Information processing system and information processing method

Abstract

Provides methods that contribute to effective organizational improvement. [Solution] The information processing system provides members of an organization with a questionnaire screen (23) for inputting evaluation values ​​for other groups, and if the evaluation value for one other group is low, conducts an interview using the function of a language model (4). The information processing system collects the input results into a specific user interface and the interview results and stores them in a server storage unit (12). Based on the input results of multiple members, the information processing system extracts pairs requiring improvement, which are combinations of groups whose relationship requires improvement, and uses the function of the language model to generate improvement-related information related to improving the relationship between the pairs requiring improvement based on the interview results related to the pairs requiring improvement, and provides the improvement-related information to parties involved in the pairs requiring improvement.

Description

[Technical Field]

[0001] The present disclosure relates to an information processing system for improving an organization and an information processing method by the information processing system. [Background technology]

[0002] Conventionally, technologies for supporting organizational improvement have been proposed for companies, associations, and other organizations. For example, Patent Document 1 describes the following technology. Specifically, the organizational improvement activity support device described in Patent Document 1 acquires a measure suitability level, which quantitatively indicates the degree of suitability for the characteristics and improvement theme of the target organization, a measure agreement level, which quantitatively indicates the degree of support for the measure among members of the target organization, and a measure activity level, which quantitatively indicates the degree of implementation of the measure in the target organization. Based on these, the device calculates an effect point, which quantitatively indicates the degree of effectiveness of organizational improvement activities in the target organization, and presents information on at least the activity status of the target organization to facilitators and / or members of the target organization based on the effect point. The technology described in Patent Document 1 makes it possible to improve the continuity of organizational improvement activities without the need for on-site expert intervention. Patent Document 2 also describes the following technology. Specifically, an information processing device 3 stores organizational status information regarding the status of the organization and presentation information, which is information to be presented to members of the organization. The information processing device 3 uses the organizational status information to determine and transmit the presentation information to be presented to members of the organization from the presentation information. According to the technology of Patent Document 2, it is possible to use organizational status information to appropriately present information to members of an organization in order to improve the organization. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2016-139323 [Patent Document 2] Japanese Patent Publication No. 2023-171503 Summary of the Invention [Problem to be solved by the invention]

[0004] As suggested by Patent Documents 1 and 2, in the field of organizational improvement, there is a constant demand for methods that contribute to effective organizational improvement.

[0005] The present disclosure has been made to solve such problems, and aims to provide a method that contributes to effective organizational improvement. [Means for solving the problem]

[0006] In order to solve the above-mentioned problems, the present disclosure has the following configuration: An information processing system comprising: a function of providing a specific user interface for inputting evaluation values ​​to members of an organization having multiple groups for other groups different from the group to which the member belongs, a function of using a language model to conduct a specific interview regarding one of the other groups for members who have input an evaluation value for the other group into the specific user interface, a function of collecting input evaluation values ​​that are evaluation values ​​input into the specific user interface and interview results of the specific interview for multiple members and storing them in a storage unit, a function of deriving statistics based on the input evaluation values ​​of the multiple members stored in the storage unit for each evaluation pair that is a combination of an evaluator group and an evaluee group in the specific interview, and a function of providing the statistics to a manager, a function of using the language model to analyze interview responses given by members in the specific interview corresponding to the evaluation pair based on the interview results of the multiple members stored in the storage unit, and deriving analysis results, and a function of providing the analysis results to the manager. [Effects of the Invention]

[0007] According to the present disclosure configured as described above, statistics for each evaluation pair are provided to the manager based on the evaluation values ​​set by members of an organization for other groups. Therefore, by referring to the statistics, the manager can accurately identify group combinations with which improved relationships are required. Furthermore, according to the present disclosure, interviews are conducted with members regarding the other groups, and analysis results based on the interview results are provided to the manager. Therefore, the manager can obtain useful information regarding organizational improvement for the evaluation pairs that reflects the responses to the interviews. In other words, according to the present disclosure, a method can be provided that contributes to effective organizational improvement. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a diagram illustrating an example of the configuration of a control system according to a first embodiment. [Figure 2] 2 is a block diagram showing an example of the functional configuration of an information processing server, an employee terminal, and a manager terminal according to the first embodiment; FIG. [Figure 3] FIG. 10 is a diagram showing an example of a questionnaire screen according to the first embodiment. [Figure 4] 10 is a flowchart showing an information processing method by an employee terminal and an information processing server. [Figure 5] 10 is a flowchart illustrating an information processing method by an employee terminal and a language model server. [Figure 6] FIG. 10 is a diagram showing an example of a questionnaire screen according to the first embodiment. [Figure 7] FIG. 10 is a diagram showing an example of the contents of survey result information. [Figure 8] FIG. 10 is a diagram showing an example of the contents of evaluation value related information. [Figure 9] FIG. 10 is a diagram used to explain processing by an information processing server. [Figure 10] FIG. 10 is a diagram used to explain processing by an information processing server. [Figure 11] FIG. 10 is a diagram illustrating an example of a department relationship screen. [Figure 12] FIG. 10 is a diagram illustrating an example of a details screen. [Figure 13] FIG. 10 is a diagram illustrating an example of an improvement list screen. [Figure 14] FIG. 11 is a diagram showing an example of a questionnaire screen according to the second embodiment. [Figure 15] 10 is a flowchart showing an information processing method by an employee terminal and an information processing server. [Figure 16] 10 is a flowchart illustrating an information processing method by an employee terminal and a language model server. [Figure 17] FIG. 10 is a diagram showing the contents of individual question answer information. [Figure 18] FIG. 10 is a diagram showing an example of an average evaluation value for each target pair in the form of a heat map. [Figure 19] FIG. 10 is a diagram illustrating an example of an improvement-related screen. [Figure 20] FIG. 1 illustrates an example of the configuration of an information processing system. DETAILED DESCRIPTION OF THE INVENTION

[0009] First Embodiment A first embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing an example of the configuration of a control system 1 (information processing system) according to this embodiment. As shown in FIG. 1, the control system 1 includes an information processing server 2 (information processing system, server, computer), an employee terminal 3E (information processing system, terminal, computer), and a manager terminal 3R (information processing system, terminal, computer). The information processing server 2, employee terminal 3E, and manager terminal 3R can all be connected to a network N including the Internet, a telephone network, and other communication networks. A language model server 4 is connected to the network N.

[0010] The information processing server 2 has a function of providing a service that supports improving relationships between groups within an organization. Hereinafter, the service provided by the information processing server 2 will be referred to as the "service." In this embodiment, a single company (organization) with multiple departments (groups) is assumed, and this company will be referred to as the "target company." The target company (a concept that also includes individuals belonging to the target company) will use the service. Examples of departments within the target company include the legal department, human resources department, corporate planning department, sales department, or development department. In this embodiment, "manager" refers to a department manager, a member of the management / management level, or other person involved in the management / administration of the target company. However, "manager" does not refer only to a person in a specific position, but also broadly includes anyone who can access information based on the results of a specific survey (described below). In this embodiment, "employee" refers to a person who belongs to a department but is not a manager. However, "employee" does not refer only to a person in a specific position, but also broadly includes anyone who can take a specific survey. Hereinafter, employees and managers will be collectively referred to as "users." Please note that the user of the Service is described as a single company for the sake of simplicity, and it is natural that multiple organizations can use the Service.

[0011] In FIG. 1 and FIG. 2 described below, the information processing server 2 is represented by a single block. However, this does not mean that the information processing server 2 is composed of a single device. The information processing server 2 may be composed of a virtual server. The information processing server 2 may also be composed of a plurality of devices. In this configuration, the devices constituting the information processing server 2 may include a web server, a web application server, or a database server. In this configuration, the information processing server 2 may also be composed of a plurality of server devices whose load is distributed by a load balancer. When the information processing server 2 is composed of a plurality of devices, each of the information processing units of the plurality of devices, or a combination of the plurality of information processing units, functions as an "information processing unit."

[0012] The employee terminal 3E is a terminal used by an employee. However, the employee terminal 3E is a convenient reference to a terminal used by an employee, and does not mean a terminal used exclusively by an employee. The employee terminal 3E may be of any type. For example, a tablet computer (including so-called smartphones), a desktop computer, or a notebook computer can function as the employee terminal 3E. The administrator terminal 3R is a terminal used by an administrator. However, the administrator terminal 3R is a convenient reference to a terminal used by an administrator, and does not mean a terminal used exclusively by an administrator. The administrator terminal 3R may be of any type.

[0013] The language model server 4 is a server equipped with a language model 5. The language model 5 according to this embodiment is a so-called large-scale language model (LLM). The provider of the language model 5, the type of the language model 5, and the machine learning method of the language model 5 are not limited. The language model 5 may be a publicly available non-commercial model or a publicly available commercial model (e.g., GPT-4, GPT-4o, or BERT). The language model 5 may also be a local LLM. The language model 5 may be fine-tuned or additionally trained by other methods from the perspective of supporting organizational improvement, being used in the present service, or other perspectives. The language model server 4 has a function of receiving a prompt (question sentence) from an external device, generating a response sentence to the prompt using the function of the language model 5, and transmitting the response sentence to the external device. The information processing server 2 is appropriately authorized to use the language model server 4.

[0014] 2 is a block diagram showing an example of the functional configuration of the information processing server 2, employee terminal 3E, and manager terminal 3R. As shown in FIG. 2, the information processing server 2 includes, as its functional configuration, a server information processing unit 10 (information processing unit), a server communication unit 11, and a server storage unit 12 (storage unit). The employee terminal 3E includes, as its functional configuration, a terminal information processing unit 13E (information processing unit), a terminal communication unit 14E, a terminal display unit 15E (display unit), a terminal input unit 16E, a voice input unit 17E, a voice output unit 18E, and a terminal storage unit 19E. The manager terminal 3R includes, as its functional configuration, a terminal information processing unit 13R (information processing unit), a terminal communication unit 14R, a terminal display unit 15R (display unit), a terminal input unit 16R, a voice input unit 17R, a voice output unit 18R, and a terminal storage unit 19R.

[0015] The server information processing unit 10 comprises a processing device including a processor and a primary storage device. The server information processing unit 10 executes processing by having the processing device read out a program stored in the storage area of ​​the server storage unit 12 (or another storage area) into the primary storage device and execute it. In other words, the server information processing unit 10 executes processing through cooperation between hardware and software. The server communication unit 11 comprises a communication device including a communication control device and a network interface. The server communication unit 11 communicates with external devices via the communication device under the control of the server information processing unit 10. Hereinafter, it is assumed that communication by the information processing server 2 is appropriately executed by the server communication unit 11, and a description of the communication will be omitted. The server storage unit 12 stores data in non-volatile memory. The non-volatile memory is, for example, a hard disk drive (or other magnetic storage device), ROM, or flash memory.

[0016] The server storage unit 12 stores a user management database 20. The user management database 20 stores information about each user (including employees and managers) in association with a user ID (identification information).

