survey results are fed back to supervisors on an organizational level, allowing the supervisors to ask questions about unclear points and receive detailed advice on how to respond to the organization. Therefore, specific advice is given to supervisors and other personnel regarding actions to be taken based on the survey results of employees within the organization, and explanations and further advice are provided regarding the advice. This allows the supervisors and other personnel to accurately understand the advice and take appropriate action in accordance with the advice. Furthermore, in this embodiment, multiple questions are prepared in advance, and questions and advice can be received in chatbot format. Therefore, supervisors and other personnel can easily learn what they want to know through advice verbalized by the language model M. Therefore, supervisors and other personnel can, for example, understand and implement actions to be taken for the entire organization before providing support to individual employees. Furthermore, in this embodiment, it is possible to understand the situation not only at the organizational level but also at the individual employee level, allowing supervisors and other personnel to take action for the entire organization while also providing support to each employee.

[0060] In this embodiment, the prompt for an organization may include information about the personality test results of employees belonging to the organization. In this case, it is possible to provide advice that takes into account, for example, the personality tendencies of employees belonging to the organization. In addition, the modified examples described in the above-mentioned first embodiment can also be applied to this embodiment as appropriate.

[0061] Independent and dependent claims may be combined with each other in any and all combinations, regardless of the reference format. Furthermore, while the claims may be written in a format in which a claim references two or more other claims (multiple claim format), this is not a limitation. Multiple claims that reference at least one other multiple claim (multiple multiple claim format) may also be written.

[0062] The embodiments disclosed herein are illustrative in all respects and should not be considered limiting. The scope of the present invention is defined by the claims, not by the above meaning, and is intended to include all modifications within the meaning and scope of the claims. [Explanation of symbols]

[0063] 1. Server (information processing device) 11 Control section 12 Main memory 13 Communications Department 14 Auxiliary storage 2. Supervisor's terminal 21 Control Unit 22 Main memory 23 Communications Department 24 Display section 25 Input section 26 Auxiliary storage M language model

Claims

1. Periodically collect employee responses to workplace surveys, Obtaining a score for each of a plurality of categories calculated based on the obtained answers; Obtaining a time change in the score for each category calculated based on the score for each category calculated from multiple answers periodically obtained; Accepting questions regarding the treatment of said employees, creating a prompt that instructs the generation of advice regarding a response to the employee based on the received question, the score for each category, and a change in the score for each category over time; The created prompt is input into a language model to obtain advice on how to respond to the employee, the advice being generated by the language model. An information processing method in which processing is performed by a computer.

2. creating a sentence for each of the plurality of categories according to the score for each category; Create the prompts that include the questions received and the sentences you create for each category. The information processing method according to claim 1 , wherein the processing is executed by the computer.

3. Sentence templates are stored in the memory in association with each category, For each category, a sentence corresponding to the score for each category is inserted into the sentence template to create a sentence corresponding to the score. The information processing method according to claim 2 , wherein the processing is executed by the computer.

4. Accepting a selection from a plurality of pre-prepared questions regarding the response to the employee 4. The information processing method according to claim 1, wherein the processing is executed by the computer.

5. Obtaining test results for the employee's personality test; Using the language model, advice regarding how to respond to the employee is generated based on the acquired test results, the received question, and the score for each category.

4. The information processing method according to claim 1, wherein the processing is executed by the computer.

6. The scores include scores for a plurality of categories that represent the employee's psychological state regarding the work environment and work content, and a total score that is calculated based on the scores for the plurality of categories and represents the state of the employee.

4. The information processing method according to claim 1.

7. outputting a screen that displays scores for the plurality of categories and has operation buttons for receiving questions regarding the employee's response; When an operation on the operation button is accepted, a screen is output that displays a plurality of prepared question options regarding a response to the employee, When a selection of one of the options is accepted, a screen is output that displays the advice generated based on the selected question and the score for each category.

4. The information processing method according to claim 1, wherein the processing is executed by the computer.

8. For each organization established in the workplace, a score for each category in each organization is calculated based on the scores for each category of employees belonging to each organization; We will be accepting questions regarding the treatment of employees of either organization. Based on the received question and the scores for each category in the organization, a language model is used to generate advice on how to respond to the employee of the organization.

4. The information processing method according to claim 1, wherein the processing is executed by the computer.

9. creating a prompt that instructs the generation of advice regarding a response for an employee of the organization based on the received question and the score for each category in the organization; Inputting the created prompt into the language model to obtain a response recommendation generated by the language model for an employee of the organization. The information processing method according to claim 8 , wherein the processing is executed by the computer.

10. Periodically collect employee responses to workplace surveys, Obtaining a score for each of a plurality of categories calculated based on the obtained answers; Obtaining a time change in the score for each category calculated based on the score for each category calculated from multiple answers periodically obtained; Accepting questions regarding the treatment of said employees, creating a prompt that instructs the generation of advice regarding a response to the employee based on the received question, the score for each category, and a change in the score for each category over time; The created prompt is input into a language model to obtain advice on how to respond to the employee, the advice being generated by the language model. A program that causes a computer to perform a process.

11. In an information processing device having a control unit, The control unit Periodically collect employee responses to workplace surveys, Obtaining a score for each of a plurality of categories calculated based on the obtained answers; Obtaining a time change in the score for each category calculated based on the score for each category calculated from multiple answers periodically obtained; Accepting questions regarding the treatment of said employees, creating a prompt that instructs the generation of advice regarding a response to the employee based on the received question, the score for each category, and a change in the score for each category over time; The created prompt is input into a language model to obtain advice on how to respond to the employee, the advice being generated by the language model. Information processing device.

Images