Interview system, information processing method, and program

Abstract

It efficiently and simply detects the target object and outputs the detection results. [Solution] The interview system has at least one control unit. The control unit outputs a question. The question includes a question for detecting a target. The control unit receives the user's answer to the question. The control unit generates a prompt that includes the question, the answer to the question, and an instruction to detect the target from the answer. The control unit inputs the prompt into a trained model. The control unit receives output information for the prompt from the trained model. The output information includes the detection result of the target. The detection result includes a value that quantitatively indicates the degree to which the target is detected. The control unit outputs the detection result to an output destination corresponding to the value.

Description

Technical Field

[0004]

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

Background Art

[0002] Patent Document 1 discloses a technique capable of detecting harmful acts such as power harassment, sexual harassment, and bullying in a workplace environment and assisting in taking countermeasures.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the technique of Patent Document 1, it is necessary to collect and transmit the voices around the target person, or to collect and transmit vital data, and there is room for improvement in the efficient and simple detection of the detection target and the output of the detection result.

Means for Solving the Problems

[0005] According to one aspect of the present invention, an interview system is provided. The interview system has at least one or more control units. The control unit outputs a question. The question includes a question for detecting a detection target. The control unit receives an answer from the user to the question. The control unit generates a prompt including the question, the answer to the question, and an instruction to detect the detection target from the answer. The control unit inputs the prompt into a learned model. The control unit receives output information for the prompt from the learned model. The output information includes a detection result of the detection target. The detection result includes a value quantitatively indicating the degree of the detection target. The control unit outputs the detection result to an output destination according to the value.

Brief Description of the Drawings

[0006] [Figure 1] Figure 1 shows an example of the system configuration of the interview system. [Figure 2] Figure 2 shows an example of the hardware configuration of a server device. [Figure 3] Figure 3 shows an example of the hardware configuration of a client device. [Figure 4] Figure 4 is a sequence diagram showing an example of information processing in an interview system. [Figure 5] Figure 5 shows an example of a screen during an interview. [Figure 6] Figure 6 shows an example of output information. [Figure 7] Figure 7 shows an example of configuration data where the output destination is set in stages. [Figure 8] Figure 8 shows an example of a detailed screen for interview results. [Modes for carrying out the invention]

[0007] Embodiments of the present invention will be described below with reference to the drawings. The various features shown in the embodiments below (including modified examples; the same applies hereinafter) can be combined with each other.

[0008] <Embodiment 1> 1. System configuration of the interview system Figure 1 shows an example of the system configuration of the interview system 1000. As shown in Figure 1, the interview system 1000 includes a server device 100, client devices 110, 120, and 130 as its system configuration. The server device 100 is an example of a computer. The server device 100, client devices 110, 120, and 130 are connected to each other via a network 150. The network 150 is either a WAN (Wide Area Network), a LAN (Local Area Network), or the Internet, or any combination thereof. The network 150 is configured to allow communication between devices connected to the network 150 via wired and / or wireless connections. The interview system 1000 is a system that provides so-called SaaS (Software as a Service) functionality.

[0009] The server device 100 performs the processing of interviews with interviewees, detects targets from questions asked of the interviewees and their answers, and outputs the detection results. The details of the server device 100's processing will be explained later using the sequence diagram in Figure 4. The interview system 1000 may consist of one server device or multiple server devices. If the interview system 1000 consists of multiple server devices, the functions of the server device 100 are provided as a so-called distributed system.

[0010] The client device 110 is a device operated by the interviewer using the interview service provided by the server. The operator of the client device 110 will hereinafter also be simply referred to as the interviewer. In Figure 1, a PC (Personal Computer) is shown as an example of the client device 110, but it is not limited to a PC and may be a tablet terminal device or a smartphone, etc. The client device 110 can be any device that can input the operator's operation information and voice information, output data sent from the server device 100 as voice, or display a GUI (Graphical User Interface). In Figure 1, only one client device 110 is shown as an example included in the interview system 1000, but multiple client devices 110 may be included in the interview system 1000.

[0011] The client device 120 is a device operated by a supervisor or other person from the same organization (for example, the same company) as the interviewee.

[0012] The client device 130 is a device operated by a person in charge of a designated department (such as the Human Resources Department) within the same organization (for example, the same company) as the interviewee.

