Processing apparatus, processing method, and processing program
The processing apparatus and method leverage generative AI to analyze health data and engage in empathetic dialogue, effectively addressing female employees' health issues and improving workplace environments.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-09-19
- Publication Date
- 2026-04-01
Smart Images
Figure 2026056322000001_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a processing device, a processing method, and a processing program.
Background Art
[0002] Due to the decrease in the working population caused by the declining birthrate, the importance of health management has been increasing. Among these, health issues specific to women affect work efficiency and job retention, and the economic loss to the entire society is estimated to be 3.4 trillion yen.
[0003] The Ministry of Health, Labour and Welfare and the Ministry of Economy, Trade and Industry are strengthening policies to promote the active participation of women for the purpose of achieving gender equality and improving the competitiveness of enterprises. Companies are obliged to formulate and disclose specific action plans regarding the increase of female management positions and the creation of a work-friendly environment.
[0004] Under such circumstances, companies are required by administrative agencies to improve the workplace environment, such as providing methods for dealing with women's health issues, and the number of companies introducing femtech services as welfare is increasing.
Prior Art Documents
Non-Patent Documents
[0005]
Non-Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0006] Although the need to introduce femtech services is increasing, necessary support may not be provided to employees due to insufficient recognition and utilization of the provided services by employees and insufficient recognition of issues by companies.
[0007] The present invention has been made in view of the above, and aims to provide a processing apparatus, a processing method, and a processing program that can smoothly provide solutions for the health challenges of female users. [Means for solving the problem]
[0008] To solve the above-mentioned problems and achieve the objective, the present invention provides a processing apparatus comprising: a first acquisition unit that acquires health management information indicating the past health status of a female user as past quantitative information of the female user; a second acquisition unit that acquires past subjective information of the female user; an analysis unit that estimates causal relationships between the information based on the past quantitative information and the past subjective information, and predicts the state of the female user's malaise, candidate causes of the malaise, and the timing based on the causal relationships; a third acquisition unit that acquires the female user's current subjective information relating to at least the prediction results by the analysis unit, using a generative AI (Artificial Intelligence) model that generates text corresponding to the information input by the female user and engages in dialogue with the female user; and a presentation unit that narrows down the cause of the female user's malaise from the candidate causes based on the current subjective information, and presents one or more countermeasures for the narrowed-down causes to the female user's terminal.
[0009] Furthermore, the processing method of the present invention is a processing method executed by a processing device, and is characterized by including the steps of: acquiring health management information indicating the past health status of a female user as past quantitative information of the female user; acquiring past subjective information of the female user; estimating causal relationships between the information based on the past quantitative information and the past subjective information, and predicting the state of the female user's malaise, candidate causes of the malaise, and the timing based on the causal relationships; acquiring the female user's current subjective information relating to the prediction results in at least the prediction step, using a generative AI model that generates text corresponding to the information input by the female user and engages in dialogue with the female user; and narrowing down the causes of the female user's malaise from the candidate causes based on the current subjective information, and presenting one or more countermeasures for the narrowed-down causes to the female user's terminal.
