Information processing method, program, and information processing device

Abstract

This provides information processing methods and other tools that can offer advice on communication between employees and their supervisors. [Solution] The computer acquires information on the employee's progress toward work goals, reflections on their work, and consultations. Based on the acquired progress information, reflections, and consultations, the computer uses a language model to generate advice regarding communication between the employee and their supervisor, and outputs the generated advice.

Description

Technical Field

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

Background Art

[0002] In the workplace or the like, interviews are conducted between each employee and the supervisor in order to improve the motivation of employees, build a trust relationship between employees and the supervisor (company), improve productivity, etc., and to grasp the psychological state and troubles of employees. In Patent Document 1, a technique is disclosed for displaying information indicating the work situation of the target person and information about the target person input by an employee (supervisor, etc.) different from the target person so that the respective times correspond.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the technique disclosed in Patent Document 1, it is possible to grasp the work situation of the target person while considering the temporal relationship between the work situation of the target person and the information input by other employees regarding the target person, and it is possible to efficiently review the work situation. However, in the interview with the supervisor, not only the review of the work situation but also the purpose of grasping the psychological state of the employee and resolving troubles, etc., but in Patent Document 1, there is no mention of a technique for supporting the implementation of an interview that can achieve such a purpose.

[0005] In one aspect, an object is to provide an information processing method or the like that can present advice regarding communication between an employee and a supervisor.

Means for Solving the Problems

[0006] One aspect of the information processing method involves a computer that acquires information on an employee's progress toward work goals, reflections on their work, and consultations, generates advice regarding communication between the employee and their supervisor using a language model based on the acquired progress information, reflections, and consultations, and outputs the generated advice. [Effects of the Invention]

[0007] One aspect of this approach is that it can offer advice on communication between employees and their supervisors. [Brief explanation of the drawing]

[0008] [Figure 1] This is an explanatory diagram showing an example of the configuration of an information processing system. [Figure 2] This block shows an example configuration of a server and a supervisor terminal. [Figure 3] This is an explanatory diagram showing an example of a record layout for the employee database. [Figure 4] This is an explanatory diagram showing an example of a prompt template. [Figure 5] This is an explanatory diagram showing an example of a prompt template. [Figure 6] This is an explanatory diagram showing an example of a prompt template. [Figure 7] A flowchart illustrating an example of the process for providing advice on how to conduct a 1-on-1 meeting. [Figure 8] This is an explanatory diagram showing an example of an employee terminal screen. [Figure 9] This is an explanatory diagram showing an example of an employee terminal screen. [Figure 10] This flowchart shows an example of the process for generating information related to 1on1 meetings. [Figure 11] This flowchart shows an example of the process for generating information related to 1on1 meetings. [Figure 12] This is an explanatory diagram showing an example screen of a supervisor's terminal. [Figure 13]This is an explanatory diagram showing an example screen of a supervisor's terminal. [Figure 14] This is an explanatory diagram showing an example screen of a supervisor's terminal. [Modes for carrying out the invention]

[0009] The information processing method, program, and information processing apparatus of this disclosure will be described in detail below with reference to the drawings illustrating their embodiments.

[0010] Figure 1 is an explanatory diagram showing an example of the configuration of an information processing system. In this embodiment, we describe an information processing system that supports 1-on-1 meetings (hereinafter referred to as 1-on-1) in which an employee belonging to a company, etc., and their supervisor (manager) have a one-on-one conversation (interview). The information processing system of this embodiment generates advice related to 1-on-1 meetings using a general-purpose large language model (LLM). In this embodiment, we will explain using Japanese as the language to be input to the large language model, but a configuration using other languages ​​is also possible.

[0011] The information processing system of this embodiment includes a server 10, employee terminals 30 (hereinafter referred to as employee terminals 30), and employee supervisor terminals 20 (hereinafter referred to as supervisor terminals 20), and each device 10, 20, and 30 is communicated with via a network N. Multiple supervisor terminals 20 and employee terminals 30 are provided. Network N may be the Internet or a public telephone network, or a LAN (Local Area Network) built within the facility where the information processing system is installed. Server 10 is a computer capable of various information processing and information transmission and reception, such as a server computer or personal computer. Server 10 may be a server managed by each company, or a server managed by a company that supports the implementation of 1on1 meetings. Supervisor terminals 20 and employee terminals 30 are computers capable of various information processing and information transmission and reception, such as personal computers, tablet terminals, smartphones, etc.

[0012] In the information processing system of this embodiment, the server 10 acquires prior information regarding one-on-one from an employee via the employee terminal 30, generates advice on how to conduct a one-on-one (advice on how to proceed with a one-on-one) based on the acquired prior information, and presents it to the employee via the employee terminal 30 and to the supervisor via the supervisor terminal 20, respectively. Further, the server 10 collects the speech audio during the execution of the one-on-one, identifies the minutes of the meeting and the actions (next actions) that the employee and the supervisor should take based on the collected audio data, and presents them to the employee and the supervisor. Furthermore, after the one-on-one ends, the server 10 generates review information for reviewing the one-on-one based on the prior information, the minutes of the meeting, etc., and presents it to the employee and the supervisor. Thereby, it is possible to support the preparation before the one-on-one, the execution of the one-on-one, and the review after the one-on-one ends.