[0013] Here, the interview system described in the claims may consist of multiple devices or of a single device. If the interview system described in the claims consists of a single device, an example of such a device is, for example, a server device 100. If the interview system described in the claims consists of multiple devices, examples of the multiple devices are, for example, a distributed system that provides the functions of server device 100, or server device 100 and client device 110, or server device 100 and a server device that provides the functions of a large-scale language model described later.

[0014] 2. Hardware Configuration Diagram (1) Hardware configuration of server device 100 Figure 2 shows an example of the hardware configuration of server device 100. As shown in FIG. 2, the server device 100 includes, as a hardware configuration, a control unit 210, a storage unit 220, a communication unit 230, and an internal bus 240. The control unit 210, the storage unit 220, and the communication unit 230 are electrically connected via the internal bus 240.

[0015] The control unit 210 is a CPU (Central Processing Unit) or the like, and controls the entire server device 100.

[0016] The storage unit 220 is any one of an HDD (Hard Disk Drive), a ROM (Read Only Memory), a RAM (Random Access Memory), an SSD (Solid Sate Drive), or any combination thereof, and stores programs and data (for example, a list of questions for an interviewee described later (hereinafter, also simply referred to as a question list), GUI image data related to an avatar, answers from an interviewee, etc.) used when the control unit 210 executes processing based on the program.

[0017] The storage unit 220 is an example of a storage medium. In the specification, data used when the control unit 210 executes processing based on a program is described as being stored in the storage unit 220, but it may be stored in the storage unit of another device communicable with the server device 100. That is, the data may be stored in the storage unit of any device as long as the control unit 210 can refer to and / or acquire it. By the control unit 210 executing processing based on the program stored in the storage unit 220, the functions of the server device 100 and the processing of the server device 100 in the sequence diagram shown in FIG. 4 described later are realized.

[0018] The communication unit 230 connects the server device 100 to a network and controls communication with other devices (for example, other server devices and / or client devices 110, etc.).

[0019] The server device 100 may include multiple hardware configurations as shown in Figure 1. For example, the server device 100 may have multiple control units. The same applies to the client devices 110, 120, and 130 shown below.

[0020] (2) Hardware configuration of client device 110 Figure 3 shows an example of the hardware configuration of the client device 110. As shown in Figure 3, the client device 110 includes, as a hardware configuration, a control unit 310, a storage unit 320, an imaging unit 330, an input unit 340, an output unit 350, an audio input unit 360, an audio output unit 370, a communication unit 380, and an internal bus 390. The control unit 310, storage unit 320, imaging unit 330, input unit 340, output unit 350, audio input unit 360, audio output unit 370, and communication unit 380 are electrically connected via the internal bus 390.

[0021] The control unit 310 is a CPU or the like, and controls the entire client device 110.

[0022] The storage unit 320 is one of the following: HDD, ROM, RAM, SSD, etc., or any combination thereof, and stores programs, data used by the control unit 310 when executing processing based on the programs, etc. The storage unit 320 is an example of a storage medium.

[0023] In this specification, the data used by the control unit 310 when executing processing based on the program is described as being stored in the storage unit 320, but it may also be stored in the storage unit of another device that can communicate with the client device 110. The data may be stored in the storage unit of any device as long as the control unit 310 can access and / or retrieve it. The functions of the client device 110 are realized when the control unit 310 executes processing based on the program stored in the storage unit 320.

[0024] The imaging unit 330 is a camera or the like that takes images of a subject. An example of a subject is the person being interviewed.

[0025] The input unit 340 is a device that inputs information to the client device 110 in response to the interviewer's actions. The input unit 340 receives operation inputs made by the user. The operation inputs are transmitted as command signals to the control unit 310 via the internal bus 390. The control unit 310 can, if necessary, perform predetermined controls and / or calculations based on the transmitted command signals. The input unit 340 may be included in the housing of the client device 110 or it may be external. For example, the input unit 340 may be implemented as a touch panel integrated with the output unit 350. When the input unit 340 is implemented as a touch panel, the user can input tap operations, swipe operations, etc. to the input unit 340. The input unit 340 may be a switch button, mouse, trackpad, keyboard, etc. instead of a touch panel.

[0026] The output unit 350 is, for example, a display unit, which outputs (displays) information as a graphical user interface (GUI) screen that can be operated by the interviewer. The output unit 350 may be included in the housing of the client device 110 or it may be externally attached. More specifically, the output unit 350 may be implemented as a display device such as a liquid crystal display, an organic EL (Electron-Luminescence) display, or a plasma display. It is preferable that these display devices be used and implemented according to the type of client device 110.