[0010] Furthermore, the processing program of the present invention causes a computer to perform the following steps: acquiring health management information indicating the past health status of a female user as past quantitative information of the female user; acquiring past subjective information of the female user; estimating causal relationships between the information based on the past quantitative information and the past subjective information, and predicting the state of the female user's malaise, candidate causes of the malaise, and the timing based on the causal relationships; acquiring the female user's current subjective information regarding the prediction results in at least the prediction step, using a generative AI model that generates text corresponding to the information input by the female user and engages in dialogue with the female user; and narrowing down the causes of the female user's malaise from the candidate causes based on the current subjective information, and presenting one or more countermeasures for the narrowed-down causes to the female user's terminal. [Effects of the Invention]
[0011] According to the present invention, it is possible to smoothly provide female users with a way to address their health issues. [Brief explanation of the drawing]
[0012] [Figure 1] FIG. 1 is a diagram showing an overview of the processing of the processing system according to the embodiment. [Figure 2] FIG. 2 is a diagram showing an overview of the processing of the processing system according to the embodiment. [Figure 3] FIG. 3 is a diagram showing an example of the configuration of the processing system according to the embodiment. [Figure 4] FIG. 4 is a diagram showing an example of the configuration of the support server shown in FIG. 3. [Figure 5] FIG. 5 is a sequence diagram showing the processing procedure of the support processing according to the embodiment. [Figure 6] FIG. 6 is a diagram for explaining the input / output information of the generative AI. [Figure 7] FIG. 7 is a diagram for explaining the overview of the contents of the questionnaire. [Figure 8] FIG. 8 is a diagram for explaining the input data to the generative AI. [Figure 9] FIG. 9 is a diagram showing an example of the prompt set for the generative AI. [Figure 10] FIG. 10 is a diagram showing an example of the prompt set for the generative AI. [Figure 11] FIG. 11 is a diagram showing an example of the report generated by the generative AI. [Figure 12] FIG. 12 is a diagram showing an example of the report generated by the generative AI. [Figure 13] FIG. 13 is a diagram showing an example of the report generated by the generative AI. [Figure 14] FIG. 14 is a diagram showing an example of the report generated by the generative AI. [Figure 15] FIG. 15 is a diagram for explaining the input / output information of the generative AI. [Figure 16] FIG. 16 is a diagram showing an example of the screen of the user terminal. [Figure 17] FIG. 17 is a diagram showing an example of the prompt set for the generative AI. [Figure 18] FIG. 18 is a diagram showing an example of the screen of the user terminal. [Figure 19]FIG. 19 is a diagram showing an example of a chat on a user terminal. [Figure 20] FIG. 20 is a diagram showing an example of a screen of a user terminal. [Figure 21] FIG. 21 is a diagram showing an example of a screen of an enterprise terminal. [Figure 22] FIG. 22 is a diagram showing an example of a screen of an enterprise terminal. [Figure 23] FIG. 23 is a diagram showing an example of a screen of an enterprise terminal. [Figure 24] FIG. 24 is a diagram showing an example of a screen of an enterprise terminal. [Figure 25] FIG. 25 is a diagram showing an example of a screen of an enterprise terminal. [Figure 26] FIG. 26 is a diagram showing an example of a screen of an enterprise terminal. [Figure 27] FIG. 27 is a diagram showing an example of a screen of an enterprise terminal. [Figure 28] FIG. 28 is a diagram showing an example of a screen of an enterprise terminal. [Figure 29] FIG. 29 is a diagram showing an example of a computer in which a support server and various terminals are realized by executing a program.
Embodiments for Carrying Out the Invention
[0013] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited by this embodiment. Also, in the description of the drawings, the same parts are denoted by the same reference numerals.
[0014] [Embodiment] First, let's describe the embodiment. In this embodiment, based on health management information specific to women and a survey of female employees (female users) regarding health issues specific to women and working styles for women (e.g., policies for women's advancement, promotion of women's advancement), the system uses a generative AI (Artificial Intelligence) (generative AI model) capable of engaging in empathetic conversations with female employees to gather their inner thoughts, narrow down the causes of their ailments, and propose solutions. In this way, the embodiment smoothly provides solutions to the health issues of female employees.
[0015] In this embodiment, the system gathers the inner thoughts of female employees that cannot be collected through surveys via chat, extracts information that female employees want from the company, and provides it to the company, thereby supporting the company in planning measures and introducing services to create a workplace where female employees can work comfortably.
[0016] [Summary of the Embodiment] Figures 1 and 2 show an overview of the processing system according to the embodiment. As shown in Figures 1 and 2, in the embodiment, a welfare service is provided to each female employee's user terminal, presenting solutions to their individual problems as a curation site or application (app).
[0017] In this embodiment, the processing system is linked to personal health records (PHRs) for women via curation sites and apps. Examples of PHRs include menstrual cycle management apps P1, fertility support apps P2, health management apps P3, etc., as well as Femtech-related services (femcare recommendation service W1, lifestyle improvement recommendation service W2, online dispensing and prescription W3, online medical consultation and treatment W4).