[0013] FIG. 2 is a block diagram showing a configuration example of the server 10 and the supervisor terminal 20. The server 10 includes a control unit 11, a storage unit 12, a communication unit 13, an input unit 14, a display unit 15, a reading unit 16, and the like. The control unit 11 includes one or more processors (arithmetic processing units) such as a CPU (Central Processing Unit), MPU (Micro-Processing Unit), GPU (Graphics Processing Unit), TPU (Tensor Processing Unit), or AI chip (semiconductor for AI). The control unit 11 executes the processes that the server 10 should perform by reading and executing the program P1 stored in the storage unit 12. When the control unit 11 includes a plurality of processors, each process may be executed by the same processor, or each process may be executed by different processors.

[0014] The memory unit 12 includes RAM (Random Access Memory), flash memory, hard disk, SSD (Solid State Drive), etc. The memory unit 12 stores the program P1 (program product, computer program) executed by the control unit 11 and various data. The memory unit 12 also temporarily stores data generated when the control unit 11 executes program P1. The server 10 has the functionality of a web server, and the memory unit 12 stores the 1on1 site S to support the implementation of 1on1 meetings. The memory unit 12 also stores the language model M. The language model M is a general-purpose large-scale language model constructed by pre-training with a large set of documents, etc. The language model M is trained to, for example, when input data including text data written in natural language is input, perform calculations to generate output data according to the content of the input data based on the input data, and output the generated output data. The language model M can be constructed using algorithms such as GPT (Generative Pre-trained Transformer)-3, GPT-3.5, and GPT-4, or by combining multiple algorithms. The language model M is not limited to the transformer-based models described above. The language model M is intended to be used as a program module constituting artificial intelligence software. The language model M performs predetermined calculations on input data and outputs the calculation results. The memory unit 12 stores data such as the coefficients and thresholds of the function that defines these calculations as the language model M. Instead of storing the language model M in the memory unit 12, the server 10 may access and read the language model M from a language processing server. Furthermore, the language model M may be fine-tuned using information from past 1on1 meetings. In addition to fine-tuning the entire model, the language model M may also be fine-tuned using techniques such as LoRA (Low-Rank Adaptation) for only some of its layers. The memory unit 12 also stores the employee database 121 and prompt templates 122, which will be described later.Note that the storage unit 12 may be composed of a plurality of storage devices. A part of the storage unit 12 may be another storage device connected to the server 10, or may be another storage device with which the server 10 can communicate.

[0015] The communication unit 13 is a communication module for performing processing related to wired communication or wireless communication, and transmits and receives information to and from other devices via the network N. The input unit 14 receives operation inputs by the user and sends a control signal corresponding to the operation content to the control unit 11. The display unit 15 is a liquid crystal display, an organic EL display, or the like, and displays various types of information according to an instruction from the control unit 11. The input unit 14 and the display unit 15 may be a touch panel configured integrally.

[0016] The reading unit 16 reads information stored in a portable storage medium 10a such as a CD (Compact Disc), a DVD (Digital Versatile Disc), a USB (Universal Serial Bus) memory, an SD card, a micro SD card, a Compact Flash (registered trademark), or the like. The program P1 and various types of data stored in the storage unit 12 may be read by the control unit 11 from the portable storage medium 10a via the reading unit 16 and stored in the storage unit 12. Further, the program P1 and various types of data may be written in the storage unit 12 at the manufacturing stage of the server 10, or may be downloaded by the control unit 11 from other devices via the communication unit 13 and stored in the storage unit 12.

[0017] In this embodiment, the server 10 is not limited to a single computer, but may be a multi-computer consisting of multiple computers, or a virtual machine virtually constructed by software within a single device. Furthermore, the server 10 may be a local server installed in a facility where the server 10 (information processing system) is located, or it may be a cloud server connected via network N. The server 10 may consist of multiple servers, each realizing multiple functions such as a function for performing support processing related to the preparation for a 1on1, a function for performing support processing during a 1on1, and a function for performing support processing after the 1on1 is completed. The program 12P may be deployed and executed on a single computer or at a single site, or it can be distributed across multiple sites and deployed to run on multiple computers interconnected via network N. Moreover, the input unit 14 and the display unit 15 are not mandatory for the server 10; it may be configured to accept operations through a connected computer, or it may be configured to output the information to be displayed to an external display device.

[0018] The supervisor terminal 20 includes a control unit 21, a storage unit 22, a communication unit 23, an input unit 24, a display unit 25, a microphone 26, etc. The control unit 21, storage unit 22, communication unit 23, input unit 24, and display unit 25 each have the same configuration as parts 11 to 15 of the server 10, so their explanation is omitted. The storage unit 22 of the supervisor terminal 20 stores a web browser (hereinafter referred to as browser AP) for accessing the web server, in addition to the program P2 executed by the control unit 21. The employee terminal 30 has the same configuration as the supervisor terminal 20, so its explanation is omitted. That is, although not shown in the diagram, the employee terminal 30 has a control unit 31, a storage unit 32, a communication unit 33, an input unit 34, and a display unit 35, which have the same configuration as parts 21 to 25 of the supervisor terminal 20.