[0027] The audio input unit 360 is, for example, an audio input unit represented by a microphone, and is a device that inputs the voice of the interviewee. The audio input unit 360 may be included in the housing of the client device 110 or it may be external. For example, the audio input unit 360 may be implemented as an external headphone with a microphone, integrated with the audio output unit 370.

[0028] The audio output unit 370 is, for example, an audio output unit such as a speaker, which outputs the voice of an avatar, as described later. The audio output unit 370 may be included in the housing of the client device 110 or it may be an external unit.

[0029] The communications unit 380 connects the client device 110 to the network 150 and manages communication with other devices.

[0030] Client devices 120 and 130 have the same hardware configuration as client device 110.

[0031] 3. Information Processing (1) Overview of the process The control unit 210 outputs a question. The question includes a question for detecting the target. The control unit 210 accepts the user's answer to the question. The control unit 210 generates a prompt that includes the question, the answer to the question, and an instruction to detect the target based on the answer. The control unit 210 inputs the prompt into the trained model. The control unit 210 receives output information for the prompt from the trained model. The output information includes the detection result of the target. The detection result includes a value that quantitatively indicates the degree to which the target is detected. The control unit 210 outputs the detection result to an output destination corresponding to the value.

[0032] By performing this process, the target can be detected efficiently and easily, and the detection results can be output to a destination appropriate to the detection result.

[0033] (2) Details of the process Figure 4 is a sequence diagram showing an example of information processing in the interview system 1000. In sequence SQ401, the control unit 210 receives a login request for the interview system 1000 from the client device 110. The login request includes the interviewee's authentication information. The authentication information includes an email address and a password. The email address is an example of an account name. For the sake of simplicity, this disclosure uses an email address as an example of an account name.

[0034] For example, the control unit 210 sends an email or other notification to the interviewee at predetermined intervals (for example, once a month or once every two months) informing them that they will be undergoing an interview using the interview system 1000. When the interviewee selects the URL (Uniform Resource Locator) etc. provided in the email address, a login request is sent from the client device 110 to the server device 100.

[0035] The control unit 210 searches for an account corresponding to the email address using the account information stored in the memory unit 220, etc. Account information refers to a dataset (or record data) used to identify a specific user within the interview system 1000. The dataset includes an email address, a password, and additional user information. Additional user information includes the user's name, registration date and time, profile information (for example, company affiliation information), and past interview record data. Past interview record data includes audio data and text information obtained by transcribing the audio data. The interview record data also includes audio data of questions asked by the AI ​​avatar (described later), text information of the audio data of these questions, audio data of the interviewee's answers to the questions, and text information of the audio data of these answers.

[0036] If the search results indicate that an account exists, the control unit 210 uses the account's password to verify whether the password entered by the interviewee is correct. If the account's password matches the password entered by the interviewee, the control unit 210 considers the password correct and allows login.

[0037] In sequence SQ402, the control unit 210 generates a screen related to the interview with the interviewee (hereinafter also simply referred to as the interview screen) and transmits it to the client device 110. The interview screen is displayed on the output unit 350 by the web browser or the like of the client device 110.

[0038] Figure 5 shows an example of the interview screen 500. The interview screen 500 includes an AI avatar 510, an image of the interviewee 520, a display area 530, and an end button 540. The AI ​​avatar 510 is a digital character created using artificial intelligence (AI) and asks questions to the interviewee. The AI ​​avatar 510 is an example of a digital avatar. The questions from the AI ​​avatar 510 are output as voice. That is, the questions are output as voice via the digital avatar. The image of the interviewee 520 is an image (or video) of the interviewee captured by the imaging unit 330. The display area 530 is an area where the questions from the AI ​​avatar 510 are displayed as text information. The end button 540 is a button that the interviewee selects when the interview with the AI ​​avatar 510 is finished. The control unit 210 controls the display of the interview screen 500, etc.

[0039] Compared to face-to-face interviews with a person, conversing with an AI avatar can reduce emotional tension and provide a more relaxed environment. Furthermore, while human questioning can be influenced by unconscious biases and emotions, questions from an AI avatar are free from such influences. Additionally, interviews with AI avatars can be scheduled at a time convenient for the interviewee, making them efficient and highly scalable, even for companies with many employees.