[0018] For example, on the company side that is supported by the processing system, the welfare officer H1 sets the task of determining what should be done to promote the advancement of women (Figure 2(A)).
[0019] The processing system's support server (processing unit) collects information about female employees (UA) (Figure 2(B)). The support server takes as input female employee UA's health management information specific to women from the menstrual cycle management app P1, fertility support app P2, and health management app P3, as well as the female employee UA's responses to the questionnaire survey 121A regarding women's specific health issues and women's work styles, and past chat history with the generating AI (Artificial Intelligence) (Figure 1(A)). Alternatively, the female employee UA may conduct a questionnaire survey on women's empowerment measures via the generating AI chat 134A on the terminal used by the female employee UA (for example, user terminals 30-1, 30-2) (Figure 2(B-1)).
[0020] Next, the support server uses the generated AI chat 134A with the female employee UA to delve deeper into the survey responses and collect current information (subjective information) about the female employee UA that could not be obtained in the survey 121A. During this process, the support server collects information while engaging in a conversation with the female employee UA through the generated AI chat 134A. The generated AI chat 134A collects current subjective information about the female employee UA, for example, regarding career aspirations, health concerns, lifestyle concerns, dissatisfaction with systems and policies, requests, and literacy issues (Figure 2 (B-2)).
[0021] The support server analyzes and extracts the causes of female employees' health issues, lifestyle issues, dissatisfaction with systems and policies, career aspirations, and / or workplace dissatisfaction (literacy, work environment) based on current subjective information collected from multiple female employees using the generated AI chat 134A (Figure 2(C)), and creates an anonymized report R1 (Figure 1(F), Figure 2(C-1)).
[0022] The support server creates visualized data of the survey's statistical results and Report R1, and sends this visualized data to Welfare Officer H1 at the company where the female employee works, to support policy planning when considering solutions (Figure 2(D), (D-1)).
[0023] Welfare Officer H1 can extract and analyze hidden issues among female employees and their satisfaction with various systems, etc., based on the questionnaire survey and report R1 (Figure 1 (G)). Welfare Officer H1 can then formulate new measures in response to the identified issues, and introduce new services to female employees UA (Figure 2 (E), (E-1)).
[0024] The support server then analyzes the state of the female employee UA's malfunction, potential causes, and timing based on the input information, visualizes the analysis results, and provides them to the female employee UA's user terminal (Figure 1(B)).
[0025] The support server receives inquiries from female employee UA via the generated AI chat 134A (Figure 1(C)), and based on the chat conversation (current subjective information), narrows down the cause of the female employee UA's discomfort from the possible causes and proposes countermeasures (Figure 1(D)). As countermeasures, the support server presents, for example, newly introduced Femtech-related services (Femcare Recommendation Service W1, Lifestyle Improvement Recommendation Service W2, Online Dispensing / Prescription W3, Online Medical Consultation / Treatment W4). The support server connects the female employee UA's user terminal to each service so that the female employee UA can use each of the presented services.
[0026] The support server then uses the generated AI chat 134B to collect current subjective information from multiple female employees about the implementation of the measures and services in order to verify their effectiveness (Figure 2(F)).
[0027] The support server collects the female employee UA's opinions (subjective information) regarding countermeasures based on the dialogue between the female employee UA and the generated AI chat 134B, and confirms the effectiveness of the countermeasures, measures, and services (Figure 1 (E)). The generated AI chat 134B collects information such as the usability of the system, impressions of the services provided, and concerns, and provides the female employee UA with information on countermeasures (videos, columns, etc.) and countermeasures (products, prescription drugs, etc.) (Figure 2 (F-1)).
[0028] The support server creates Report R2 from the current subjective information of multiple female employees collected (Figure 2 (F-2)), and then creates an anonymized report (Figure 1 (F)). Welfare officer H1 can refer to this Report R2 to understand the effects of introducing new measures and services (Figure 2 (F-3)), and efficiently proceed with planning the next measures and introducing services.