[0017] The terminal information processing unit 13E of the employee terminal 3E includes a processing device including a processor and a primary storage device. The terminal information processing unit 13E executes processing by having the processing device read a program stored in the storage area of ​​the terminal storage unit 19E (or another storage area) into the primary storage device and execute it. In other words, the terminal information processing unit 13E executes processing through cooperation between hardware and software. The terminal communication unit 14E includes a communication device including a communication control device and a network interface. The terminal communication unit 14E communicates with external devices via the communication device under the control of the terminal information processing unit 13E. The terminal display unit 15E includes a liquid crystal panel, an organic EL panel, or other display device. The terminal display unit 15E displays images on the display device under the control of the terminal information processing unit 13E. The terminal input unit 16E detects input to the input device and outputs the detection result to the terminal information processing unit 13E. The input device is, for example, a keyboard, a mouse, a touch panel, or a camera. The voice input unit 17E includes a microphone and a signal processing circuit. The audio input unit 17E has a function of outputting an audio signal based on audio picked up by the microphone to the terminal information processing unit 13E. The terminal information processing unit 13E has a function of performing analog / digital conversion processing including sampling, quantization, and encoding processing, and other necessary signal processing on the audio signal input from the audio input unit 17E to generate audio data, and stores the audio data in the terminal storage unit 19E. The audio output unit 18E has a speaker and a signal processing circuit. The audio output unit 18E inputs an audio signal from the terminal information processing unit 13E, drives the speaker based on the input audio signal, and causes the speaker to emit audio based on the audio signal. The terminal storage unit 19E stores data in a non-volatile memory.

[0018] The terminal information processing unit 13R, terminal communication unit 14R, terminal display unit 15R, terminal input unit 16R, audio input unit 17R, audio output unit 18R, and terminal storage unit 19R of the manager terminal 3R each have the same hardware configuration and the same "functions based on the hardware configuration" as the terminal information processing unit 13E, terminal communication unit 14E, terminal display unit 15E, terminal input unit 16E, audio input unit 17E, audio output unit 18E, and terminal storage unit 19E of the employee terminal 3E. In the following, it is assumed that communication between the employee terminal 3E and the manager terminal 3R is appropriately performed by the terminal communication units 14E and 14R, and a description of the communication will be omitted.

[0019] As shown in FIG. 2, a dedicated application AP (hereinafter referred to as the "dedicated application AP") is installed on the employee terminal 3E and the manager terminal 3R. The dedicated application AP is dedicated software that implements functions such as providing various screens related to the service, transmitting and receiving various information related to the service with the information processing server 2 and other external devices, and executing other processes related to the service. Unless otherwise specified in the following description, the terminal information processing units 13E and 13R basically execute processes using the functions of the dedicated application AP (including the terminal's OS, web applications that the dedicated application AP can use, and other programs that can cooperate with the dedicated application AP). Furthermore, even if not otherwise specified in the following description, the terminal information processing units 13E and 13R execute the following processes as needed. That is, the terminal information processing units 13E and 13R communicate with the server information processing unit 10 (which may be an information processing unit of another external device; the same applies below) as needed to obtain information required for processing. In particular, when displaying various screens, the terminal information processing units 13E and 13R obtain information required for displaying the screens from the server information processing unit 10. The server information processing unit 10, user management database 20, survey event management database 21 (described later), and other databases / data are referenced, and necessary information is transmitted to the terminal information processing units 13E and 13R. The terminal information processing units 13E and 13R also communicate with the server information processing unit 10 as necessary to request the execution of processing and obtain the processing results. In the following, the display of various screens on the terminal display units 15E and 15R by the terminal information processing units 13E and 13R using the functions of the dedicated application AP and other programs may be expressed simply as "the terminal information processing units 13E and 13R display a screen."

[0020] Next, the operation of the control system 1 will be described. In the following description, it is assumed that the target company has eight departments: Department A, Department B, Department C, ... Department H. The target company periodically holds an investigation event. During this investigation event, a specific questionnaire (described below) is administered to employees, and the results of the specific questionnaire are used to perform various analyses and provide various information to the manager. The server storage unit 12 of the information processing server 2 stores an investigation event management database 21 for managing various information related to each investigation event. In the following description, when focusing on a certain employee, the department to which the employee belongs is referred to as the "own department" (own group), and departments other than the department to which the employee belongs are referred to as the "other departments" (other groups). In the following, one investigation event will be referred to as the "current investigation event," and the operation of the control system 1 during the current investigation event will be described in detail by scene. In the following, the database of the investigation event management database 21 corresponding to the current investigation event will be specifically referred to as the "current investigation event management database 21-1."

[0021] <Scene of an employee responding to a specific survey> First, one of the employees will be referred to as a "notable employee," and the operation of the control system 1 in a scene where the notable employee answers a specific questionnaire will be explained. The notable employee belongs to Department B. The significance of the specific questionnaire will become clear later.

[0022] When the survey event is held, each employee of the target company is notified in a predetermined manner that they must answer a specific questionnaire. In response to this notification, the target employee launches the dedicated application AP on their employee terminal 3E and logs in to the service. When logging in, the server information processing unit 10 performs authentication in cooperation with the terminal information processing unit 13E. The server information processing unit 10 acquires the target employee's user ID during authentication. The server information processing unit 10 can acquire information about the target employee from the user management database 20 using the target employee's user ID. Even if not specifically described below, the server information processing unit 10 acquires information about the target employee from the user management database 20 as needed. The above points also apply when an administrator logs in to the service. In other words, a state is created in which the server information processing unit 10 can acquire information about the administrator from the user management database 20 in response to the administrator's login.

[0023] After logging in, the target employee performs a predetermined operation to instruct the display of a questionnaire screen 23 (specific user interface). In response to this instruction, the terminal information processing unit 13E displays the questionnaire screen 23 on the terminal display unit 15E. FIG. 3A shows an example of the questionnaire screen 23 in its initial state. The series of questionnaires implemented by the questionnaire screen 23 corresponds to a "specific questionnaire." As shown in FIG. 3A, a character image 24 representing an interviewer is displayed on the left side of the questionnaire screen 23. The questionnaire screen 23 also has an other department selection field 25 for selecting one of the other departments with which the employee completing the questionnaire collaborates or frequently works. The questionnaire will be conducted for the other department selected in the other department selection field 25. Hereinafter, the other department with which the employee collaborates or frequently works will be referred to as the "other collaborating department."

[0024] When focusing on this questionnaire, the employee who completes this questionnaire is the one who evaluates the "other department selected in the other department selection field 25," and the "other department selected in the other department selection field 25" is the department being evaluated by the employee. With this in mind, hereinafter, the employee who completes this questionnaire will be referred to as the "evaluator," the employee's own department will be referred to as the "evaluating department," and the "other department selected in the other department selection field 25" will be referred to as the "evaluated department." The evaluating department corresponds to the "group evaluating in a specific interview," and the evaluated department corresponds to the "group being evaluated in a specific interview." If there are multiple collaborating departments, the employee will complete a specific questionnaire for each of the other departments.

[0025] The employee of interest selects one of the other departments that he or she collaborates with as the department to be evaluated from the other department selection field 25. When the department to be evaluated is selected, the terminal information processing unit 13E displays an evaluation value input field 26 below the other department selection field 25. (B) of FIG. 3 shows the evaluation value input field 26 displayed on the questionnaire screen 23. The evaluation value input field 26 is an area for inputting the evaluation value to be set for the department to be evaluated.

[0026] The evaluation value is a score that can range from 1 to 10. An employee's evaluation value for a department functions as an index showing the degree of positive or negative impression the employee has of that department. The evaluation value increases the more positive (or less negative) the employee's impression of the department, and decreases the more negative (or less positive) the employee's impression of the department. If an employee has thoughts, feelings, or experiences regarding another department, such as having a good relationship with that department on a regular basis or finding collaboration with that department helpful, the employee is likely to have a positive impression of that department. In this case, the employee is likely to set a high evaluation value for that department. On the other hand, if an employee has thoughts, feelings, or experiences regarding a department, such as having a strained relationship with that department or having previously failed collaboration with that department, the employee is likely to have a negative impression of that department. In this case, the employee is likely to set a low evaluation value for that department.

[0027] The evaluation value input field 26 and the questionnaire screen 23 including the evaluation value input field 26 correspond to a "specific user interface for inputting an evaluation value to be set for a group other than the group to which the member belongs, for a member of an organization having multiple groups. The terminal information processing unit 13E has a function to provide such a specific user interface.

[0028] The notable employee inputs an evaluation value to be set for the department being evaluated in the evaluation value input field 26 based on his or her own thoughts, feelings, and experiences. Hereinafter, the evaluation value input in the evaluation value input field 26 will be referred to as the "input evaluation value." When the notable employee inputs an evaluation value, the control system 1 executes the following process. Below, an example of the operation of the control system 1 when an evaluation value is input will be explained using a flowchart. Figure 4 is a flowchart showing an example of the operation of the control system 1 when an input evaluation value is input. Flowchart FA in Figure 4 shows the information processing method by the employee terminal 3E, and flowchart FB shows the information processing method by the information processing server 2.

[0029] As shown in flowchart FA, when the evaluation value is input by the target employee (step SA1), the terminal information processing unit 13E executes interview processing (step SA2). Interview processing is processing in which the terminal information processing unit 13E uses the functions of the language model 5 to conduct an interview with an employee (the target employee in this example). Hereinafter, an interview realized by interview processing will be referred to as a "specific interview." A specific interview is conducted with the purpose of obtaining answers to at least the following questions: ·The appraiser's positive opinion of the department being evaluated - "Issues of the department being evaluated" that the evaluator considers in relation to himself or her own department. - The evaluator's "improvement proposals to solve / improve the issues of the department being evaluated" Hereinafter, the issues of the department being evaluated will be referred to as "the issues of the department being evaluated." The issues of the department being evaluated are equivalent to "other group issues." Furthermore, the "improvement plan to solve / improve the issues being evaluated" proposed by the evaluator will be referred to as the "evaluator improvement plan," which is equivalent to the "member improvement plan."

[0030] The questions to be answered in the specific interview are not limited to those given above. Other examples of such questions are given below. · Negative opinions that the evaluator has about the department being evaluated ·The evaluator's thoughts on "the causes of the issues in the department being evaluated" -Anecdotes about experiences working in collaboration with the department being evaluated

[0031] FIG. 5 is a diagram showing details of interview processing by the terminal information processing unit 13E together with processing by the language model server 4. In FIG. 5, flowchart FC shows the information processing method by the employee terminal 3E, and flowchart FD shows the information processing method by the language model server 4. As shown in flowchart FA, the terminal information processing unit 13E generates an initial prompt (step SC1). The initial prompt is a prompt that requests the language model server 4 (language model 5) to generate the first question (hereinafter referred to as the "initial question") to be presented to the evaluator (the target employee in this example) in the specific interview. The initial prompt describes the following items: The purpose of the specific interview and information requesting the rater to conduct the interview that reflects the purpose of the specific interview. Here, "conducting an interview" includes the concept of "generating questions to be provided to the rater in the interview." Attribute information of the evaluator, the evaluating department, and the evaluated department, as well as information requesting that the interview be conducted in a way that reflects these attribute information

[0032] In this embodiment, "attribute information" refers to information including attribute values ​​of various attributes. Employee attribute information includes, for example, name, gender, age, years of employment, career history before joining the company, and career history after joining the company (these are merely examples). Department attribute information includes, for example, department name, job description, and number of employees (these are merely examples). The server information processing unit 10 obtains information to be written in the initial prompt from the user management database 20, the survey event management database 21, and other databases / data as needed, and provides it to the terminal information processing unit 13E. This also applies to other prompts. The terminal information processing unit 13E (including other information processing units) transmitting a prompt containing a certain piece of information to the language model server 4 (language model 5) is equivalent to providing the certain piece of information to the language model server 4 (language model 5).

[0033] An example of an initial prompt is shown below. However, the purpose of this example is to provide a simplified overview of the initial prompt, and the content and structure have been abstracted and simplified. Furthermore, this example does not consider the appropriateness of the response.