[0040] In sequence SQ403, the control unit 210 causes the AI ​​avatar 510 to output questions as voice data based on the question list. Here, the question list is a list of questions prepared for each interviewer's company in the storage unit 220 of the server device 100. The question list is stored in the storage unit 220. The control unit 210 may add and / or modify the content of questions to the default question list based on a request from a client device of a person in charge of a designated department (such as the Human Resources Department) in the same organization (e.g., the same company) as the interviewer. The control unit 210 may also add and / or modify the content of the question list based on the history of past interviews. For example, if the history of past interviews records that harassment was detected, the control unit 210 may increase the number of questions related to harassment detection compared to the number of questions in the normal question list, or make the content of the questions related to harassment detection more detailed than the content of the normal questions.

[0041] In sequence SQ404, the control unit 210 receives audio data of the interviewer's responses to questions from the AI ​​avatar 510 from the client device 110. Based on the received audio data of the interviewer's responses, the control unit 210 transcribes the text information of the responses and stores it in the storage unit 220 or the like, associating it with the corresponding questions in the question list. For example, the control unit 210 transcribes the audio data using a trained model that has been trained with audio data as input data and text data as output data. The transcribed responses are used for prompts, which will be described later. The control unit 210 stores the text information of the questions asked to the interviewer, the video recording data of the interview, and the text information of the interviewer's answers in the storage unit 220 or the like as the interview history of the interviewer's (user's) profile information.

[0042] In sequence SQ405, the control unit 210 determines whether to terminate the repetition of sequences SQ403 and SQ404. For example, the control unit 210 determines to terminate the repetition if the end button 540 is selected on the interview screen 500. The control unit 210 also determines to terminate the repetition if, for example, all questions in the question list have been output and all answers to the questions have been received. If the control unit 210 determines to terminate the repetition, it proceeds to sequence SQ406; if it determines not to terminate the repetition, it returns to sequence SQ403. If it returns to sequence SQ403, the control unit 210 causes the AI ​​avatar 510 to output the next question from the question list as audio data, based on the question list.

[0043] In sequence SQ406, the control unit 210 generates a prompt to be input to the large-scale language model. This prompt includes multiple questions for the interviewer, answers to each of the multiple questions, and instructions to detect the detection targets from the answers. The detection targets are harassment and the risk of employee turnover. The prompt further includes instructions to detect the type of harassment if harassment is detected. The prompt includes the types of harassment and examples of harassment for each type. The prompt also includes examples of respondent answers that may lead to employee turnover. The prompt may further include instructions to detect the reasons for employee turnover if the risk of employee turnover is above a threshold.

[0044] Types of harassment include power harassment, sexual harassment, moral harassment, maternity harassment, paternity harassment, and alcohol harassment. Examples of power harassment include "a superior berating a subordinate in front of many people for making a mistake," "intimidating someone by banging on a table or throwing objects," and "assigning someone with an unrealistically short deadline." Examples of sexual harassment include "persistently commenting on someone's clothing or appearance, such as 'your skirt is too short' or 'you have a nice figure'," "making physical contact," "forcing someone to sit next to a superior or a male business associate at a company drinking party," and "sending sexual images or videos through company communication tools." Examples of moral harassment include "making fun of someone by saying, 'You don't even understand something this simple?' when pointing out a mistake," "continuing to ignore someone's greetings and refusing to acknowledge their existence," and "constantly monitoring work progress and putting pressure on them by saying things like, 'You're too slow! You should be able to do more!'" Examples of maternity harassment include "being told that reporting a pregnancy would be a nuisance because it would reduce the workload, and being advised to quit," "being gossiped about for taking breaks due to morning sickness, and "having full-time employee benefits revoked when an employee returns to work after giving birth by saying, 'You're fine as a part-timer,'" and "having full-time employee benefits revoked."

[0045] Responses from respondents that lead to leaving their jobs include, in terms of the types of responses that cause them to leave, responses indicating dissatisfaction with the work environment, responses indicating decreased motivation for work, responses indicating dissatisfaction with work-life balance, and responses indicating dissatisfaction with evaluation or compensation. Examples of responses indicating dissatisfaction with the work environment include "It is difficult to communicate with superiors and colleagues," "I have experienced harassment or unfair treatment," and "I feel that the atmosphere at work is tense." Examples of responses indicating decreased motivation for work include "I do not feel a sense of fulfillment in my work," "I do not feel that I am growing," and "My skills and abilities are not being utilized." Examples of responses indicating dissatisfaction with work-life balance include "There is a lot of overtime and I do not have time for my private life," "I feel that it is difficult to take time off," and "It is difficult to balance work with family and childcare." Examples of responses indicating dissatisfaction with evaluation or compensation include "I feel that my efforts are not being fairly evaluated," "The method for determining the amount of salary and bonuses is unclear," and "There is no prospect of promotion."