[0019] Figure 3 is an explanatory diagram showing an example of the record layout of Employee DB121. Employee DB121 is a database that stores information about employees belonging to companies and other organizations using this system. Employee DB121, as shown in Figure 3, includes columns such as Employee ID, Name, Employee Information, Personality Test Results, and 1on1 Information, and stores various information about employees associated with the Employee ID. The Employee ID column stores the identification information (Employee ID) assigned to each employee. The Name column stores the employee's name, and the Employee Information column stores information about the employee, such as their department, position, length of service, and work history. The Personality Test Results column stores information representing personality traits obtained from personality tests taken by employees. The information representing personality traits includes scores for each item representing personality traits, and personality types classified based on the scores of each item. Personality tests are conducted immediately after joining the company, once every few months to several years, and at appropriate times such as when an employee changes departments. The test results from each test may be stored, or only the most recent test result may be stored. The 1on1 information column stores information related to 1on1s, such as the date and time, the supervisor ID (employee ID) of the supervisor conducting the 1on1, pre-meeting information, meeting minutes, next actions, survey results, and reflection information. The pre-meeting information includes information entered (registered) by the employee and supervisor before the 1on1 is conducted, and the meeting minutes and next actions include meeting minutes and next actions generated and extracted from the content of the exchange between the employee and supervisor during the 1on1. The survey results include the results of the survey answered by the employee after the 1on1 is conducted, and the reflection information includes the content of the exchange during the 1on1 after it has been conducted. The employee DB121 is not limited to the configuration shown in Figure 3. For example, if a well-being survey is conducted on employees on a regular basis, information regarding the survey results may be stored along with the date and time of implementation. In addition, the employee DB121 may be divided into multiple DBs, with each piece of information registered in each, and some of the information may be registered in DBs on other servers. For example, employee personality test results or 1-on-1 information may be stored on another server that manages employee information, associated with employee IDs, etc.

[0020] As described above, Server 10 stores the results of each employee's personality test in Employee DB 121. Server 10 also obtains preliminary information from employees whose 1on1 sessions have been scheduled and stores it in Employee DB 121. Based on this preliminary information and the personality traits of the employee and their supervisor, Server 10 uses Language Model M to generate and present advice on how to conduct the 1on1 session. Server 10 also collects the content of the conversation between the employee and their supervisor during the 1on1 session and generates meeting minutes using Language Model M based on the collected conversation content. Furthermore, after the 1on1 session is completed, Server 10 uses the Language Model to generate 1on1 review information based on various information related to the 1on1 session and presents it to the employee and their supervisor. Note that Language Model M may be a general-purpose large-scale Language Model, or it may be a finely tuned version using information on the employee's (subordinate's) personality traits obtained from the personality test results, corporate information of the company using this system, etc.

[0021] The following describes the prompts that the server 10 inputs to the language model M when generating various types of information (advice, meeting minutes, etc.) using the language model M. Figures 4 to 6 are explanatory diagrams showing examples of prompt templates 122. Figures 4A and 4B show examples of prompt templates for generating advice on how to conduct a 1on1 based on prior information, etc. Figure 4A shows a prompt template for generating advice for an employee, and Figure 4B shows a prompt template for generating advice for a supervisor. The prompt template shown in Figure 4A includes a command statement instructing the generation of advice on how to conduct a 1on1, text about prior information entered by the employee in advance (before the 1on1 is conducted), text about the personalities of the employee and supervisor, and text specifying the output format. The text about the employee's prior information includes information entered via the input screen shown in Figure 8B, which will be described later. The text about personalities includes the personality types of the employee and supervisor obtained from the personality test, and the scores of each item of personality traits included in the personality test results, or explanatory text corresponding to the scores. The text about personalities may also include information (explanatory text) about the relationship that takes into account the personality traits of the employee and supervisor. The explanatory text corresponding to the score of each item is stored in the memory unit 12 in advance according to the score. The text regarding the specification of the output format includes a message for controlling the output result from the language model M and a message regarding constraints on the output result. The prompt template shown in Figure 4B has the same configuration as the prompt template shown in Figure 4A, but the command statement and the text regarding the specification of the output format have been changed for the supervisor. By creating and inserting text regarding the employee's prior information, as well as text regarding the personalities of the employee and the supervisor, into the prompt template shown in Figure 4A or Figure 4B, a prompt is created for generating advice on how to conduct a 1on1 using the language model M. The prompt template for generating advice on how to conduct a 1on1 is not limited to the example shown in Figure 4. For example, if past information about the employee's prior information is stored in the employee DB 121, the past prior information may be included in the prompt. That is, time-series data of the prior information may be included in the prompt.Additionally, past 1-on-1 information (meeting minutes and review information) stored in employee DB121 may be included in the prompt. Furthermore, the personality-related text may be structured to include descriptions of the personality of at least one employee and their supervisor in the prompt.

[0022] Figure 5A shows an example of a prompt template for generating 1on1 meeting minutes. The prompt template shown in Figure 5A includes an instruction to generate 1on1 meeting minutes, a sample format for the minutes and text regarding constraints, and transcribed text data. The transcribed text data is obtained by converting audio data collected during the 1on1 meeting into text. The text regarding constraints may include messages indicating constraints on the content of the minutes, as well as messages to control the output from the language model M. By creating and inserting transcribed minutes into the prompt template shown in Figure 5A, a prompt is created for generating 1on1 meeting minutes using the language model M. The prompt template for generating meeting minutes is not limited to the example shown in Figure 5A. For example, information about the personality traits of the employee and the supervisor may be included in the prompt. This makes it possible to create minutes that extract important points considering the personality traits of the employee and the supervisor. Information from past 1on1 meetings may also be included in the prompt.