[0034] ***Beginning of initial prompt*** You are a professional interviewer. The following <employee>, who belongs to the following <own department> of the following <company>, has given the following <input evaluation value> as an evaluation value for the following <other department>. The evaluation value is a score that comprehensively evaluates the other departments and ranges from 1 to 10 points. <Company> Company name: ○ Co., Ltd. Industry: System Integrator ... <My department> Department Name: Sales Department ·Number of members: 30 people ... <Employee> Name: ○Yama○o Age: 45 ... <Other departments> Department name: System Development Department ·Number of members: 50 people ... <Input evaluation value> 3 points You are planning to interview this employee with the goal of obtaining answers to at least the following questions: <Matters> (1) <Employee>'s positive opinion of <Other departments>. (2) "Issues in other departments" that employees consider in relation to themselves or their own department. (3) <Employee>'s improvement proposals for solving / improving issues in <other departments>. First, create and answer the first questions you should ask this employee. ***End of initial prompt***

[0035] Note that the initial prompt may be configured to include information other than the example information as a context. Examples of items that may be included as a context include the following: - Results of specific surveys conducted by the evaluator in the past -Information accumulated within the target company regarding the relationship between the evaluator / evaluation department and the department being evaluated

[0036] The terminal information processing unit 13E generates the initial prompt, for example, in the following manner. A template for the initial prompt is prepared in advance, and the terminal information processing unit 13E processes the template based on the information to be written in the initial prompt to generate the initial prompt. The terminal information processing unit 13E requests a predetermined language model server to generate an initial prompt, and obtains (generates) the initial prompt as a response sentence to the request. However, the method for generating the initial prompt (as well as other prompts) is not limited to the example method.

[0037] After step SC1, the terminal information processing unit 13E transmits the generated initial prompt to the language model server 4 (step SC2). As shown in flowchart FD, the language model server 4 receives the initial prompt and generates a response sentence based on the initial prompt using the language model 5 (step SD1). The response sentence includes an "initial question" that reflects the purpose of the specific interview and other matters described in the initial prompt. The language model server 4 transmits the generated response sentence to the terminal information processing unit 13E (step SD2).

[0038] The terminal information processing unit 13E receives and acquires the response sentence (step SC3). Note that "the terminal information processing unit 13E (and other information processing units) receiving and acquiring the response sentence from the language model server 4" corresponds to generating information described in the response sentence. Next, the terminal information processing unit 13E controls the voice output unit 18E to voice output the wording indicated by the first question based on the first question in the response sentence (step SC4). Note that the terminal information processing unit 13E voice-outputs the first question using voice synthesis technology or other technology. At this time, the terminal information processing unit 13E creates an effect in which the character image 24 appears to be speaking in sync with the voice output of the question. Along with the voice output of the first question, the terminal information processing unit 13E displays a question content field 27 displaying the content of the first question below the evaluation value input field 26 (step SC5). Figure 6A shows the question content field 27 displayed on the questionnaire screen 23. Hereinafter, questions provided to the evaluator in a specific interview will be referred to as "interview questions."

[0039] The target employee, having heard the audio output of the initial question, orally responds to the initial question (step SC6). Hereinafter, the response given by the employee to the interview question in the specific interview is referred to as the "interview response." The spoken voice is collected by the microphone of the audio input unit 17E, and the audio data is stored in the device storage unit 19E. The device information processing unit 13E converts the audio data into text data (hereinafter referred to as "interview response data") containing text (wording) corresponding to the interview response (step SC7). Next, the device information processing unit 13E displays an answer content field 28 showing the content of the interview response below the question content field 27 based on the interview response data (step SC8). (B) of FIG. 6 shows the answer content field 28 displayed on the questionnaire screen 23. Next, the terminal information processing unit 13E writes the interview question based on the response sentence acquired in step SC3 and the interview question based on the interview response data generated in step SC7 in the interview log according to the format (step SC9). An interview log is log data in which interview questions and answers exchanged in a specific interview are written in chronological order along with accompanying information (for example, date and time).

[0040] Next, the terminal information processing unit 13E generates a prompt requesting the generation of a second question that reflects at least the interview answer (step SC10). The prompt describes the interview answer (i.e., the answer given by the target employee to the first question) and includes information requesting the generation of a second question that further reflects the answer. Furthermore, prompts for the second and subsequent questions include information requesting that, if the question recorded in the response sentence is the "final question" in the specific interview, information to that effect be included in the response sentence. After generating the prompt, the terminal information processing unit 13E sends the generated prompt to the language model server 4 (step SC11). Thereafter, the second, third, ..., and final interview questions are provided to the target employee in the same manner as the first question, and the target employee provides interview answers to the second, third, ..., and final interview questions. The terminal information processing unit 13E writes information in the interview log each time the target employee provides an interview answer to an interview question.

[0041] For example, in a specific interview, the following dialogue takes place between the employee of interest and the AI ​​interviewer (a convenient way of describing the entity that interviews the employee of interest): The following example is merely a simplified example. [AI Interviewer] Thank you. A score of 7 means you have a high opinion of Department A. Please tell us any positive comments you have about Department A. [Featured Employee] The other day, when I went to consult with them, they were very helpful. [AI Interviewer] I see. Is there anything else? [Featured Employee] Speaking of which... [AI Interviewer] So, what do you think are the issues facing Department A? [Featured Employee] I think they're always slow to respond. [AI Interviewer] What do you think is the reason for Department A's slow response? [Featured Employee] I think it's simply a lack of staff. [AI Interviewer] So, how do you think this can be resolved? [Featured Employee] In my opinion... [AI Interviewer] This is my last question. What ideas do you have for improving the issues you just mentioned? [Employee of Interest] I think it would be good to have an environment where we could check in with each other about how busy we are. [AI Interviewer] This concludes the interview. Thank you for your hard work.

[0042] As mentioned above, the targeted interview is conducted in a conversational format, which reduces the burden on the employee of interest. Also, because the interviewer is not human, the employee can respond to the targeted interview without feeling nervous, and it is easy for them to speak their true feelings without worrying about their personal relationship with the interviewer. Furthermore, because the interviewer is not human, anonymity can be ensured for the employee's comments, and this anonymity makes it easier for employees to speak their true feelings.

[0043] As described above, the terminal information processing unit 13E has a function of conducting a specific interview with the evaluator using the functions of the language model 5. The terminal information processing unit 13E also has a function of providing employee attribute information to the language model 5 and conducting a specific interview by reflecting the employee attribute information. The terminal information processing unit 13E also has a function of providing attribute information of the evaluating department (own group) and attribute information of the evaluated department (other group) to the language model 5 and conducting a specific interview by reflecting the attribute information of the own department and the attribute information of the other department. The above is the details of the interview processing.

[0044] As shown in the flowchart FA of FIG. 4, once the interview process in step SA2 is complete, the terminal information processing unit 13E executes the interview log analysis process (step SA3). The interview log analysis process analyzes the interview log, extracts positive opinion responses from the interview responses given by the evaluator in a specific interview, and further extracts combinations of issue-related responses and improvement-proposal-related responses for each issue in the department being evaluated. Positive opinion responses are interview responses that express positive opinions about the department being evaluated (responses that mention the opinions). Issue-related responses are interview responses that relate to the issues in the department being evaluated (responses that mention the issues). Improvement-proposal-related responses are interview responses that relate to the evaluator's improvement proposal for the issues in the department being evaluated (responses that mention the improvement proposals). Hereinafter, a combination of an issue-related response and an improvement-proposal-related response corresponding to one issue in the department being evaluated is referred to as a "issue-improvement-proposal pair." In the interview log analysis process, either no positive opinion responses or one or more positive opinion responses may be extracted. Similarly, there are cases where no problem-improvement-proposal pairs are extracted, and cases where one or more pairs are extracted.

[0045] The terminal information processing unit 13E executes the extraction process in the following manner. That is, the terminal information processing unit 13E generates a prompt that includes an interview log and requests the extraction of positive opinion response and problem improvement proposal pairs from the interview log. The prompt also includes information requesting that the extraction be performed taking into consideration the chronological order of interview questions and interview answers in a specific interview, the contents of the interview questions and interview answers, and other factors. The terminal information processing unit 13E transmits the prompt to the language model server 4 and obtains a response sentence from the server. The terminal information processing unit 13E obtains (extracts) positive opinion response and problem improvement proposal pairs from the response sentence.

[0046] However, the method for extracting pairs of positive opinion responses and problem improvement proposals is not limited to the exemplified method. Furthermore, the entity that performs the extraction and the timing of the extraction are also not limited to the exemplified method. The above is the details of step SA3.

[0047] After step SA3, the server information processing unit 10 generates survey result information (step SA4). FIG. 7 is a diagram showing the contents of the survey result information. As shown in FIG. 7, the survey result information includes an evaluating department ID, which is the department ID of the evaluating department, an evaluated department ID, which is the department ID of the evaluated department, the user ID of the noted employee, an input evaluation value entered by the noted employee, an interview log for the specific interview, and log analysis result information. The log analysis result information includes each positive opinion response and each problem / improvement proposal pair extracted in the interview log analysis process of step SA3. The interview log and log analysis result information correspond to the interview results of the specific interview. Based on this, the survey result information includes the input evaluation value and the interview result. In the log analysis result information, type information indicating that the response is a positive opinion response is associated with each positive opinion response, and type information indicating that the pair is a problem / improvement proposal pair is associated with each problem / improvement proposal pair. Furthermore, in a problem-improvement pair, type information indicating that the answer is a problem-related answer is associated with the problem-related answer, and type information indicating that the answer is a problem-related answer is associated with the improvement suggestion-related answer.

[0048] After step SA2, the terminal information processing unit 13E transmits the survey result information to the server information processing unit 10 (step SA5). As shown in flowchart FB, the server information processing unit 10 receives and acquires the survey result information (step SB1). Next, the server information processing unit 10 records the received survey result information in the current survey event management database 21-1 in an appropriate manner (step SB2).

[0049] In the survey event according to this embodiment, all employees are required to complete a specific questionnaire. Consequently, the server information processing unit 10 ultimately registers the survey result information for each collaborating department for all employees in the current survey event management database 21-1. Specifically, the server information processing unit 10 has the function of collecting and storing in the server storage unit 12 (storage unit) the input evaluation values, which are the evaluation values ​​entered into the survey screen 23 (specific user interface), and the interview results of the specific interview for multiple employees (members). While it is possible that some employees may not complete the specific questionnaire due to certain circumstances, for the sake of simplicity, in this embodiment, it is assumed that all employees completed the specific questionnaire. This concludes the operation of the control system 1 in the "scene in which employees complete a specific questionnaire."

[0050] <Scene for generating information to be provided> After all employees have completed the specific questionnaire, the server information processing unit 10 generates evaluation value-related information, positive opinion-related information, and assignment-related information for each evaluation pair. Hereinafter, evaluation value-related information, positive opinion-related information, and assignment-related information are collectively referred to as "information to be provided." An evaluation pair is a combination of an evaluating department and an evaluated department. In this example, the target company has eight departments, so there are 56 types of evaluation pairs. Therefore, in this example, the server information processing unit 10 generates evaluation value-related information, positive opinion-related information, and assignment-related information for each of the 56 types of evaluation pairs. Below, evaluation pairs are expressed in the format of "evaluating department → evaluated department," such as "Department A → Department B," "Department B → Department A," "Department C → Department D," etc. The operation of the server information processing unit 10 when generating each piece of information is described below.

[0051] ●Evaluation value related information First, the server information processing unit 10 will be described when generating evaluation value-related information for a given evaluation pair. In the following description of evaluation value-related information, the evaluation pair for which evaluation value-related information is generated will be referred to as the "processing target pair." FIG. 8 is a diagram showing the contents of the evaluation value-related information. As shown in FIG. 8, the evaluation value-related information includes average value-related information. The average value-related information includes range averages for each evaluation value range. As described above, in this embodiment, the lower limit of the evaluation value is 1 point, and the upper limit is 10 points. The evaluation value range refers to the range that can exist between 1 point and 10 points. The evaluation value range can also be described as the combination of different scores that can be obtained. For example, "1 point - 2 points," "1 point - 3 points"... "2 points - 3 points," "2 points - 4 points"... "8 points - 9 points" are each evaluation value ranges.