[0046] Furthermore, the prompt includes instructions that, if harassment is detected, the system should retrieve information about the person suspected of committing the harassment based on the questions and answers.

[0047] In sequence SQ407, the control unit 210 inputs the generated prompt as input information to the Large Language Model. The Large Language Model is a type of artificial intelligence model for natural language processing that is trained using a vast amount of text data. The Large Language Model may be implemented in another server device that can communicate with the server device 100, or it may be implemented in the server device 100. As another example, instead of a large-scale language model, a multimodal model may be used. A multimodal model is an artificial intelligence (AI) model that integrates and processes multiple different types of data (modalities). These different types of data include text, images, audio, video, and numerical data.

[0048] In sequence SQ408, the control unit 210 receives output information from the large-scale language model.

[0049] Figure 6 shows an example of output information 600. Output information 600 includes at least the detection results for the detected targets. The detected targets are harassment and the risk of employee turnover. The risk of employee turnover is an example of employee turnover risk. The detection results also include values ​​that quantitatively indicate the degree of each of the multiple detected targets.

[0050] In the example of information 610, the quantitative value indicating the degree of harassment is shown to be 6 on a 10-point scale. On a 10-point scale, 10 is the highest and 0 is the lowest. In other words, regarding the quantitative value indicating the degree of harassment, a value closer to 10 indicates a higher degree of harassment, and a value closer to 0 indicates a lower degree of harassment. A quantitative value indicating the degree of harassment of 0 indicates that no harassment was detected.

[0051] In the example of information 620, the quantitative value indicating the degree of employee turnover is shown to be 3 on a 10-point scale. On a 10-point scale, 10 is the highest and 0 is the lowest. In other words, regarding the quantitative value indicating the degree of employee turnover, a value closer to 10 indicates a higher degree of risk of turnover, and a value closer to 0 indicates a lower degree of risk of turnover. In addition, output information 600 includes the reported content as a detection result.

[0052] The report includes the date the interview took place, the location and department to which the interviewer belongs, the type of detected incident to be reported, the interviewer, and, if the detected incident was power harassment, the person suspected of having committed the harassment (power harassment perpetrator 630 in the example in Figure 6), as well as a URL to access detailed information about the interview results (details URL 640 in the example in Figure 6). The control unit 210 determines that the harassment should be reported if any form of harassment is detected (i.e., if the value quantitatively indicating the degree of harassment is 1 or greater).

[0053] Returning to the explanation of Figure 4, in sequence SQ409, the control unit 210 determines the output destination according to a value that quantitatively indicates the degree of the detected object to be reported. In the example in Figure 6, harassment (more specifically, power harassment) is detected as the detected object to be reported, so the control unit 210 determines the output destination (or reporting destination) of the detection result based on a value that quantitatively indicates the degree of harassment (6 / 10, or 6).

[0054] Figure 7 shows an example of setting data 700 in which output destinations are set in stages. The setting data 700 shown in Figure 7 is stored in, for example, the storage unit 220. The setting data 700 in Figure 7 is an example of setting data related to the output destination of harassment. For each detection target, setting data like that shown in Figure 7 is stored in the storage unit 220.

[0055] In the setting data 700 in Figure 7, if the quantitative value indicating the degree of harassment is 0, no report is made; if the quantitative value indicating the degree of harassment is 1 to 3, the Human Resources Department (also simply called Human Resources) is set as the reporting destination (output destination); if the quantitative value indicating the degree of harassment is 4 to 6, the Harassment Committee is set as the reporting destination; and if the quantitative value indicating the degree of harassment is 7 or higher, an external reporting destination is set as the reporting destination. Each reporting destination has its own set of information, such as the recipient's email address. Please note that the personnel department, harassment committee, and external reporting channels are examples of output destinations corresponding to quantitative values ​​indicating the degree of harassment, and are not limited to these.