[0029] The support server may provide the statistical results of the survey and the visualized data of the report to third-party companies that provide various products and services for women. By providing the report to various service companies, these service companies can target corporate welfare managers, diversity promotion managers, etc., and propose the adoption of welfare services that are better suited to the challenges faced by female employees.
[0030] [Processing System] Next, the configuration of the processing system will be described. Figure 3 is a diagram showing an example of the configuration of the processing system according to the embodiment.
[0031] As shown in Figure 3, the processing system according to this embodiment has a support server 10 provided on a platform that provides support services. The support server 10 communicates with the corporate terminal 20 of company A, which is the target of the service, and with the user terminals 30-1 and 30-2 of female employees working at company A. The support server 10 also communicates with the generation AI server 200.
[0032] The Generative AI Server 200 is equipped with Generative AI 210 (Generative AI Model), a large-scale natural language processing model. Generative AI 210 processes input text data according to set prompts, creates text, and outputs it. For example, Generative AI 210 interacts with female employees by generating and outputting text corresponding to information entered by them. In addition, Generative AI 210 processes input text data according to set prompts, creates a summary (report), and outputs it.
[0033] The support server 10 sets a prompt for the generation AI 210 and inputs the text data (or transcript data of audio data) entered into user terminals 30-1 and 30-2 into the generation AI 210. The support server 10 sends the text data output by the generation AI 210 to user terminals 30-1 and 30-2, causing them to display it in the chat. In this way, the support server 10 uses the generation AI 210 to conduct conversations with female employees, capturing their inner thoughts and feelings that cannot be gathered through surveys.
[0034] The support server 10 narrows down the cause of the female employee's distress and proposes solutions through a chat that uses empathetic conversations with the female employee. Then, the support server 10 creates a report from the chat history that extracts the information the female employee wants from the company and sends it to the company terminal 20.
[0035] Corporate terminals 20 include, for example, PCs (Personal Computers), notebook PCs, and tablet devices. User terminals 30-1 and 30-2 include, for example, PCs, notebook PCs, tablet devices, and smartphones. When referring to user terminals 30-1 and 30-2 collectively, they are referred to as user terminal 30.
[0036] [Generation Server] Figure 4 shows an example of the configuration of the support server 10 shown in Figure 3. The support server 10 includes, for example, a communication unit 11, a storage unit 12, and a control unit 13.
[0037] The communication unit 11 controls communications related to various types of information. For example, the communication unit 11 controls communications between the generation AI server 200, the enterprise terminal 20, and the user terminal 30.
[0038] The memory unit 12 stores data and programs necessary for various processes performed by the control unit 13. For example, the memory unit 12 may be a semiconductor memory element such as RAM (Random Access Memory) or flash memory, or a storage device such as a hard disk or optical disc. The memory unit 12 contains survey results 121, user information 122, health management information 123, chat history 124, analysis results 125, prompt data 126, report 127, countermeasure information 128, and visualization data 129.
[0039] The survey results 121 consist of questions and answers from a questionnaire administered to each female employee regarding health issues specific to women and working styles for women (e.g., policies for promoting women's advancement, and the promotion of women's advancement). The survey results 121 may be received from the company terminal 20, or collected via user terminals 30 used by each female employee.
[0040] User information 122 is information about each female employee, including, for example, the female employee's age, address, family structure, etc.
[0041] Health management information 123 is information about the health of each female employee, and is obtained through, for example, a menstrual cycle management app P1, a fertility support app P2, a health management app P3, etc., installed on the user terminal 30. For example, health management information 123 includes menstrual cycle information such as menstrual cycle (including start date), basal body temperature, and mood (for example, symptoms of mental and physical instability due to premenstrual syndrome (PMS)). Health management information 123 also includes fertility information such as timing of hospital visits, purpose of hospital visits, treatments and medications, and health management information such as steps taken, exercise, distance traveled, vital information such as heart rate, sleep information, dietary history, and headache status.