[0023] Figure 5B shows an example of a prompt template for identifying (generating) next actions based on 1on1 meeting minutes, etc. The prompt template shown in Figure 5B includes an instruction statement that directs the generation of next actions, a document defining and formatting the next actions, and the 1on1 meeting minutes. By creating and inserting 1on1 meeting minutes into the prompt template shown in Figure 5B, prompts are created for generating next actions for employees and supervisors using the language model M. The prompt template for generating next actions is not limited to the example shown in Figure 5B. Here too, information about the personality traits of employees and supervisors may be included in the prompt. This makes it possible to identify next actions that take into account the personality traits of employees and supervisors. Information from past 1on1 meetings may also be included in the prompt.

[0024] Figure 6 shows an example of a prompt template for generating 1on1 review information. The prompt template shown in Figure 6 includes an instruction to generate 1on1 review information, text about the employee's prior information, text about the employee's and supervisor's personalities, 1on1 meeting minutes, text about the employee's and supervisor's next actions, text about the employee's survey results regarding the 1on1, and text specifying the output format. By creating and inserting text about the employee's prior information, the employee's and supervisor's personalities, 1on1 meeting minutes, text about next actions, and text about the employee's survey results into the prompt template shown in Figure 6, a prompt is created for generating 1on1 review information using the language model M. The prompt template for generating 1on1 review information is not limited to the example shown in Figure 6; for example, the prompt may also include the employee's prior information and past 1on1 information.

[0025] The following describes the process in which, in the information processing system of this embodiment, the server 10 generates and presents advice on how to conduct a 1-on-1 meeting based on the employee's prior information obtained via the employee terminal 30. Figure 7 is a flowchart showing an example of the procedure for presenting advice on how to conduct a 1-on-1 meeting, and Figures 8 and 9 are explanatory diagrams showing examples of the screen of the employee terminal 30.

[0026] In the information processing system of this embodiment, employees determine a scheduled date and time for a 1-on-1 meeting with their supervisor. Subsequently, at any time before the 1-on-1 meeting, the employee accesses server 10 (1on1 site S) based on a predetermined URL (Uniform Resource Locator) and inputs (sets) preliminary information for the 1on1 meeting. If 1on1 site S is linked to a scheduler that manages each employee's schedule, the employee may also check their supervisor's schedule via 1on1 site S and adjust the date accordingly.

[0027] When the control unit 11 of the server 10 receives an access request to the 1on1 site S from the employee terminal 30, it sends an initial screen as shown in Figure 8A to the employee terminal 30. The control unit 31 of the employee terminal 30 displays the initial screen received from the server 10 on the display unit 35. The screen in Figure 8A has an input field A1 for the employee's department, and input field A1 has a pull-down menu for selecting any one of several selectable departments. When any department is selected via input field A1, the control unit 31 displays a list of names of employees belonging to the selected department, and as a result, the screen in Figure 8A is displayed. The screen in Figure 8A has a "Set up 1on1" button associated with each employee, and when the "Set up 1on1" button is operated, the control unit 31 displays an input field for the date and time of the 1on1, as shown in Figure 8A. The employee operates the "Set up 1on1" button corresponding to their name, enters the date and time of the 1on1, and then operates the create button to register the 1on1 at the entered date and time.

[0028] The control unit 31 of the employee terminal 30 determines whether or not it has received input for the date and time of the 1on1 meeting via the input field (S11). If it determines that it has not received the input (S11: NO), it waits until it does. If it determines that it has received input for the date and time of the 1on1 meeting (S11: YES), the control unit 31 displays a pre-meeting information input screen, as shown in Figure 8B, on the display unit 35 (S12). The screen in Figure 8B has input fields that accept input for pre-meeting information, including the employee's condition (mood), progress information toward work goals, reflection information on work (what went well, what didn't go well, etc.), and consultation information, as preparation for the 1on1 meeting. The employee's condition is entered as one of five levels, with the worst state represented by a rain icon and the best state by a sunny icon. Progress information is entered as progress toward the set goal (for example, progress shown with a minimum value of 0 and a maximum value of 100), importance, feasibility, and motivation (for example, a score shown with a minimum value of 1 and a maximum value of 5). The goals here may include goals related to work content, medium- to long-term career goals, etc. The control unit 31 may also retrieve information about the employee's past 1on1 meetings from the server 10 and display it on the pre-information input screen. In the example in Figure 8B, the date and time of the previous (past) 1on1, meeting minutes, next actions, and reflection information (advice for the next meeting) are displayed on the left side of the screen. This allows the employee to input pre-information while considering the content of the previous 1on1 meeting.

[0029] The control unit 31 accepts pre-information input via the screen shown in Figure 8B (S13) and displays each entered piece of information in its respective input field. In addition to accepting information input into each input field by employees, the control unit 31 may also be linked with a system that manages employee work progress, etc., and extract and input the information to be entered into each input field from that system. The control unit 31 determines whether the registration button on the screen has been operated (S14). If it determines that it has not been operated (S14: NO), it returns to step S13 and continues to accept pre-information input. If it determines that the registration button has been operated (S14: YES), the control unit 31 sends the entered pre-information along with the employee ID and the date and time of implementation to the server 10 (S15). The control unit 11 of the server 10 acquires the pre-information from the employee terminal 30 and stores the acquired date and time of implementation and pre-information in the employee DB 121, associating them with the acquired employee ID (S16). Specifically, the control unit 11 acquires and stores information on the employee's progress toward their work goals, reflections on their work, and consultation information (prior information). The control unit 11 also identifies the supervisor with whom the employee will conduct a 1on1 meeting based on the employee ID, and stores the supervisor ID of the identified supervisor in the employee database 121. If the supervisor conducting the 1on1 meeting is pre-configured, it can be identified from a table where the correspondence between each employee and their supervisor is registered. Furthermore, the supervisor conducting the 1on1 meeting is not limited to the pre-configured supervisor; it may also be specified by the employee via the prior information input screen, in which case the control unit 11 can receive the supervisor ID from the employee terminal 30.