[0056] The Human Resources Department is the department within a company or other organization that is responsible for matters related to human resources. The Human Resources Department is responsible for all aspects of human resource management within the organization, from employee recruitment to retirement. A Harassment Committee is a specialized committee or group established within a company or other organization to deal with harassment issues (power harassment, sexual harassment, moral harassment, etc.). The official name and composition vary depending on the organization. Harassment Committees are established with the purpose of preventing harassment, providing consultation, investigating, and resolving harassment. An external reporting destination refers to a consultation service or organization established outside the company or other organization. External reporting destinations are often used when victims of harassment cannot resolve the issue internally or when they have distrust in the internal response. Examples of external reporting destinations include the Labor Standards Inspection Office, lawyers or law firms, external professional organizations contracted by the company (e.g., counselors or consultants), external labor unions such as external unions, human rights consultation services at the Legal Affairs Bureau, and NPOs (Non-Profit Organizations) or private support groups that provide support to victims of harassment.

[0057] As shown in Figure 7, having multiple output destinations depending on the severity of the harassment allows for a swift response appropriate to the level of harassment (the magnitude of the harassment problem). Furthermore, having multiple output destinations prevents the processing from becoming concentrated in a single point of contact. The existence of multiple output destinations also indicates the presence of multiple response measures corresponding to the severity of the harassment, sending a message to the entire organization to which the interviewer belongs that "harassment will not be overlooked." This also creates a deterrent against harassment. Moreover, even if minor harassment is overlooked, or if the situation worsens after intervention, the harassment information is sent to a higher-level output destination, reducing the risk of the victim being left unattended.

[0058] The control unit 210 determines, based on the value indicating the degree of harassment and the setting data 700, the output destination corresponding to the value indicating the degree of harassment is the output destination to which the detection result is output. Furthermore, the control unit 210 may also transmit training information to the person identified in the person information (in the example of Figure 6, the perpetrator of power harassment 630), corresponding to a value that quantitatively indicates the content and / or degree of the harassment. Examples of training include discussions using actual harassment cases and simulated cases, workshops to reflect on one's own actions, and a review of the code of conduct and / or workplace rules established by the interviewee's company. The format of the training may include individual training, online training, and training by external instructors. The training information transmitted will include the content of the training, the purpose of the training, the date and time of the training, and the location of the training.

[0059] In sequence SQ410, the control unit 210 transmits output information 600, as shown in Figure 6, to the output destination determined in sequence SQ409. When the person in charge who receives the output information 600 selects the detailed URL 640 in Figure 6, the control unit 210 controls the person in charge's device to display the detailed screen 800 of the interview results for the corresponding interviewer, as shown in Figure 8.

[0060] Figure 8 shows an example of the detailed interview results screen 800. The detailed screen 800 includes a video playback area 810, an AI evaluation display area 820, and an interview text information display area 830. The video playback area 810 is the area where the video recorded during the interview is played and displayed. The AI ​​evaluation display area 820 includes the reasons why the detected object was identified. If the detected object is harassment and power harassment is detected from the questions and answers, the AI ​​evaluation display area 820 will include, for example, what kind of questions were asked, what kind of answers were given, and the reasons why which part of the answers was identified as power harassment.

[0061] The interview text information display area 830 displays text information (text information) that is a transcription of the interview video. In the example in Figure 8, text information 831, text information 833, and text information 835 are text information that is a transcription of the questions asked by the AI ​​avatar 510. Text information 832, text information 834, and text information 836 are text information that is a transcription of the interviewer's answers corresponding to the questions asked by the AI ​​avatar 510. Text information 832 is the text information of the interviewer's answer corresponding to the question from the AI ​​avatar in text information 831. Text information 834 is the text information of the interviewer's answer corresponding to the question from the AI ​​avatar in text information 833. Text information 836 is the text information of the interviewer's answer corresponding to the question from the AI ​​avatar in text information 835.

[0062] By referring to the output information 600 shown in Figure 6 and the detailed interview results screen 800 shown in Figure 8, the person in charge who receives this information can check the questions and answers from the interview, thereby detecting harassment and the risk of employee turnover, and taking appropriate measures to address these risks.

[0063] According to the processing of this embodiment, efficient and simple detection of the target to be detected and output of the detection results can be performed.

[0064] <Note> This embodiment includes the following disclosures.