[0042] Chat history 124 is the chat history of each female employee with the chatbot provided by the support server 10. Chat history 124 is a combination of greetings, questions, solutions, etc. output by the generating AI 210 and the text entered by the female employee in response. Since the text entered by the female employee is based on the female employee's subjective opinion, chat history 124 is one of the female employee's past subjective information. Subjective information is the results of the female employee's responses to a questionnaire on women's specific health issues and women's working styles, and / or the chat history of female employees on women's specific health issues and women's working styles.
[0043] Analysis result 125 includes the analysis results from analysis department 133 (described later). Analysis result 125 includes the state of health problems among female employees, as well as predictions of possible causes and timing of these problems.
[0044] The prompt data 126 consists of each prompt to be set in the generating AI 210. For example, it may include prompts to instruct the AI to conduct a more in-depth survey of female employees, or prompts to instruct the AI to create a report for the company.
[0045] Report 127 is a report created by the generating AI 210 under the direction of the report creation unit 136 (described below). Report 127 includes a report summarizing the health issues, lifestyle issues, dissatisfaction with systems and policies, career aspirations, and / or workplace dissatisfaction (literacy, work environment) of female employees.
[0046] Coping Strategy Information 128 includes various coping strategies that can be adopted when female employees experience discomfort. For example, Coping Strategy Information 128 provides information that, for each type of discomfort, corresponds to lifestyle improvements such as drinking a warm caffeine-free beverage before bedtime or going to bed earlier, examples of recommended medications, and recommendations for medical consultations and the appropriate medical departments to visit. Coping Strategy Information 128 is based on content supervised by medical professionals, for example.
[0047] Visualization data 129 includes visualization data created by the visualization unit 137 (described later). Visualization data includes, for example, the statistical results of the questionnaire survey 121 and the report 127.
[0048] The control unit 13 has an internal memory for storing programs that define various processing procedures and required data, and executes various processes using these. Here, the control unit 13 may be, for example, an electronic circuit such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit), or an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).
[0049] The control unit 13 includes a questionnaire data collection unit 131 (second acquisition unit), a health management information acquisition unit 132 (first acquisition unit), an analysis unit 133, a chat control unit 134 (third acquisition unit), a coping method suggestion unit 135 (suggestion unit), a report creation unit 136 (creation unit), and a visualization unit 137.
[0050] The survey data collection unit 131 collects the results of surveys conducted by female employees (response data from the surveys). The surveys concern health issues specific to women and women's work styles. For example, the surveys are conducted via the web or through a chatbot using the generation AI 210, and the survey responses are aggregated on the corporate terminal 20 with the permission of each female employee.
[0051] The health management information acquisition unit 132 acquires health management information showing the past health status of each female employee. The health management information consists of quantitative information on the past health status of the female employees.
[0052] The health management information acquisition unit 132 acquires health management information for each female employee via the menstrual cycle management app P1, fertility support app P2, health management app P3, etc., installed on the user terminal 30. For example, health management information 123 includes menstrual cycle information such as menstrual cycle (including start date), basal body temperature, and mood (PMS symptoms, etc.). Health management information 123 also includes fertility information such as timing of hospital visits, purpose of hospital visits, treatment / medication details, etc. Health management information also acquires vital information such as steps taken, exercise, distance traveled, heart rate, sleep information, dietary history, and headache status. The health management information acquisition unit 132 also acquires time-series data of atmospheric pressure from weather information.
[0053] The analysis unit 133 estimates causal relationships between pieces of information based on past quantitative and subjective information of the female employee being supported, and predicts the state of the female employee's poor health, as well as potential causes and timing of the poor health, based on these causal relationships. The analysis unit 133 then transmits the predicted state of the female employee's poor health, potential causes, and timing to the user terminal 30.
[0054] The analysis unit 133 estimates the causal relationships between weather conditions, vital signs, and female-specific cycles (e.g., menstrual cycle and PMS period) related to the female employee's health problems, based on her past health management information and past chat history. Based on the estimated causal relationships, the analysis unit 133 predicts when the female employee's next health problems will occur, the state of the health problems that will occur, and potential causes of the health problems. The analysis unit 133 may also perform processing using a trained machine learning model or a trained machine learning model that predicts the state of the female employee's health problems, potential causes, and timing based on causal relationships.