[0030] The screen in Figure 8B includes a "Request Advice" button for employees to request advice on conducting 1on1 meetings. The control unit 31 determines whether the "Request Advice" button (labeled "Advice Button" in Figure 7) has been operated (S17). If it determines that the button has not been operated (S17: NO), it waits. The control unit 31 also terminates processing when it is instructed to end the display of the 1on1 site S screen. If the control unit 31 determines that the "Request Advice" button has been operated (S17: YES), it sends the employee ID to the server 10 to request advice (S18).

[0031] When advice is requested, the control unit 11 of server 10 creates a prompt to generate advice for the employee on how to conduct a 1-on-1 (S19). Here, the control unit 11 reads the prior information entered by the employee and the personality test results (personality traits) of the employee and their supervisor from the employee DB 121. Then, the control unit 11 creates the prompt by inserting the employee's prior information and the personality traits of the employee and their supervisor into the prompt template shown in Figure 4A. At that time, the control unit 11 may also create explanatory text for the employee's prior information and the personality traits of the employee and their supervisor and insert it into the prompt template.

[0032] The control unit 11 inputs the created prompt into the language model M, and the language model M generates advice for the employee on how to conduct the 1on1 (S20). That is, based on the prior information entered by the employee (progress information on work goals, reflection information on work, and consultation information), the control unit 11 uses the language model M to generate advice (advice for the employee) regarding the 1on1 (communication) between the employee and their supervisor. The control unit 11 retrieves the advice generated by the language model M from the language model M and sends it to the employee terminal 30 (S21), and the control unit 31 of the employee terminal 30 retrieves the advice from the server 10 and displays it on the display unit 35 (S22). For example, as shown in Figure 9, the control unit 31 displays the advice on the right side of the prior information input screen. Through the above process, advice for conducting 1on1s can be presented to the employee, taking into account the prior information and the personality traits of the employee and their supervisor. Furthermore, if information from past 1on1s is included in the prompt, advice that takes past 1on1 information into consideration can also be presented. Therefore, employees can use the advice as a reference when conducting one-on-one meetings with their supervisors, enabling them to have more meaningful meetings. Furthermore, if work objectives include medium- to long-term career goals, advice can be provided that takes into account the employee's personality traits and medium- to long-term career goals, thus supporting communication between employees and supervisors regarding work situations. The advice display area may also be configured to accept questions about the advice in a chatbot format. In this case, employees can accurately understand the advice by asking further questions and conduct one-on-one meetings efficiently in accordance with the advice.

[0033] Next, we will explain the process by which server 10 generates meeting minutes based on the content of speech acquired by the supervisor terminal 20 during the 1on1 meeting, and generates reflection information after the 1on1 meeting. Figures 10 and 11 are flowcharts showing an example of the procedure for generating information related to the 1on1 meeting, and Figures 12 to 14 are explanatory diagrams showing example screens of the supervisor terminal 20.

[0034] A supervisor conducting a 1-on-1 with a subordinate (employee) accesses server 10 (1on1 site S) based on a designated URL before or during the 1on1 to confirm the agenda for that 1on1. The following explanation describes the procedure for starting the 1on1 immediately after the supervisor confirms the agenda, but agenda confirmation is not limited to immediately before the 1on1; the supervisor can confirm the agenda at any time before the 1on1.

[0035] When the control unit 11 of the server 10 receives access to the 1on1 site S from the supervisor terminal 20, it reads the pre-information entered by the employee who will be having a 1on1 with their supervisor from the employee DB 121, generates a pre-information screen displaying the pre-information as shown in Figure 12A, and sends it to the supervisor terminal 20. The screen displayed here may also show the date and time of the previous (past) 1on1, meeting minutes, next actions, and review information (advice for the next time), similar to the screens in Figures 8B and 9. The control unit 21 of the supervisor terminal 20 displays the pre-information screen received from the server 10 on the display unit 25 (S31). The screen in Figure 12A has an "Confirm Agenda" button for confirming the agenda for the 1on1 based on the displayed pre-information. The control unit 21 of the supervisor terminal 20 determines whether the "Confirm Agenda" button has been operated (S32), and if it determines that it has not been operated (S32: NO), it waits.

[0036] If the control unit 21 determines that the "Confirm Agenda" button has been pressed (S32: YES), it sends the supervisor's supervisor ID and the employee ID of the employee conducting the 1on1 to the server 10 to request advice (agenda) (S33). If advice is requested, the control unit 11 of the server 10 creates a prompt to instruct the server 10 to generate advice for the supervisor on how to proceed with the 1on1 (S34). Here, the control unit 11 reads the preliminary information entered by the employee subject to the 1on1 and the personality test results of the employee and the supervisor from the employee DB 121. Then, the control unit 11 creates the prompt by inserting the employee's preliminary information and the personality traits of the employee and the supervisor into the prompt template shown in Figure 4B. Here as well, the control unit 11 may create sentences explaining the employee's preliminary information and the personality traits of the employee and the supervisor and insert them into the prompt template.