[0052] The average value of a range of evaluation values ​​for a given evaluation pair refers to the average value based on the input evaluation values ​​corresponding to that evaluation pair that fall within that range. For example, the average value of the range "1 point - 3 points" for "Department A → Department B" is the average value based on the 1 point, 2 point, and 3 point input evaluation values ​​set by employees of Department A (the evaluating department) for Department B (the evaluated department). In this example, if five employees gave a score of 1, one employee gave a score of 2, and two employees gave a score of 3, the range evaluation value is "1.6 points" because "(1 point x 5 employees + 2 points x 1 employee + 3 points x 2 employees) / 8 = 1.6" (rounded to the nearest tenth). Here, the average value of the range of evaluation values ​​"1 point - 10 points" is the average value of all input evaluation values ​​corresponding to the corresponding evaluation pair. Based on this, the average value of the range of evaluation values ​​"1 point - 10 points" is referred to as the "overall average."

[0053] The server information processing unit 10 derives the range average value for each evaluation value range in the following manner. That is, the server information processing unit 10 acquires each input evaluation value corresponding to the pair to be processed from the "survey result information corresponding to the pair to be processed" in the current survey event management database 21-1. Next, the server information processing unit 10 uses each input evaluation value to derive the range average value for each evaluation value range. The range average value corresponds to a "statistic." The server information processing unit 10 has a function to derive a statistical quantity based on the input evaluation value for each evaluation pair.

[0054] As shown in Fig. 8, the evaluation value-related information includes an overall average change amount. The overall average change amount is the amount of change in the overall average value for the current investigation event relative to the overall average value for the previous investigation event for the corresponding evaluation pair. The server information processing unit 10 derives the overall average change amount for the pair to be processed by referring to the current investigation event management database 21-1 and the investigation event management database 21 related to the previous investigation event.

[0055] After generating the evaluation value related information for the pair to be processed, the server information processing unit 10 associates the evaluation related value information with the identification information of the pair to be processed and registers the evaluation related value information in the current investigation event management database 21-1 in an appropriate manner. This concludes the description of the processing of the server information processing unit 10 when generating evaluation value related information.

[0056] ●Positive opinion related information Next, the server information processing unit 10 will be described when generating positive opinion related information for one evaluation pair. In the following description of positive opinion related information, the evaluation pair for which positive opinion related information is to be generated will be referred to as the "processing target pair."

[0057] FIG. 9 is a diagram used to explain the processing of the server information processing unit 10. First, the server information processing unit 10 refers to the current investigation event management database 21-1 and acquires each of the positive opinion replies corresponding to the pair to be processed (see symbol SX1 in FIG. 9). Next, the server information processing unit 10 groups each of the positive opinion replies based on the similarity of the content of the replies (symbol SX2). Hereinafter, the groups generated by grouping will be referred to as "positive opinion groups." The processing related to grouping will be described in detail below.

[0058] The server information processing unit 10 generates a prompt that includes each positive opinion response and requests that the positive opinion responses be grouped according to the similarity of the content of each response. The server information processing unit 10 then sends the prompt to the language model server 4 and obtains a response sentence. The server information processing unit 10 groups the positive opinion responses based on the content of the response sentence. The server information processing unit 10 assigns a positive opinion group ID, which serves as identification information, to each existing positive opinion group. The above is the details of the grouping process.

[0059] After grouping, the server information processing unit 10 generates a summary (hereinafter referred to as a "positive opinion summary") for each positive opinion group based on each of the positive opinion replies belonging to the group (symbol SX3). Hereinafter, one of the positive opinion groups will be referred to as a target group, and the process of generating a positive opinion summary for the target group will be described in detail. The server information processing unit 10 generates a prompt that includes each of the positive opinion replies belonging to the target group and requests that a summary be generated for each of the positive opinion replies. Next, the server information processing unit 10 sends the prompt to the language model server 4 and obtains a response. The server information processing unit 10 obtains (generates) a positive opinion summary from the response. The above is the detailed process of generating a positive opinion summary for the target group.

[0060] After generating the positive opinion summary for each positive opinion group, the server information processing unit 10 generates positive opinion related information including each positive opinion response and the positive opinion summary, in association with the positive opinion group ID for each positive opinion group (symbol SX4). After generating the positive opinion related information, the server information processing unit 10 associates the positive opinion related information with the identification information of the pair to be processed and registers the positive opinion related information in the current investigation event management database 21-1 in an appropriate manner.

[0061] Note that using the functions of the language model 5 to generate summaries (not limited to positive opinion summaries) based on interview responses corresponds to analyzing the interview responses. The resulting generated summaries correspond to analysis results. This also applies to the processing related to AI summaries, which will be described later, and to processing interview responses in some form using other language models 5. The above is an explanation of the processing of the server information processing unit 10 when generating positive opinion related information.

[0062] ● Issue-related information Next, the server information processing unit 10 will be described when generating assignment-related information for one evaluation pair. In the following description of assignment-related information, the evaluation pair for which assignment-related information is to be generated will be referred to as a "processing target pair."

[0063] FIG. 10 is a diagram used to explain the processing of the server information processing unit 10. First, the server information processing unit 10 refers to the current investigation event management database 21-1 and acquires each of the problem improvement plan pairs corresponding to the pair to be processed (see symbol SY1 in FIG. 10). Next, the server information processing unit 10 groups each of the problem improvement plan pairs based on the similarity of the content of the problem-related answers (symbol SY2). Hereinafter, the groups generated by grouping will be referred to as "problem groups." The processing related to grouping will be described in detail below.

[0064] The server information processing unit 10 generates a prompt that includes each problem improvement plan pair and requests that the problem improvement plan pairs be grouped based on the similarity in the content of the problem-related answers. The server information processing unit 10 then sends the prompt to the language model server 4 and receives a response sentence. The server information processing unit 10 groups the problem improvement plan pairs based on the content of the response sentence. The server information processing unit 10 assigns a problem group ID, which serves as identification information, to each existing problem group. The above is the details of the grouping process.

[0065] After grouping, the server information processing unit 10 generates a summary (hereinafter referred to as "task summary") for each task group based on each of the task-related answers belonging to the group (symbol SY3). Hereinafter, one of the task groups will be referred to as the focus group, and the process for generating a task summary for the focus group will be described in detail. The server information processing unit 10 generates a prompt that includes each of the task-related answers belonging to the focus group and requests that a summary be generated for each of the task-related answers. The server information processing unit 10 then sends the prompt to the language model server 4 and obtains a response. The server information processing unit 10 obtains (generates) a task summary from the response. The process for generating a task summary corresponds to the process of analyzing the "interview answers," and the task summary corresponds to the "analysis result." The above is a detailed description of the process for generating a task summary for the focus group.

[0066] After generating a task summary for each task group, the server information processing unit 10 executes the following process for each task group. That is, the server information processing unit 10 groups each of the improvement plan-related answers belonging to the task group based on the similarity of the answer content (symbol SY4). Hereinafter, the groups generated by grouping will be referred to as "improvement plan groups." The process related to grouping will be described in detail below.

[0067] The server information processing unit 10 generates a prompt that includes each of the improvement suggestion-related answers and requests that the improvement suggestion-related answers be grouped based on the similarity of the content of each answer. The server information processing unit 10 then sends the prompt to the language model server 4 and receives a response sentence. The server information processing unit 10 groups the improvement suggestion-related answers based on the content of the response sentence. The server information processing unit 10 assigns an improvement suggestion group ID, which serves as identification information, to each existing improvement suggestion group. The above is the details of the grouping process.

[0068] After grouping, the server information processing unit 10 generates a summary (hereinafter referred to as "improvement proposal summary") for each improvement proposal group based on each of the improvement proposal-related answers belonging to the group (symbol SY5). Hereinafter, one of the improvement proposal groups will be referred to as the "target group," and the process for generating the improvement proposal summary for the target group will be described in detail. The server information processing unit 10 generates a prompt that includes each of the improvement proposal-related answers belonging to the target group and requests that a summary be generated for each of the improvement proposal-related answers. The server information processing unit 10 then sends the prompt to the language model server 4 and obtains a response. The server information processing unit 10 obtains (generates) the improvement proposal summary from the response. The above is a detailed description of the process for generating the improvement proposal summary for the target group. The process for generating the improvement proposal summary corresponds to the process of analyzing the "interview answers," and the improvement proposal summary corresponds to the "analysis result." The above is a detailed description of the process for generating the improvement proposal summary for the target group.

[0069] Furthermore, the server information processing unit 10 uses the function of the language model 5 to generate, for each improvement plan group, an AI improvement plan that is an improvement plan for improving the assessed department's problem corresponding to the group, based on the problem-related answers and improvement plan-related answers corresponding to the group (symbol SY6). The server information processing unit 10 generates a prompt that includes each of the corresponding problem-related answers and each of the corresponding improvement plan-related answers and requests the generation of an improvement plan for improving the assessed department's problem, which is different from the plan indicated by the improvement plan-related answers. The server information processing unit 10 sends the prompt to the language model server 4 and obtains a response sentence from the server. The server information processing unit 10 obtains (generates) the AI ​​improvement plan from the response sentence. The process of generating the AI ​​improvement plan corresponds to the process of analyzing the "interview answers," and the AI ​​improvement plan corresponds to the "analysis result."

[0070] After generating improvement plan summaries and AI improvement plans for all improvement plan groups, the server information processing unit 10 generates assignment-related information (SY7). As shown in Figure 10, the assignment-related information includes, for each assignment group, assignment-related answers and assignment summaries associated with the assignment group ID. Furthermore, the assignment-related information includes, for each assignment group, the corresponding assignment summary, the corresponding AI improvement plan, and the corresponding improvement plan-related answers for each improvement plan group.

[0071] After generating the assignment-related information, the server information processing unit 10 associates the assignment-related information with the identification information of the pair to be processed and registers the assignment-related information in the current investigation event management database 21-1 in an appropriate manner. This concludes the description of the processing of the server information processing unit 10 when generating assignment-related information.

[0072] This concludes the description of the operation of the control system 1 in the scene where information to be provided is generated.

[0073] <Scene in which the administrator refers to the provided screen> Next, we will explain an example of the operation of the control system 1 in a scene where an administrator refers to a provided screen. A provided screen is a general term for various screens provided to an administrator based on information to be provided and other information. Below, we will explain an example of the operation of the control system 1, assuming that one of the administrators is the "noted administrator" and that this noted administrator performs various tasks. Even if not otherwise specified, when the terminal information processing unit 13R displays various screens, the server information processing unit 10 provides the terminal information processing unit 13R with information necessary for displaying the screen. In particular, the server information processing unit 10 can obtain information necessary for displaying the screen from the "evaluation value-related information, positive opinion-related information, and issue-related information" of each evaluation pair and provide it to the terminal information processing unit 13R.

[0074] The administrator launches the dedicated application AP on his / her administrator terminal 3R and logs in to the service. After logging in, the administrator performs a predetermined operation to instruct the display of the department relationship screen 30. In response to this instruction, the terminal information processing unit 13R cooperates with the server information processing unit 10 to display the department relationship screen 30 on the terminal display unit 15R. Figure 11 is a diagram showing an example of the department relationship screen 30 in its initial state. As shown in Figure 11, the department relationship screen 30 includes a department relationship heat map 31 (statistics table). The department relationship heat map 31 is a heat map that displays range average values ​​for each evaluation pair. The department relationship heat map 31 has a cell 32 for each evaluation pair, and the range average value of the evaluation pair corresponding to each cell 32 is displayed.