[0055] The chat control unit 134 controls the chat with the female employee via a chatbot using the generation AI 210. The chat control unit 134 obtains the female employee's current subjective information, at least regarding the prediction results, via the chatbot. The chat control unit 134 instructs the generation AI 210 to engage in a dialogue with the female employee, listening to her current concerns and instructing it to narrow down the possible causes of her discomfort.
[0056] Along with this, the chat control unit 134 instructs the generation AI 210 to perform a conversation that is empathetic to the woman. The chat control unit 134 causes the generation AI 210 to generate text that is empathetic to the woman and includes words that show empathy (for example, "I understand how you feel," "That's right," etc.).
[0057] Furthermore, the chat control unit 134 instructs the generating AI 210 to change the tone of the conversation depending on the female employee's physical condition. For example, the chat control unit 134 instructs the generating AI 210 to make the endings of the conversation gentler for female employees who have menstrual cramps or headaches, or female employees with PMS.
[0058] The chat control unit 134 changes the timing of the response from the generating AI 210 depending on the content. For example, if a female employee enters a question of a predetermined type, the chat control unit 134 does not return an answer immediately, but instead makes the generating AI 210 respond slowly, for example, after 2 seconds, so as not to rush the female employee's chat input.
[0059] Furthermore, the chat control unit 134 instructs the generating AI 210 to engage in conversations with female employees that inquire about health issues specific to women and concerns related to women's work styles. For example, the chat control unit 134 instructs the generating AI 210 to ask questions about health issues, lifestyle issues, dissatisfaction with systems and policies, career aspirations, and / or workplace dissatisfaction (literacy, work environment). The chat control unit 134 then acquires the female employees' dissatisfaction and concerns in response to these questions as their current subjective information. The chat control unit 134 also acquires chat histories from multiple female employees regarding health issues specific to women and women's work styles.
[0060] The troubleshooting unit 135 narrows down the cause of the female employee's malaise from the list of possible causes. For this narrowing down, the troubleshooting unit 135 uses the female employee's current subjective information obtained through chat using the generating AI 210.
[0061] The solution suggestion unit 135 presents one or more solutions for the narrowed-down cause to the female employee's user terminal 30. The solution suggestion unit 135 presents one or more solutions to the female employee's user terminal 30 in a state that allows it to link with service providers corresponding to the solutions for the narrowed-down cause. The solution suggestion unit 135 proposes solutions that are easily accepted by the female employee herself, based on subjective information such as how she feels, through a chatbot conversation using the generation AI 210.
[0062] The solution suggestion unit 135, for example, based on the current subjective information of the female employee UA obtained through chat, narrows down the cause of the female employee UA's discomfort from the candidate causes and sets a prompt instructing the generation AI 210 to present one or more solutions for the narrowed-down cause. The solution suggestion unit 135 may, for example, extract one or more solutions from the solutions information 128 that correspond to the narrowed-down cause and present them to the female employee's user terminal 30. The solution suggestion unit 135 may estimate one or more solutions for the cause using a trained machine learning model.
[0063] The report creation unit 136 instructs the generation AI 210 to create a report summarizing health issues, lifestyle issues, dissatisfaction with systems and policies, career aspirations, and / or workplace dissatisfaction (literacy, work environment) based on the current subjective information of multiple female employees obtained by the chat control unit 134. The report creation unit 136 instructs the generation AI 210 to create a report that shows hidden issues, necessary countermeasures, and solutions regarding health issues, lifestyle issues, dissatisfaction with systems and policies, career aspirations, and / or workplace dissatisfaction.
[0064] The visualization unit 137 creates visualization data of the survey's statistical results and reports, and transmits the created visualization data to the company terminal 20 of the company where the female employees work. The visualization unit 137 changes the data to be visualized according to the filtering of the company terminal 20 and displays it on the company terminal 20. In the company, for example, the welfare officer H1 can refer to the visualization data to efficiently plan policies and consider new services.