[0037] The control unit 11 inputs the created prompt into the language model M, and the language model M generates advice (agenda) for the supervisor (S35). The control unit 11 sends the advice generated by the language model M to the supervisor terminal 20 (S36), and the control unit 21 of the supervisor terminal 20 retrieves the advice from the server 10 and displays it on the display unit 25 (S37). For example, as shown in Figure 12B, the control unit 21 displays the advice (agenda) on the right side of the pre-information screen. In the screen of Figure 12B, information (advice) regarding two agendas is displayed, and each agenda has a checkbox, which allows the user to select the agenda to be discussed in the 1on1. When one or more agendas are selected via the checkboxes and the "Start 1on1" button is pressed, the control unit 21 stores the selected agenda in the storage unit 22 as the agenda for this 1on1. The control unit 21 may also send the agenda for this 1-on-1 meeting to the server 10 and store it in the employee database 121 as 1-on-1 information.

[0038] The control unit 21 determines whether the "Start 1on1" button on the screen has been operated (S38). If it determines that the button has not been operated (S38: NO), it waits while accepting input from the checkbox. If the control unit 21 determines that the "Start 1on1" button has been operated (S38: YES), it displays the "1on1 in progress" screen on the display unit 25 (S39). The "in progress" screen, as shown in Figure 12B, displays a "Start transcription" button and an "End 1on1" button on the pre-information screen. The "Start transcription" button instructs the start of the transcription process, which converts the conversation between the supervisor and employee during the 1on1 into text data and displays it. The "End 1on1" button instructs the end of the 1on1.

[0039] The control unit 21 determines whether the "Start transcription" button has been pressed on the execution screen (S40). If it determines that the button has not been pressed (S40: NO), it waits. If the control unit 21 determines that the "Start transcription" button has been pressed (S40: YES), it collects audio data (dialogue information) spoken by the supervisor and employee via the microphone 26 (S41), converts the collected audio data into text data (S42), and displays the obtained text data on the display unit 25 as shown in Figure 13A (S43). In the screen of Figure 13A, preliminary information is displayed on the left, the agenda is displayed in the center, and the transcribed text data is displayed on the right. In addition to being converted into text data at the supervisor terminal 20, the audio data collected via the microphone 26 may also be sent to the server 10 and converted into text data at the server 10. In this case, the text data converted at the server 10 is sent to the supervisor terminal 20 and displayed as shown in Figure 13A.

[0040] The screen in Figure 13A is equipped with a "Summarize" button to instruct the generation of 1on1 meeting minutes based on the transcribed text data, and a "Extract Next Actions" button to instruct the identification of actions (next actions) that the supervisor and employee should take. Therefore, the control unit 21 determines whether or not the "Summarize" button has been operated (S44), and if it determines that it has been operated (S44: YES), it sends the text data displayed on the screen to the server 10 to request the generation of meeting minutes (S45). When the control unit 11 of the server 10 receives a request to generate meeting minutes, it creates a prompt to instruct the generation of meeting minutes (S46). Here, the control unit 11 creates the prompt by inserting the text data obtained from the supervisor terminal 20 into the prompt template shown in Figure 5A.

[0041] The control unit 11 inputs the created prompt into the language model M, and the language model M generates the minutes of the 1on1 meeting (S47). The control unit 11 stores the generated minutes in the employee DB 121, associating them with the employee ID of the employee who is the subject of the 1on1. The control unit 11 sends the minutes generated by the language model M to the supervisor terminal 20 (S48), and the control unit 21 of the supervisor terminal 20 retrieves the minutes from the server 10 and displays them on the display unit 25 (S49). For example, as shown in Figure 13B, the control unit 21 displays the minutes below the transcribed text data. If the control unit 21 determines that the "Summarize" button has not been operated (S44: NO), it proceeds to step S50 and determines whether the "Extract Next Action" button (labeled as the "Next Action" button in Figure 11) has been operated (S50).

[0042] If the control unit 21 determines that the "Extract Next Action" button has been pressed (S50: YES), it sends the text data displayed on the screen (transcribed text data) to the server 10 to request the generation of the next action (S51). If the control unit 11 of the server 10 is requested to generate the next action, it creates a prompt to instruct the server to generate the next action (S52). Here, the control unit 11 creates the prompt by inserting the text data obtained from the supervisor terminal 20 into the prompt template shown in Figure 5B. That is, in the prompt template of Figure 5B, the text data obtained from the supervisor terminal 20 is inserted in place of the 1on1 meeting minutes.

[0043] The control unit 11 inputs the created prompt into the language model M, and the language model M generates information about the next action (action information) (S53). Here again, the control unit 11 stores the generated information about the next action in the employee DB 121, associating it with the employee ID. The control unit 11 sends the information about the next action generated by the language model M to the supervisor terminal 20 (S54), and the control unit 21 of the supervisor terminal 20 retrieves the information about the next action from the server 10 and displays it on the display unit 25 (S55). For example, as shown in Figure 13B, the control unit 21 displays the information about the next action below the meeting minutes. The display field for the next action is provided with a "Create from Meeting Minutes" button to instruct the creation of the next action from the meeting minutes. When the "Create from Meeting Minutes" button is pressed, the control unit 21 sends the meeting minutes displayed on the screen to the server 10 to request the generation of the next action. In this case, the control unit 11 of the server 10 creates a prompt by inserting the meeting minutes obtained from the supervisor terminal 20 into the prompt template shown in Figure 5B, and then inputs the created prompt into the language model M to generate information about the next action.