[0075] In the department relationship heat map 31, a caution mark 33 is displayed in a cell 32 corresponding to a caution pair. A caution pair is an evaluation pair with a low overall average value. Specifically, in this embodiment, the terminal information processing unit 13R (or another information processing unit) determines that an evaluation pair whose overall average value is equal to or less than a threshold value (in this example, "4 points") is a caution pair. The terminal information processing unit 13R compares the overall average value with the threshold value for each evaluation pair to identify a caution pair. Note that the threshold value may be set by the administrator, may be derived using a predetermined formula that uses information about the evaluation value, or may be set by other means. The function of the caution mark 33 allows a concerned administrator to accurately identify evaluation pairs with relatively low overall average values ​​and that require attention to improving relationships between departments by referring to the department relationship heat map 31.

[0076] The method for determining whether an evaluation pair corresponds to a suspicious pair is not limited to the method exemplified in this embodiment. For example, the terminal information processing unit 13R may be configured to determine that an evaluation pair is a suspicious pair when the current overall average value for a certain evaluation pair is lower than the previous overall average value by a certain amount or more. Furthermore, for example, the terminal information processing unit 13R may be configured to determine that an evaluation pair is a suspicious pair when the value for a certain evaluation pair is gradually lowering from the previous overall average value toward the current overall average value.

[0077] Above the department relationship heat map 31, an evaluation value range setting bar 34 is provided for setting the evaluation value range for the department relationship heat map 31. The default evaluation value range set using the evaluation value range setting bar 34 is "1 point - 10 points." In this state, the department relationship heat map 31 displays the overall average value for each evaluation pair. If the focused manager sets the evaluation value range to "5 points - 10 points" using the evaluation value range setting bar 34, the department relationship heat map 31 displays the range average value corresponding to "5 points - 10 points" for each evaluation pair. By arbitrarily changing the evaluation value range using the evaluation value range setting bar 34, the focused manager can accurately recognize the range average value for each evaluation value range. In addition, the focused manager can grasp the bias in the input evaluation value scores that cannot be detected by the overall average value alone.

[0078] An overall average change setting bar 35 is provided to the right of the evaluation value range setting bar 34. The overall average change setting bar 35 is configured to allow a value to be specified in the range of "-9 points" to "+9 points." The terminal information processing unit 13R changes the background of the cells 32 of the evaluation pairs with the overall average change value that matches the value specified in the overall average change setting bar 35 from the default white to yellow. As described above, the overall average change value is the change in the overall average value for the current survey event relative to the overall average value for the previous survey event. For example, assume that "+2 points" is specified in the overall average change setting bar 35. Then, among the cells 32 in the department relationship heat map 31, the background of the cell 32 corresponding to the evaluation pair with an overall average change value of "2 points" turns yellow, and the background of the other cells 32 turns white. The focused manager can accurately recognize the overall average change value for each evaluation pair by arbitrarily changing the overall average change value using the overall average change setting bar 35.

[0079] As described above, the server information processing unit 10 has a function to provide the administrator with range average values ​​(statistics). The server information processing unit 10 also has a function to provide the administrator with a department relationship heat map 31 (statistics table), which is a display showing the range average values ​​of each evaluation pair. The server information processing unit 10 also has a function to display a caution mark 33 (applying decoration indicating a low level of the overall average value) for evaluation pairs with a low level of the overall average value on the department relationship heat map 31. However, the form of decoration indicating a low level of the overall average value is not limited to the form exemplified. For example, a configuration such as thickening the border or changing the background color may also be used.

[0080] Each cell 32 in the department relationship heat map 31 is selectable. When the focused manager selects a cell 32, the terminal information processing unit 13R pops up a details screen 37 corresponding to the selected cell 32. Hereinafter, the selected cell 32 is referred to as the selected cell, and the evaluation pair corresponding to the selected cell is referred to as the "target evaluation pair." FIG. 12 is a diagram illustrating an example of the details screen 37. As shown in FIG. 12, the details screen 37 includes a block-shaped positive opinion block 38 for each positive opinion group corresponding to the target evaluation pair. In the example of FIG. 12, the topmost block of the three blocks corresponds to the positive opinion block 38. The positive opinion block 38 for a positive opinion group displays a summary of the positive opinions for that positive opinion group. By referring to each positive opinion block 38, the focused manager can accurately understand the content / tendency of the positive opinions expressed by the evaluators for the target evaluation pair.

[0081] 12, the positive opinion block 38 has a response list button 39. When the noted manager selects the response list button 39, the terminal information processing unit 13R pops up a screen displaying a list of the positive opinion responses belonging to the corresponding positive opinion group. By referring to the screen, the noted manager can accurately recognize the specific responses given by the evaluators.

[0082] As shown in FIG. 12, the details screen 37 provides a block-shaped task block 40 for each task group corresponding to the target evaluation pair. In the example of FIG. 12, of the three blocks, the bottom two blocks correspond to the positive opinion block 38. The task block 40 for one task group displays a task summary for that task group. By referring to each task block 40, the focused manager can accurately grasp the content / tendency of the tasks for the evaluated department that the evaluator has raised for the target evaluation pair.

[0083] As shown in Fig. 12, the task block 40 has an answer list button 41. When the focused manager selects the answer list button 41, the terminal information processing unit 13R pops up a screen that displays a list of each of the task-related answers that belong to the corresponding task group. By referring to this screen, the focused manager can accurately recognize what specific answers the evaluators have provided.

[0084] As shown in FIG. 12, the task block 40 has a dig deeper button 60. When the dig deeper button 60 is selected, the server information processing unit 10 pops up an answer selection screen displaying a list of task-related answers belonging to the corresponding task group, each of which can be selected. When the target manager selects one or more task-related answers from the list, a notification is sent to the employee terminal 3E of each employee who issued the task-related answer. When the employee responds to the notification, the terminal information processing unit 13E uses the language model 5 to conduct an interview with the employee to learn more about the tasks being evaluated. In the interview, the terminal information processing unit 13E sends a prompt to the language model server 4, stating the purpose of "asking the employee more details about the tasks being evaluated," and uses the language model 5 to conduct the interview. The results of the interview are displayed on a screen viewable by the manager and provided to the manager.

[0085] As shown in FIG. 12, the issue block 40 has an improvement proposal list button 42. When the noted manager selects the improvement proposal list button 42, the terminal information processing unit 13R pops up an improvement proposal list screen 43. FIG. 13 is a diagram showing an example of the improvement proposal list screen 43. As shown in FIG. 13, the improvement proposal list screen 43 has an evaluator improvement proposal block 44 for each improvement proposal group belonging to the corresponding issue group. In the example of FIG. 13, the top two of the three blocks are evaluator improvement proposal blocks 44. The evaluator improvement proposal block 44 displays a summary of the improvement proposal. By referring to each evaluator improvement proposal block 44, the noted manager can accurately understand the content / tendency of the evaluator improvement proposal proposed by the evaluator for the corresponding evaluated department issue. The evaluator improvement proposal block 44 has an answer list button 45. When the noted manager selects the answer list button 45, the terminal information processing unit 13R pops up a screen displaying a list of improvement proposal-related answers belonging to the corresponding improvement proposal group. By referring to this screen, the administrator in question can accurately recognize what specific improvement proposals the evaluator has responded with. The evaluator improvement proposal block 44 is provided with a group of radio buttons 46 for selecting adopt, hold, or pass on. The administrator in question can use the group of radio buttons 46 to register the action (adopt, hold, or pass on) for each improvement proposal. The registered actions are shared among authorized personnel.

[0086] As shown in FIG. 13, the improvement proposal list screen 43 has an AI improvement proposal block 47 for each AI improvement proposal that belongs to the corresponding task group. In the example of FIG. 13, of the three blocks, the bottom block corresponds to the AI ​​improvement proposal block 47. The AI ​​improvement proposal block 47 displays the AI ​​improvement proposal. By referring to each of the AI ​​improvement proposal blocks 47, the focused manager can accurately understand the content of the AI ​​improvement proposal generated by the language model 5 function for the corresponding evaluated department task. The AI ​​improvement proposal block 47 has a group of radio buttons 46, allowing the focused manager to register actions for the AI ​​improvement proposal.

[0087] As described above, the information processing unit of the information processing system according to this embodiment has the following functions. A function that provides a questionnaire screen 23 (specific user interface) for employees (members) of a focused company (organization) that has multiple departments (groups) to input evaluation values ​​to be set for departments (other groups) different from the department (own group) to which the employee belongs. A function that uses the language model 5 to conduct a specific interview, which is an interview about one other department, with an employee who has entered an evaluation value for that other department on the questionnaire screen 23. A function for collecting input evaluation values, which are evaluation values ​​entered on the questionnaire screen 23, and interview results of specific interviews for multiple employees and storing them in the server memory unit 12 (memory unit). A function to derive a range average value (statistic) based on the input evaluation values ​​of multiple members stored in the server memory unit 12 for each evaluation pair, which is a combination of an evaluating department (group evaluating) and an evaluated department (group being evaluated) in a specific interview. -A function that provides range average values ​​(statistics) to administrators. A function of analyzing interview responses given by employees in specific interviews corresponding to evaluation pairs based on the interview results of multiple employees stored in the server memory unit 12 using the function of the language model 5, and deriving analysis results. -A function that provides analysis results to administrators.

[0088] According to this configuration, a range average value for each evaluation pair is provided to the manager based on the evaluation values ​​assigned by organizational members to other departments. Therefore, by referring to the range average value, the manager can accurately identify department combinations for which improved relationships are required. Furthermore, according to the present disclosure, interviews are conducted with members regarding other departments, and analysis results based on the interview results are provided to the manager. Therefore, the manager can obtain useful information regarding organizational improvement for the evaluation pairs, reflecting the responses to the interviews. In other words, the above configuration can provide a method that contributes to effective organizational improvement.

[0089] Second Embodiment Next, a second embodiment will be described. In this embodiment, the operation of the control system 1 in the investigation event is different. Below, the operation of the control system 1 according to this embodiment will be described for each scene.

[0090] <Scene of an employee responding to a specific survey> In the following, one of the employees will be referred to as a "notable employee," and the operation of the control system 1 in a scene where the notable employee responds to a specific questionnaire will be described. In this embodiment, the notable employee belongs to Department B. Furthermore, the other departments with which the notable employee collaborates are Departments A, C, F, and G.

[0091] As in the first embodiment, when a survey event is held, employees of the target company are notified in a predetermined manner that they must complete a specific questionnaire. In response to this notification, the target employee launches the dedicated app AP on their employee terminal 3E and logs in to the service. After logging in, the target employee performs a predetermined operation to instruct the display of a questionnaire screen 50 (specific user interface). In response to this instruction, the terminal information processing unit 13E displays the questionnaire screen 50 on the terminal display unit 15E. FIG. 14 shows an example of the questionnaire screen 50. However, the questionnaire screen 50 in FIG. 14 is not the initial state of the questionnaire screen 50, but rather shows a screen in which evaluation values ​​have been entered. As shown in FIG. 14, the questionnaire screen 50 has an input field for entering evaluation values ​​for each of the other collaborative departments. The questionnaire screen 50 corresponds to a "specific user interface for members of an organization with multiple groups to enter evaluation values ​​for groups other than the group to which they belong." The terminal information processing unit 13E has a function for providing such a specific user interface.

[0092] In the specific questionnaire in this example, the employee in question corresponds to the "evaluator," Department B to which the employee in question belongs corresponds to the "evaluating department," and Departments A, C, F, and G, which are other departments with which the employee in question collaborates, correspond to the "evaluated department."