[0065] (Note 1) It is an interview system, Having at least one control unit, The control unit, Output the question, The aforementioned questions include questions for detecting the target, We accept user responses to the above questions. A prompt is generated that includes the aforementioned question, the answer to the aforementioned question, and an instruction to detect the target from the aforementioned answer. The aforementioned prompt is input to the trained model, The trained model receives the output information for the prompt, The output information includes the detection result of the target to be detected. The detection results include a value that quantitatively indicates the degree of the detected target. The detection result is output to an output destination corresponding to the aforementioned value. Interview system. (Note 2) The interview system described in Appendix 1, The detected object is harassment. Interview system. (Note 3) The interview system described in Appendix 1, The target of detection is the risk of employee turnover. Interview system. (Note 4) The interview system described in Appendix 2, The aforementioned prompt includes examples of harassment. Interview system. (Note 5) The interview system described in Appendix 2, The prompt further includes instructions to detect information about a person suspected of having committed harassment. The aforementioned detection results include personal information about the person suspected of having committed harassment. Interview system. (Note 6) The interview system described in Appendix 5, The control unit, Information regarding harassment-related training will be sent to the person indicated in the aforementioned person information. Interview system. (Note 7) An interview system described in any one of the appendices 1 through 6, The detection result includes the reason why the target was detected. Interview system. (Note 8) The interview system described in Appendix 2, The output destinations corresponding to the aforementioned value include the Human Resources Department, the Harassment Committee, and external reporting destinations. Interview system. (Note 9) An interview system described in any one of the appendices 1 through 8, The control unit, Display a digital avatar, The question is output via the aforementioned digital avatar. Interview system. (Note 10) An information processing method performed by an interview system, Output the question, The aforementioned questions include questions for detecting the target, We accept user responses to the above questions. A prompt is generated that includes the aforementioned question, the answer to the aforementioned question, and an instruction to detect the target from the aforementioned answer. The aforementioned prompt is input to the trained model, The trained model receives the output information for the prompt, The output information includes the detection result of the target to be detected. The detection results include a value that quantitatively indicates the degree of the detected target. The detection result is output to an output destination corresponding to the aforementioned value. Information processing methods. (Note 11) It is a program, Computers, A program to function as an interview system as described in any one of the items from Appendix 1 to Appendix 9.

[0066] Although embodiments have been described above, these are presented as examples and are not intended to limit the scope of the invention. Novel embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. The embodiments are included in the scope and spirit of the invention, as well as in the claims of the invention and its equivalents. [Explanation of symbols]

[0067] 100: Server device 110: Client device 150: Network 210: Control Unit 220: Storage section 230: Communications Department 1000: Interview System

Claims

1. It is an interview system, Having at least one control unit, The control unit, Output the question, The aforementioned questions include questions to detect harassment, We accept user responses to the above questions. A prompt is generated that includes the aforementioned question, an answer to the aforementioned question, and an instruction to detect harassment from the aforementioned answer. The aforementioned prompt is input to the trained model, The trained model receives the output information for the prompt, The output information includes the results of the harassment detection, The detection results include a value that quantitatively indicates the degree of the harassment. An interview system that outputs the detection result to an output destination corresponding to the aforementioned value.

2. The interview system according to claim 1, An interview system in which the aforementioned prompts include examples of harassment.

3. The interview system according to claim 1, The prompt further includes instructions to detect information about a person suspected of having committed harassment. The aforementioned detection results include personal information about individuals suspected of having committed harassment, as part of an interview system.

4. The interview system according to claim 3, The control unit, An interview system that sends information about harassment-related training to the person indicated in the aforementioned person information.

5. The interview system according to claim 1, An interview system in which the detection results include the reason for detecting the harassment.

6. The interview system according to claim 1, The interview system outputs to various destinations based on the aforementioned values, including the Human Resources Department, the Harassment Committee, and external reporting channels.

7. The interview system according to claim 1, The control unit, Display a digital avatar, An interview system that outputs the questions via the aforementioned digital avatar.

8. An information processing method performed by an interview system, Output the question, The aforementioned questions include questions to detect harassment, We accept user responses to the above questions. A prompt is generated that includes the aforementioned question, an answer to the aforementioned question, and an instruction to detect harassment from the aforementioned answer. The aforementioned prompt is input to the trained model, The trained model receives the output information for the prompt, The output information includes the results of the harassment detection, The detection results include a value that quantitatively indicates the degree of the harassment. An information processing method that outputs the detection result to an output destination corresponding to the aforementioned value.

9. It is a program, Computers, A program for functioning as an interview system according to any one of claims 1 to 7.

Images