[0044] If the control unit 21 determines that the "Extract Next Action" button has not been operated (S50: NO), it proceeds to step S56 and determines whether the "End 1on1" button has been operated (S56). If the control unit 21 determines that the "End 1on1" button has been operated (S56: YES), it displays a questionnaire screen as shown in Figure 14A on the display unit 25 (S57). The screen in Figure 14A displays a questionnaire for the employee, and accepts input of a 5-point scale evaluation for "Goal Fit" regarding whether the work goals discussed in the 1on1 are feasible, should be done, or would like to do. It also accepts input of a 5-point scale evaluation for "1on1 Satisfaction" regarding whether work progress was shared with the supervisor (Progress Sharing), whether the matters to be discussed were discussed (Consultation Implementation), and whether future activities were understood (Prospects for Actions).

[0045] The control unit 21 receives the questionnaire response via the input unit 24 (S58). If the OK button on the questionnaire screen is pressed, it sends the meeting minutes and next action information displayed on the screen in Figure 13B, along with the questionnaire response, to the server 10 to request the generation of 1on1 review information (S59). When the server 10's control unit 11 receives a request to generate review information, it creates a prompt to instruct the server to generate the review information (S60). Here, the control unit 11 creates the prompt by inserting the meeting minutes and next action information obtained from the supervisor terminal 20, the employee's questionnaire results (questionnaire response), prior information, and the personality traits of the employee and supervisor into the prompt template shown in Figure 6. The control unit 11 stores the meeting minutes, next action information, and questionnaire response (questionnaire result) obtained from the supervisor terminal 20 in the employee DB 121, associating them with the employee ID.

[0046] The control unit 11 inputs the created prompt to the language model M, and the language model M generates 1on1 review information (S61). The control unit 11 stores the review information generated by the language model M in the employee DB 121, associating it with the employee ID of the employee who is the subject of this 1on1 (S62). If the control unit 21 of the supervisor terminal 20 determines that the "End 1on1" button has not been pressed (S56: NO), it returns to step S41 and repeats the collection of audio data and conversion to text data. The control unit 21 is configured to convert the collected audio data into text data and display it at predetermined intervals (e.g., 10 seconds, 30 seconds, etc.) after the "Start transcription" button is pressed. As a result, the transcribed text data is updated and displayed at predetermined intervals.

[0047] The above process allows for the presentation of advice (agenda) to supervisors regarding the 1on1, taking into account the pre-meeting information entered by the employee before the 1on1, as well as the personality traits of both the employee and the supervisor. If information from past 1on1s is included in the prompts, the advice can also take past 1on1 information into consideration. Therefore, supervisors can approach the 1on1 with a correct understanding of the employee's situation by referring to the advice, enabling a more meaningful 1on1. The advice display area may also be configured in a chatbot format to accept questions regarding the advice. Furthermore, text data of the supervisor's and employee's speech can be collected during the 1on1, and meeting minutes and information on next actions can be generated and presented from the transcribed text data. Therefore, supervisors can appropriately proceed with the 1on1 while referring to the advice (agenda) and meeting minutes, enabling appropriate discussion on measures to achieve goals, and as a result, clarifying the employee's next actions. Furthermore, after the 1on1 meeting, a review of the 1on1 can be generated and presented from the meeting minutes and information on next actions. Therefore, after the 1on1, supervisors and employees can confirm future activities (next actions) based on the meeting minutes, information on next actions, and the review information, and support employees in realizing those next actions. The above process enables meaningful 1on1 meetings, supporting the resolution of stress and worries felt by employees, improving goal achievement rates, and is expected to improve employee performance. Thus, by using 1on1 meetings to understand employees' current situation regarding work goals, identify issues and consider countermeasures, and clarify next actions, and by reviewing after the 1on1 whether the content discussed in the 1on1 has been implemented, it becomes possible to realize a good management cycle.

[0048] In the processes shown in Figures 10 and 11, the pre-1on1 processes (steps S31-S37), the in-1on1 processes (steps S38-S55), and the post-1on1 processes (steps S56-S62) may be executed at different times. That is, by performing steps S31-S37 at any time before the 1on1, the supervisor can confirm the employee's prior information and agenda before the 1on1. Furthermore, by performing steps S38-S55 during the 1on1, the supervisor can confirm the progress of the 1on1 (whether the discussion is proceeding according to the agenda). In addition, by performing steps S56-S62 after the 1on1, both the supervisor and the employee can confirm the actions they should take going forward.