[0093] The notable employee inputs an evaluation value into the input field for each collaborating department on the questionnaire screen 50 based on his or her own thoughts, feelings, and experiences, and selects the confirm button 51. The input evaluation value corresponds to the "input evaluation value." In this example, the notable employee inputs the evaluation value for each collaborating department in the format shown in FIG. 14. When the notable employee selects the confirm button 51, the control system 1 executes the following process. Below, an example of the operation of the control system 1 when the confirm button 51 is selected will be explained using a flowchart. FIG. 15 is a flowchart showing an example of the operation of the control system 1 when the confirm button 51 is selected. Flowchart FE in FIG. 15 shows the information processing method by the employee terminal 3E, and flowchart FF shows the information processing method by the information processing server 2.

[0094] As shown in flowchart FE, when the Confirm button is selected by the noted employee (step SE1), the terminal information processing unit 13E determines whether there are any other collaborating departments for which the noted employee has assigned a low evaluation value (step SE2). In this embodiment, a "low evaluation value" means that the evaluation value is equal to or less than a predetermined fixed first threshold value. In this example, the first threshold value is assumed to be "4 points." Note that the terminal information processing unit 13E may be configured to switch the display content of the terminal display unit 15E to a display content corresponding to the interview being conducted in response to the selection of the Confirm button 51.

[0095] The method for determining whether an evaluation value is low is not limited to the method exemplified in this embodiment. For example, the terminal information processing unit 13E may be configured to determine that a certain collaborating department's evaluation value is low if the evaluation value in the current specific questionnaire for that collaborating department is lower than the evaluation value in the previous specific questionnaire by a certain amount or more. Alternatively, the terminal information processing unit 13E may be configured to determine that a certain collaborating department's evaluation value is low if the evaluation value for that collaborating department gradually decreases from the previous specific questionnaire to the current specific questionnaire. The terminal information processing unit 13E may also be configured to dynamically change the first threshold based on the content of past questionnaires for the focused employee or the content of questionnaires for employees other than the focused employee. In other words, any method can be used to determine whether an employee has a negative impression of the target department, or if their impression has worsened.

[0096] If it is determined in step SE2 that there are no other collaborating departments with low evaluation values ​​(step SE2: NO), the terminal information processing unit 13E proceeds to step SE5. On the other hand, if it is determined that there are other collaborating departments with low evaluation values ​​(step SE2: YES), the terminal information processing unit 13E identifies other collaborating departments with low evaluation values ​​(step SE3). Referring to FIG. 14, in this example, the first threshold is 4 points, so in step SE3, the terminal information processing unit 13E identifies departments C and G as "other collaborating departments with low evaluation values." Hereinafter, the other collaborating departments identified in step SE3 will be referred to as "low evaluation departments."

[0097] After step SE3, the terminal information processing unit 13E executes interview processing (step SE4). The interview processing is processing for realizing a specific interview. The specific interview realized in the interview processing according to this embodiment is conducted for the purpose of obtaining answers to at least the following questions for each of the low-rated departments: · Challenges in the low-rated departments (What do you think are the challenges in the low-rated departments?) · What caused the issues in the low-rated departments? (What do you think are the causes of the issues?) - Suggestions for improving the relationship between your department and the department with low ratings (What suggestions can you think of to improve the relationship between your department and the department with low ratings?) Hereinafter, "issues with low-rated departments" will be referred to as "issues with low-rated departments," "causes of issues with low-rated departments" will be referred to as "causes of issues," and "improvement proposals to improve the relationship between the low-rated departments and one's own department" will be referred to as "relationship improvement proposals." Note that the questions are not limited to those exemplified. For example, they may include "experiences that led to a negative impression" or "reason for assigning a low evaluation value to the low-rated department."

[0098] FIG. 16 is a diagram showing details of interview processing by the terminal information processing unit 13E together with processing by the language model server 4. In FIG. 16, flowchart FG shows the information processing method by the employee terminal 3E, and flowchart FH shows the information processing method by the language model server 4. As shown in flowchart FG, the terminal information processing unit 13E generates an initial prompt (step SG1). The initial prompt is a prompt that requests the language model server 4 (language model 5) to generate the first question to be posed to the target employee in the specific interview (hereinafter referred to as the "initial question"). The initial prompt describes information that requests a response with the "department targeted by the question (=department related to the question)" and the "content of the question (specific wording to be posed as a question in the specific interview)."

[0099] The initial prompt also describes the above-mentioned purpose of the specific interview, and also describes information requesting that the interview be conducted with the employee reflecting that purpose (generating questions to be used in the interview; the same applies below). The initial prompt also describes attribute information for the notable employee, and also describes information requesting that the interview be conducted with those attributes reflected. The initial prompt also describes attribute information for the notable employee's own department (Department B) and attribute information for the low-rated departments (Departments C and G), and also describes information requesting that the interview be conducted with those attribute information reflected.

[0100] After generating the initial prompt, the terminal information processing unit 13E transmits the generated initial prompt to the language model server 4 (step SG2). As shown in flowchart FH, the language model server 4 receives the initial prompt and generates a response sentence based on the initial prompt using the language model 5 (step SH1). The language model server 4 generates the response sentence by reflecting the purpose of the specific interview and other matters described in the initial prompt. As described above, the response sentence includes the department (evaluated department) targeted by the question (initial question) and the specific content of the question (initial question). The language model server 4 transmits the generated response sentence to the terminal information processing unit 13E (step SH2).

[0101] The terminal information processing unit 13E receives and acquires the response sentence (step SG3). Next, based on the question content of the response sentence, the terminal information processing unit 13E controls the audio output unit 18E to output the wording indicated by the question content (i.e., the question) by voice (step SG4). The noted employee, who has heard the question output by voice, orally responds to the question (step SG5). The spoken voice is picked up by the microphone of the audio input unit 17E, and the voice data is stored in a buffer. The terminal information processing unit 13E converts the voice data into text data containing text (wording) corresponding to the answer (step SG6). Next, the terminal information processing unit 13E generates individual question answer information based on the question output by voice in step SG4 (hereinafter referred to as the "most recent question") and the answer given by the noted employee in step SG5 (hereinafter referred to as the "most recent answer") (step SG7).

[0102] FIG. 17 is a diagram showing the contents of the individual question answer information. As shown in FIG. 17, the individual question answer information includes an evaluating department ID, an evaluated department ID, question content, and answer content. The evaluating department ID is the department ID of the department to which the employee of interest belongs. The evaluated department ID is the department ID of the department targeted by the most recent question. The most recent answer is considered to have been made for the evaluated department. The terminal information processing unit 13E stores the evaluated department ID in the individual question answer information based on the "department targeted by the question" in the response sentence acquired in step SG3. The question content is information indicating the content of the most recent question, and the answer content is information indicating the content of the most recent answer.

[0103] After step SG7, the terminal information processing unit 13E generates a prompt requesting the generation of a second question based on the text data generated in step SG6 (step SG8). The prompt includes the answer described in the text data (i.e., the answer given by the noted employee to the first question) and also includes information requesting the generation of a second question that further reflects the answer. Furthermore, prompts for the second and subsequent questions include information requesting that, if the question is the last, information to that effect be included in the response sentence. The terminal information processing unit 13E sends the generated prompt to the language model server 4 (step SG9). Thereafter, the second, third, ..., and final questions are provided to the noted employee in the same manner as the first question, and the noted employee answers the second, third, ..., and final questions. The terminal information processing unit 13E generates individual question answer information each time the noted employee answers a question.

[0104] For example, the following conversation might take place between a featured employee and an AI interviewer: [AI Interviewer] What do you think are the challenges facing Department C? [Featured Employee] They're always slow to respond. [AI Interviewer] What do you think is the reason for Department C's slow response? [Featured Employee] I think it's simply a lack of staff. [AI Interviewer] So how do you think this can be resolved? ... [AI Interviewer] This is my last question. What ideas do you have for improving the relationship between your Departments B and C? [Featured Employee] I think it would be good to have an environment where both departments could check how busy they are. [AI Interviewer] This concludes the interview. Thank you for your hard work.

[0105] When the interview process is completed, individual question and answer information is generated for each combination of a question posed to the target employee and the answer given by the target employee to the question. As described above, the specific interview is conducted interactively, which reduces the workload on the target employee. As described above, the terminal information processing unit 13E has the function of, when the evaluation value (evaluation value) for one other department (one other group) entered by the employee on the questionnaire screen 50 (specific user interface) is low, conducting an interview in accordance with the low evaluation value of the one other department using the function of the language model 5. The terminal information processing unit 13E also has the function of providing the employee's attributes to the language model 5 and conducting an interview that reflects the employee's attributes. The terminal information processing unit 13E also has the function of providing the attributes of the employee's own department (own group) and the attributes of the other department (other group) to the language model 5 and conducting an interview that reflects the attributes of the own department and the other department. The above is a detailed description of the interview process.

[0106] As shown in flowchart FE of FIG. 15, when the interview process is completed, the terminal information processing unit 13E proceeds to step SE5. In step SE5, the terminal information processing unit 13E transmits the input results on the questionnaire screen 50 (hereinafter referred to as "questionnaire input results") and all of the individual question answer information generated during the specific interview (however, there may be cases where individual question answer information is not generated) to the server information processing unit 10. The questionnaire input results are information based on the input on the questionnaire screen 50 and include evaluation values ​​assigned by the noted employee for each collaborating department. The individual question answer information is information equivalent to the interview results. As shown in flowchart FF, the server information processing unit 10 receives the questionnaire input results and the individual question answer information (step SB1). Next, the server information processing unit 10 records the received information in the current survey event management database 21-1 in an appropriate manner (step SB2). This completes the operation of the control system 1 in the <scene in which an employee responds to a specific questionnaire>.

[0107] <Scene for generating improvement-related information> After all employees have answered the specific questionnaire, the server information processing unit 10 generates improvement-related information (described later). The operation of the server information processing unit 10 in a scene where the improvement-related information is generated will be described below. In this scene, the server information processing unit 10 first executes an improvement-required pair extraction process. The improvement-required pair extraction process will be described in detail below.

[0108] In the process of extracting pairs requiring improvement, the server information processing unit 10 executes the following process based on the questionnaire entry results for each employee stored in the survey event management database 21. That is, the server information processing unit 10 derives the average evaluation value (statistic) for each evaluation pair. For example, the average evaluation value for "Division A → Division B" refers to the average evaluation value given by employees belonging to Division A, the evaluating department, to Division B, the evaluated department. For example, if two employees E1 and E2 belong to Division A, and employee E1 gives Division B an evaluation value of 10 points and employee E2 gives Division B an evaluation value of 5 points, the average evaluation value for "Division A → Division B" is 7.5 points. The average evaluation value for "Division B → Division A" is the average evaluation value given by employees belonging to Division B to Division A.

[0109] After deriving the average evaluation value for each evaluation pair, the server information processing unit 10 extracts, from among the evaluation pairs, evaluation pairs whose average evaluation value is equal to or less than a second threshold as pairs requiring improvement. In this embodiment, the second threshold is set to "4 points." A pair requiring improvement is a combination of departments whose relationship requires improvement. FIG. 18 is a diagram showing an example of the average evaluation value for each evaluation pair using a heat map. When the average evaluation value for each evaluation pair is as shown in FIG. 18, the server information processing unit 10 extracts "Department A → Department B" (average evaluation value: 2.1 points), "Department B → Department D" (average evaluation value: 3.2 points), "Department B → Department G" (average evaluation value: 3.4 points), "Department D → Department C" (average evaluation value: 3.2 points), "Department G → Department A" (average evaluation value: 3.8 points), and "Department G → Department B" (average evaluation value: 3.1 points) as pairs requiring improvement.

[0110] For an evaluation pair in which one department is the evaluating department and the other department is the evaluated department, the lower the average evaluation value, the stronger the tendency for employees of the first department to have a negative impression of the other department. Based on this, it is reasonable to identify evaluation pairs with an average evaluation value below a certain level as "pairs of departments requiring relationship improvement." However, the method by which the server information processing unit 10 extracts pairs requiring improvement is not limited to the example method. For example, the server information processing unit 10 may be configured to determine an evaluation pair as requiring improvement if the average evaluation value for the current survey event for a certain evaluation pair is lower than the average evaluation value for the previous survey event by more than a certain amount. Alternatively, the server information processing unit 10 may be configured to determine an evaluation pair as requiring improvement if the average evaluation value for the evaluation pair gradually decreases from the previous survey event to the current survey event. The server information processing unit 10 may also be configured to dynamically change the second threshold. In other words, the extraction of pairs requiring improvement may be performed by any method as long as evaluation pairs with a high likelihood of needing relationship improvement are extracted as improvement pairs.

[0111] As described above, the server information processing unit 10 has a function to extract pairs requiring improvement, which are combinations of groups whose relationships require improvement, based on the input evaluation values ​​of multiple employees (members) stored in the server storage unit 12 (storage unit). More specifically, the server information processing unit 10 has a function to derive statistics of evaluation values ​​for each combination of an evaluating department (evaluated group) and an evaluated department (evaluated group) based on the input results of multiple members stored in the server storage unit 12 (storage unit), and to determine whether or not the combination of the evaluating department and the evaluated department is a pair requiring improvement based on the statistics.

[0112] After the improvement-requiring pair extraction process is completed, the server information processing unit 10 executes improvement-related information generation process for each of the improvement-requiring pairs. Hereinafter, the improvement-requiring pairs to be processed are referred to as "target pairs," and the improvement-related information generation process executed by the server information processing unit 10 for the target pairs will be described in detail.

[0113] In the improvement-related information generation process for the pair of interest, the server information processing unit 10 first generates an analysis request prompt. The analysis request prompt is a prompt requesting the generation of the following information based on the individual question and answer information related to the pair of interest: - Information listing the issues of the department being evaluated (hereinafter referred to as "Comprehensive Issue Information") - Information listing the causes (reasons) for the issues that arose in the department being evaluated (hereinafter referred to as "general cause information") - Improvement plan to improve the relationship between the evaluating department and the evaluated department (hereinafter referred to as the "Comprehensive Improvement Plan") The comprehensive improvement plan may include multiple plans. The comprehensive issue information, comprehensive cause information, and comprehensive improvement plan are information related to improving the relationship between pairs requiring improvement, and correspond to "improvement-related information." Hereinafter, the comprehensive issue information, comprehensive cause information, and comprehensive improvement plan are collectively referred to as "improvement-related information." The analysis request prompt includes each of the individual question and answer information related to the pair of interest. The individual question and answer information related to the pair of interest refers to the individual question and answer information whose evaluating department and evaluated department match the pair of interest. The analysis request prompt also includes information requesting the generation of comprehensive issue information, comprehensive cause information, and comprehensive improvement plan based on each of the individual question and answer information related to the pair of interest.

[0114] The analysis request prompt may include the following information: Attributes of the evaluating department and the evaluated department, and information requesting the generation of improvement-related information that reflects those attributes Improvement-related information related to past survey events related to the pair of interest, and information requesting the generation of improvement-related information reflecting such information -Information on individual questions and answers related to past survey events related to the pair of interest, and information requesting the generation of improvement-related information by reflecting each of the information

[0115] After generating the analysis request prompt, the server information processing unit 10 transmits the prompt to the language model server 4. The language model server 4 receives the analysis request prompt and performs the following processing based on the analysis request prompt. That is, the language model server 4 uses the functions of the language model 5 based on the response sentence to generate a response sentence that includes comprehensive problem information, comprehensive cause information, and comprehensive improvement plans, reflecting the individual question and answer information related to the pair of interest. The language model server 4 transmits the generated response sentence to the server information processing unit 10.

[0116] The server information processing unit 10 receives and acquires the response text. Next, the server information processing unit 10 acquires (generates) the comprehensive issue information, comprehensive cause information, and comprehensive improvement plan contained in the response text. The comprehensive improvement plan acquired here is an improvement plan that reflects the results of a specific interview conducted with employees belonging to the evaluation department of the target pair who assigned a low evaluation value to the evaluated department of the target pair. Therefore, there is a high probability that it is a feasible improvement plan that reflects real voices from the field. The same applies to the comprehensive issue information and comprehensive cause information. The server information processing unit 10 associates this acquired information with the identification information of the target pair and registers it in the investigation event management database 21 in a predetermined format. The above is the details of the improvement-related information generation process for the target pair.

[0117] As described above, the server information processing unit 10 executes the improvement-related information generation process for each pair requiring improvement. As a result, comprehensive issue information, comprehensive cause information, and comprehensive improvement plan (improvement-related information) are generated for each pair requiring improvement, and are registered in the investigation event management database 21.

[0118] As described above, the server information processing unit 10 has a function of generating, by the function of the language model 5, improvement-related information related to the improvement of the relationship between the improvement-requiring pair based on the interview results related to the improvement-requiring pair stored in the server storage unit 12 (storage unit). The terminal information processing unit 13E also has a function of conducting an interview one of the purposes of which is to obtain an answer regarding an improvement plan for improving the relationship between the improvement-requiring pair and the other department based on the interview results related to the improvement-requiring pair stored in the server storage unit 12, by the function of the language model 5. The terminal information processing unit 13E also has a function of conducting an interview one of the purposes of which is to obtain an answer regarding at least one of the issues in the other department and the causes of the issues in the other department, in addition to the improvement plan for improving the relationship between the improvement-requiring pair and the other department, by the function of the language model 5. The server information processing unit 10 also has a function of generating, by the function of the language model 5, improvement-related information including an improvement plan for improving the relationship between the improvement-requiring pair based on the interview results related to the improvement-requiring pair stored in the server storage unit 12.

[0119] <Scene for providing improvement-related information> Next, the operation of the control system 1 in a scene where improvement-related information is provided will be described. In this embodiment, improvement-related information is provided to a manager for each pair requiring improvement. The managers of one pair requiring improvement are, for example, the manager of the evaluating department of the one pair requiring improvement and the manager of the evaluated department. Below, the operation of the control system 1 when improvement-related information is provided to the manager of department B (hereinafter referred to as the "monitored manager") will be described.

[0120] The noted manager launches the dedicated application AP on his / her manager terminal 3R and logs in to the service. After logging in, the noted manager performs a predetermined operation to instruct the display of a list screen of pairs requiring improvement (not shown). In response to this instruction, the terminal information processing unit 13R displays the list screen of pairs requiring improvement on the terminal display unit 15R. The list screen of pairs requiring improvement is a screen that displays a selectable list of pairs requiring improvement related to the noted manager's own department (Department B in this example). A pair requiring improvement related to the own department refers to a pair requiring improvement for which the own department is the evaluating department or the evaluated department. Referring to FIG. 18, in this example, "Department A → Department B" (average evaluation score: 2.1 points), "Department G → Department B" (average evaluation score: 3.1 points), "Department B → Department D" (average evaluation score: 3.2 points), and "Department B → Department G" (average evaluation score: 3.4 points) are pairs requiring improvement related to Question B, and these pairs requiring improvement are displayed in a list on the list screen of pairs requiring improvement.

[0121] The noted manager selects one of the pairs requiring improvement on the improvement-requiring pair list screen. In this example, it is assumed that "Department A → Department B" is selected. Hereinafter, the selected pair requiring improvement will be referred to as the "selected pair." When one pair requiring improvement is selected, the terminal information processing unit 13R displays an improvement-related screen 53 corresponding to the selected pair on the terminal display unit 15R. Figure 19 is a diagram showing an example of the improvement-related screen 53. As shown in Figure 19, the improvement-requiring related screen 53 displays the evaluating department and the evaluated department. The improvement-requiring related screen 53 also displays the average evaluation value for the current investigation event and the average evaluation value for the previous investigation event. The improvement-requiring related screen 53 also displays comprehensive issue information (symbol 54), comprehensive cause information (symbol 55), and comprehensive improvement proposals (symbol 56).

[0122] The comprehensive issue information is information indicating issues that reflect the results of interviews conducted with employees of the evaluating department (in this example, Department A) about the department being evaluated (in this example, Department B). The focused manager can refer to the comprehensive issue information to accurately understand what issues his / her department has in relation to the evaluating department. The comprehensive cause information is information indicating causes that reflect the results of interviews conducted with employees of the evaluating department (in this example, Department A) about the department being evaluated (in this example, Department B). The focused manager can refer to the comprehensive issue information to accurately understand what causes the issues his / her department has in relation to the evaluating department. The comprehensive improvement plan is information indicating improvement plans that reflect the results of interviews conducted with employees of the evaluating department (in this example, Department A) about the department being evaluated (in this example, Department B). As described above, the comprehensive improvement plan is highly likely to be a feasible improvement plan that reflects real voices from the field. The focused manager can refer to the comprehensive improvement plan to recognize or develop effective improvement plans.

[0123] As shown in FIG. 19 , the improvement-related screen 53 has input fields 57, 58, and 59 for entering the details of the issues, causes, and improvement plans agreed upon between the evaluation department and the evaluated department. The person in charge of the issue can use the comprehensive issue information, comprehensive cause information, and comprehensive improvement plans displayed on the improvement-related screen 53 to discuss with the manager of the evaluation department (in this example, the manager of Department A) and reach an agreement on the issues in their department, the causes of the issues, and improvement plans for improving the relationship between their department and the evaluation department. The person in charge of the issue can enter the details of the issues, causes, and improvement plans agreed upon with the evaluator (hereinafter referred to as "agreements") in the input fields 57, 58, and 59 and then save the contents. The agreements are recorded in the investigation event management database 21. The person in charge of the issue can then submit the saved agreements to the manager or other appropriate person. Submission includes making the agreements available for reference. The above is a detailed description of the operation of the control system 1 in a scenario in which improvement-related information is provided.

[0124] As described above, the server information processing unit 10 has a function of providing a questionnaire screen 50 (specific user interface) to employees (members) of a target company (organization) having multiple departments (groups) to input evaluation values ​​(evaluation values) for departments (other groups) different from the employee's own department (own group). If the evaluation value for one other department input on the questionnaire screen 50 is low, the terminal information processing unit 13E has a function of using the language model 5 to conduct a specific interview corresponding to the low evaluation value for the one other department. The server information processing unit 10 has a function of collecting the input results on the questionnaire screen 50 and the interview results of the specific interview for multiple employees and storing them in the server storage unit 12. The server information processing unit 10 has a function of extracting improvement-required pairs, which are combinations of departments whose relationship requires improvement, based on the input results of multiple employees stored in the server storage unit 12. The server information processing unit 10 has a function of using the language model 5 to generate improvement-related information related to improving the relationship between the improvement-required pairs based on the interview results related to the improvement-required pairs stored in the server storage unit 12. The terminal information processing unit 13R has a function of providing improvement-related information to the manager (stakeholder) of the pair requiring improvement.

[0125] According to the above configuration, a combination of departments (improvement-required pairs) whose relationship needs to be improved is extracted based on the evaluation values ​​given by employees of the organization to other groups. Therefore, it is possible to accurately extract a combination of departments whose relationship needs to be improved while reflecting the employees' actual evaluations of other departments. Furthermore, according to the above configuration, if the evaluation value of other departments is low, an interview is conducted. Furthermore, based on the results of the interview, improvement-related information related to improving the relationship between the improvement-required pairs is generated and provided to the manager. Therefore, it is possible to provide the manager with accurate information that reflects the employees' thoughts about other departments regarding the improvement of the relationship between the improvement-required pairs, and the manager can effectively improve the organization using the provided improvement-related information. In other words, the above configuration provides a method that contributes to effective organizational improvement. As described above, the terminal information processing unit 13R has a function of providing improvement-related information to the manager (stakeholder) of the improvement-required pair by displaying the improvement-related screen 53 displaying the improvement-related information on the terminal display unit 15R.