[0049] Figure 14B is an explanatory diagram showing an example of a 1on1 feedback screen. As described above, each device 10, 20, and 30 performs its respective processing before, during, and after the 1on1, so that information about the 1on1 is stored in the employee DB 121. Therefore, the supervisor can access the 1on1 site S using the supervisor terminal 20 and view the feedback screen shown in Figure 14B on the supervisor terminal 20 for review. Similarly, the employee can access the 1on1 site S using the employee terminal 30 and view a feedback screen with a similar configuration on the employee terminal 30, allowing them to review the 1on1. When a feedback screen is requested, the control unit 11 of the server 10 reads the information about the 1on1 stored in the employee DB 121, generates a feedback screen based on the read information, and sends it to the supervisor terminal 20 or employee terminal 30. The screen in Figure 14B displays the score for the previous (most recent) 1on1, including the degree of achievement against business goals, the sense of goal fit and 1on1 satisfaction included in the survey results, and the status of the 1on1. The degree of achievement against business objectives is calculated based on the progress information for each objective included in the pre-information. The 1on1 implementation status includes the degree of preparation calculated based on the employee's input of pre-information and the degree of recording calculated based on the recording status during the 1on1. Each of these pieces of information is calculated by the control unit 11. In addition, the screen in Figure 14B displays reflection information (a brief summary of the previous session, what should be done next time based on the content of the previous session), next actions, pre-information (previous input), and meeting minutes. Through such a feedback screen, supervisors and employees can grasp the employee's status, concerns, work progress, etc., and confirm their next actions. Therefore, by checking the feedback screen even after the 1on1 has been conducted, it is possible to utilize the results of the 1on1 in future actions, thereby realizing a meaningful 1on1. Furthermore, the screen shown in Figure 14B may be provided to managers who manage employees and supervisors, such as the executives or personnel department staff of the company to which the employee and supervisor belong.In this case, the administrator accesses the 1on1 site S using their own terminal, and the control unit 11 of the server 10 generates the feedback screen shown in Figure 14B and sends it to the administrator's terminal, allowing the administrator to understand the status of 1on1s conducted by each employee and their supervisor. This ensures that 1on1s are not solely between the employee and the supervisor, but also allow management and HR personnel to understand whether supervisors are providing appropriate management to employees by considering time-series data such as the status of 1on1s, the progress of work discussed during the 1on1s, reflection information, and consultation information, as well as any notes (e.g., next actions) generated using the language model M.

[0050] The independent and dependent claims described in the claims can be combined with each other in any combination, regardless of the form of reference. Furthermore, while the claims use a multi-claim format in which claims refer to two or more other claims (multi-claim format), this is not the only option. Claims may also be described using a multi-claim format in which at least one multi-claim is referenced (multi-multi-claim format).

[0051] The embodiments disclosed herein should be considered in all respects to be illustrative and not restrictive. The scope of the invention is indicated by the claims, not in the sense described above, and all modifications are intended to be in the sense and scope equivalent to the claims. [Explanation of symbols]

[0052] 10. Server (Information Processing Device) 11 Control Unit 12 Storage section 13 Communications Department 20 Supervisor terminal 21 Control Unit 22 Memory section 23 Communications Department 24 Input section 25 Display section 26 Mike 30 Employee terminals 122 Prompt Templates M language model

Claims

1. We obtain information on employees' progress toward their work goals, information on their work reflections, and information on consultations. Based on the acquired progress information, reflection information, and consultation information, a language model is used to generate advice regarding communication between the employee and their supervisor. Output the generated advice. An information processing method in which a computer performs the processing.

2. When a request for advice is received from the employee, the language model is used to generate advice for the employee. The generated advice is output to the employee's terminal. When a request for advice is received from a supervisor who is having a one-on-one conversation with the employee, the language model is used to generate advice for the supervisor. Output the generated advice to the supervisor's terminal. The information processing method according to claim 1, wherein the processing is performed by the computer.

3. The personality information of the employee and / or the personality information of the supervisor is obtained. Based on the acquired progress information, reflection information, consultation information, and personality information, the language model is used to generate the advice. The information processing method according to claim 1 or 2, wherein the computer performs the processing.

4. The employee's past progress information, reflection information, and consultation information are acquired. Based on the time-series data of the progress information, reflection information, and consultation information, the language model is used to generate the advice. The information processing method according to claim 1 or 2, wherein the computer performs the processing.

5. The employee's personality information and the employee's medium- to long-term career goals are obtained. Based on the acquired progress information, reflection information, consultation information, personality information, and medium- to long-term career goals, the language model is used to generate the advice. The information processing method according to claim 1 or 2, wherein the computer performs the processing.

6. The information of the conversation between the employee and the supervisor is obtained, Based on the acquired dialogue information, a language model is used to generate minutes of the dialogue information. Based on the aforementioned meeting minutes, a language model is used to generate behavioral information regarding the actions that the employee should take. Output the meeting minutes and the activity information. The information processing method according to claim 1 or 2, wherein the computer performs the processing.

7. Based on the employee's progress information, reflection information, and consultation information, and the minutes of the dialogue with the employee, a language model is used to generate reflection information on the dialogue between the employee and their supervisor. Output the generated review information. The information processing method according to claim 1 or 2, wherein the computer performs the processing.

8. The reflection information regarding the dialogue between the employee and the supervisor is output to the terminals of the employee and the supervisor's manager. The information processing method according to claim 7, wherein the processing is performed by the computer.

9. The system outputs a screen displaying information about past conversations between the employee and their supervisor. Through the aforementioned screen, the current progress information, reflection information, and consultation information of the employee are obtained. Based on the acquired progress information, reflection information, and consultation information, a language model is used. , generate advice for the employee or advice for the supervisor, A screen is generated that displays the information regarding past conversations, the current progress information, reflection information, and consultation information of the employee, and the generated advice in association with each other. Output the generated screen. The information processing method according to claim 1 or 2, wherein the computer performs the processing.

10. We obtain information on employees' progress toward their work goals, information on their work reflections, and information on consultations. Based on the acquired progress information, reflection information, and consultation information, a language model is used to generate advice regarding communication between the employee and their supervisor. Output the generated advice. A program that instructs a computer to perform a process.

11. In an information processing device having a control unit, The control unit, We obtain information on employees' progress toward their work goals, information on their work reflections, and information on consultations. Based on the acquired progress information, reflection information, and consultation information, a language model is used to generate advice regarding communication between the employee and their supervisor. Output the generated advice. Information processing device.

Images