[0126] Although one embodiment of the present invention has been described above, the above embodiment merely illustrates one example of a specific embodiment of the present invention, and the technical scope of the present invention should not be construed as being limited thereby. In other words, the present invention can be embodied in various forms without departing from the gist or main features thereof. Below, variations of the above embodiment are presented. The following variations may be combined when possible.

[0127] In the above embodiment, the company corresponds to the "organization," the employees correspond to the "members," and the managers of the target company correspond to the "manager." However, the organizations, members, and managers are not limited to the objects exemplified in the above embodiment. In other words, an organization may be an object that has multiple groups and for which it may become necessary to consider the relationships between the groups. A member may be an object that belongs to an organization. A manager may be an object to which improvement-related information can be provided. In the above embodiment, a department corresponds to a "group." However, the group is not limited to the examples given in the above embodiment. In other words, the group may be any group in which there may be a need to improve relationships among different groups. For example, a department, division, team, group, section, or business division may function as a group. · Specific interviews can be conducted via chat rather than through conversation.

[0128] In the above embodiment, an example of the purpose of the specific interview was given, but the purpose is not limited to the example. For example, in the second embodiment, the purpose may not include obtaining answers about the issues of low-rated departments or the causes of the issues. However, the purpose is determined to be content that takes into account the content / direction of the information that will ultimately be provided to the administrator. The terminal information processing unit 13E writes information according to the purpose in the prompt for the specific interview. In the second embodiment, the content of the information provided to the manager as improvement-related information is not limited to that exemplified in the above embodiment. For example, the configuration may be such that the comprehensive issue information or comprehensive cause information is not provided. In the above embodiment, the specific interview was conducted when the employee entered the evaluation value. However, the timing of the specific interview is not limited to the timing illustrated. For example, the specific interview may be conducted when the employee explicitly instructs it to be conducted after the employee has entered the evaluation value.

[0129] In the above embodiment, the average is used to calculate the statistical quantity. However, the statistical method for calculating the statistical quantity is not limited to the average, and other methods may be used. For example, the statistical quantity may be a value derived by weighted average or a median. In the above embodiment, information is provided by displaying the screen, but the information may be provided by a method other than displaying the screen. For example, the information may be provided by "provision of a file via email, communication tools, or other means," or "printing out a document containing the information." The form of the evaluation value is not limited to the example shown. For example, the evaluation value may be a three-level score. In the above embodiment, the entity using the service is a single company, but this is for the sake of simplicity, and the service can naturally be used by multiple organizations. The method of the survey is not limited to the example method. For example, the survey may be conducted through dialogue or chat. The specific user interface is a concept that includes all user interfaces that allow members to input evaluation values. The format of the statistics table is not limited to a heat map. For example, a list displaying statistics for each evaluation pair may be used.

[0130] In the above embodiment, a configuration may be adopted in which a browser executes some or all of the processing executed by the dedicated application AP. In the above embodiment, a part or all of the data stored in the server storage unit 12 may be stored in an external device accessible by the server information processing unit 10 (for example, a database server connected to the network N). The functional blocks shown in the above embodiments can be realized by any hardware or by a combination of any hardware and any software. In other words, these functional blocks are not limited to specific hardware.

[0131] The embodiments may also include the provision of a program executed by a computer in the information processing server 2, the employee terminal 3E, or the manager terminal 3R. The embodiments may also include the provision of a recording medium on which the program is recorded in a computer-readable manner. The recording medium may be a magnetic or optical recording medium or a semiconductor memory device. Specific examples include portable or fixed recording media such as flexible disks, hard disk drives (HDDs), compact disk read-only memories (CD-ROMs), digital versatile disks (DVDs), Blu-ray (registered trademark) discs, magneto-optical disks, flash memories, and card-type recording media.

[0132] Furthermore, in the above-described embodiments, a device having a certain function is not limited to the device exemplified as a device having that function. For example, let us consider a function for generating a positive opinion summary (referred to as the "first function") and a function for providing the positive opinion summary to an administrator (referred to as the "second function"). In the above-described embodiment, as shown in FIG. 20A, the server information processing unit 10 of the information processing server 2 has the first function, and the terminal information processing unit 13R of the administrator terminal 3R has the second function. In this configuration, the combination of the information processing server 2 and the administrator terminal 3R functions as an "information processing system." Furthermore, the server information processing unit 10 and the terminal information processing unit 13R each function as an "information processing unit." On the other hand, as shown in FIG. 20B, the terminal information processing unit 13R of the administrator terminal 3R may have the first and second functions. In this configuration, the administrator terminal 3R functions as an "information processing system," and the terminal information processing unit 13R functions as an "information processing unit." 20(C), the server information processing unit 10 may have the first and second functions. In this configuration, the information processing server 2 functions as an "information processing system," and the server information processing unit 10 functions as an "information processing unit." When the information processing units of multiple devices (e.g., the server information processing unit 10 and the terminal information processing unit 13R) each function as an "information processing unit," the functions provided in each information processing unit are not limited. A single function may also be realized by multiple information processing units. A device that realizes a certain function may be a device other than the information processing server 2, the employee terminal 3E, and the manager terminal 3R. [Explanation of symbols]

[0133] 1. Control system (information processing system) 2. Information processing server (information processing system) 3E Employee terminal (information processing system) 3R Administrator terminal (information processing system) 10 Server information processing unit (information processing unit) 12 Server memory unit (memory unit) 13E Terminal information processing unit (information processing unit) 13R Terminal information processing unit (information processing unit) 23, 50 Questionnaire screen (specific user interface)

Claims

1. A function of providing a specific user interface for inputting evaluation values ​​set for groups other than the group to which the member of an organization having multiple groups belongs, to the member of the organization; a function of conducting a specific interview regarding one other group with a member who has input an evaluation value of the other group into the specific user interface using a language model function; a function of collecting and storing in a storage unit input evaluation values, which are evaluation values ​​input into the specific user interface, and interview results of specific interviews for a plurality of members; a function of deriving statistics based on the input evaluation values ​​of the plurality of members stored in the storage unit for each evaluation pair, which is a combination of a group of evaluators and a group of people being evaluated in the specific interview; providing said statistics to an administrator; a function of analyzing interview responses given by members in the specific interview corresponding to the evaluation pair based on the interview results of the plurality of members stored in the storage unit by using the function of the language model, and deriving an analysis result; an information processing unit having a function of providing the analysis result to the administrator; An information processing system comprising:

2. The information processing unit The language model is provided with a function of providing attribute information of members and conducting the specific interview by reflecting the attribute information of members.

2. The information processing system according to claim 1, wherein:

3. The information processing unit The language model includes a function of providing attribute information of the own group and the other group to the language model, and conducting the specific interview while reflecting the attribute information of the own group and the other group.

2. The information processing system according to claim 1, wherein:

4. The information processing unit a function of providing the administrator with a statistics table that displays the statistics of each of the evaluation pairs; 2. The information processing system according to claim 1, wherein:

5. The information processing unit The statistical table has a function of applying a decoration indicating that the level of the statistical quantity is low for the evaluation pairs having a low level of the statistical quantity.

5. The information processing system according to claim 4.

6. The information processing unit a function of conducting the specific interview, one of the purposes of which is to obtain answers about the other group's problem, using the function of the language model; A function of using the language model to group, for one evaluation pair, task-related answers regarding the other group task given by members in each of the specific interviews corresponding to the one evaluation pair based on the interview results of multiple members stored in the storage unit, and to generate task summaries that are summaries based on the task-related answers for each task group generated by the grouping; and a function of providing the problem summary to the administrator.

2. The information processing system according to claim 1, wherein:

7. The information processing unit a function of providing the administrator with, for each of the task groups, the task summary and each of the task-related answers in a selectable state; and a function of, when one task-related answer is selected by the administrator, using the function of the language model, conducting an interview with the member who issued the task-related answer to ask more detailed questions about the other group task.

7. The information processing system according to claim 6.

8. The information processing unit A function of conducting the specific interview, one of the purposes of which is to obtain a member improvement plan, which is an improvement plan for improving the other group task, by using the function of the language model; a function of using the language model to group improvement proposal-related answers regarding the member improvement proposals for one other group task given by members in each of the specific interviews corresponding to one evaluation pair for one other group task based on the interview results of multiple members stored in the storage unit, and to generate improvement proposal summaries that are summaries based on the improvement proposal-related answers for each improvement proposal group generated by the grouping; and a function of providing the manager with a summary of the improvement plan.

7. The information processing system according to claim 6.

9. The information processing unit A function of generating an AI improvement plan, which is an improvement plan for improving the other group task corresponding to one task group, based on the task-related answer and the improvement plan-related answer corresponding to the one task group using the function of the language model; and a function of providing the AI ​​improvement plan to the administrator.

9. The information processing system according to claim 8.

10. The information processing unit a function of conducting the specific interview according to the low level of the evaluation value of one other group input to the specific user interface by using a language model function when the evaluation value of the one other group is low; a function of extracting pairs requiring improvement from the evaluation pairs, the pairs requiring improvement of a relationship, based on the statistics of each of the evaluation pairs; a function of generating, by the function of the language model, improvement-related information related to the improvement of the relationship of the pair requiring improvement based on the interview results related to the pair requiring improvement stored in the storage unit; and a function of providing the improvement-related information to the administrator.

2. The information processing system according to claim 1, wherein:

11. The information processing unit a function of generating, by the function of the language model, improvement-related information including an improvement plan for improving the relationship of the pair requiring improvement based on the interview results related to the pair requiring improvement stored in the storage unit; 11. The information processing system according to claim 10.

12. The information processing unit A function of conducting the specific interview, one of the purposes of which is to obtain answers regarding improvement plans for improving the relationship between the self-group and the other group, using the function of the language model; and a function of generating, by the function of the language model, the improvement-related information including an improvement plan for improving the relationship of the pair requiring improvement, based on the interview results related to the pair requiring improvement stored in the storage unit.

11. The information processing system according to claim 10.

13. The information processing unit A function of conducting the specific interview, one of the purposes of which is to obtain answers regarding at least one of the issues of the other group and the causes of the issues of the other group, in addition to improvement proposals for improving the relationship between the self-group and the other group, using the function of the language model; and a function of generating, by the function of the language model, the improvement-related information including an improvement plan for improving the relationship of the pair requiring improvement, based on the interview results related to the pair requiring improvement stored in the storage unit.

11. The information processing system according to claim 10.

14. an information processing unit of the information processing system providing a specific user interface for inputting evaluation values ​​to be set for groups other than the group to which the member belongs, to a member of an organization having a plurality of groups; the information processing unit of the information processing system, using a language model function, conducting a specific interview regarding the one other group with the member who inputs the evaluation value of the one other group into the specific user interface; the information processing unit of the information processing system collects input evaluation values, which are evaluation values ​​input to the specific user interface, and interview results of specific interviews for a plurality of members, and stores them in a storage unit; The information processing unit of the information processing system derives statistics based on the input evaluation values ​​of the plurality of members stored in the storage unit for each evaluation pair, which is a combination of a group of evaluators and a group of people being evaluated in the specific interview; the information processing unit of the information processing system providing the statistics to an administrator; The information processing unit of the information processing system analyzes interview responses given by members in the specific interview corresponding to the evaluation pair based on the interview results of the multiple members stored in the storage unit using the function of the language model, and derives analysis results; and a step in which the information processing unit of the information processing system provides the administrator with the analysis result. An information processing system comprising:

Images