[0065] [Processing Procedure] Next, the processing procedure for the support process according to the embodiment will be described. Figure 5 is a sequence diagram showing the processing procedure for the support process according to the embodiment. In Figure 5, the case in which the target of the proposed solution is a female employee UA using user terminal 30-1 will be explained as an example.
[0066] The support server 10 receives the results of a questionnaire submitted by a female employee UA from the corporate terminal 20 (step S11). The support server 10 also receives health management information (past quantitative information) indicating the past health status of the female employee UA from the user terminal 30-1 used by the female employee UA (step S12).
[0067] The support server 10 accesses the past chat history (past subjective information) of the female employee UA from the memory unit 12 (step S13).
[0068] The support server 10 estimates the causal relationships between the pieces of information based on the female employee UA's past health management information and past chat history, and performs analysis processing to predict the female employee UA's state of ill health, as well as potential causes and timing of the ill health, based on the causal relationships (step S14).
[0069] The support server 10, and the analysis unit 133, transmit the predicted state of the female employee's illness, as well as the possible causes and timing of the illness, to the user terminal 30-1 (step S15), and display it on the user terminal 30-1.
[0070] When the support server 10 receives a chat start request from the user terminal 30-1 (step S16), it creates a prompt (step S17).
[0071] The support server 10 sets a prompt in the generating AI 210 of the generating AI server 200, inputting the attributes, past health management information, weather information, and past chat history of the female employee UA (step S18), and starts a chat with the user terminal 30-1 (steps S19-1, S19-2). At this time, the support server 10 sets a prompt in the generating AI 210 that instructs the AI to narrow down the possible causes of the female employee UA's ailment while listening to her concerns. In the prompt, the support server 10 instructs the female employee to engage in conversation that asks about health issues specific to women and concerns about women's work styles. The support server 10 records the history of this chat as the female employee UA's current subjective information.
[0072] Based on the current subjective information of the female employee UA obtained through chat, the support server 10 narrows down the cause of the female employee UA's discomfort from the list of possible causes and sets a prompt instructing the generation AI 210 to present one or more solutions for the narrowed-down cause. The support server 10 receives one or more solutions presented by the generation AI 210 in response to this prompt from the generation AI server 200 (step S20-1) and sends them to the user terminal 30-1 (step S20-2).
[0073] The support server 10 instructs the generating AI 210 to create a report summarizing the health issues, lifestyle issues, dissatisfaction with systems and policies, career views, and / or workplace dissatisfaction based on the current subjective information of multiple female employees (step S21). In response, the generating AI 210 creates the report (step S22), and the generating AI server 200 sends the report to the support server 10 (step S23).
[0074] The support server 10 creates visualization data of the survey statistics and report (step S24). The support server 10 sends the created visualization data to the corporate terminal 20 (step S25) and displays it (step S26).
[0075] Furthermore, the support server 10 may instruct the generating AI 210 to conduct a chat with the user terminal 30-1 in order to collect current subjective information from multiple female employees regarding the introduction of the measures or services, in order to confirm the effectiveness of the introduction of the measures or services (steps S19-1, S19-2). Then, based on the dialogue between the female employee UA and the generating AI 210, the support server 10 instructs the generating AI 210 to create a report summarizing the current subjective information from multiple female employees regarding the introduction of the measures or services (steps S21, S22). The report created by the generating AI 210 is visualized on the support server 10, for example (step S24), and sent to the corporate terminal 20 (step S25).
[0076] [Generating AI Input / Output] Next, we will explain the input and output of the generating AI210. Figure 6 is a diagram illustrating the input and output information of the generating AI210. Figure 6 explains the case where the generating AI210 outputs a report.
[0077] To initiate a chat with the employee UA, a prompt is set in the generating AI210 (Figure 6(1)). The generating AI210 is then populated with the employee UA's survey responses 121-1, chat history 124-1, attributes, health management information (menstrual cycle information, fertility / infertility information, vital signs information, sleep information, dietary history, headache status, etc.), and weather information (time-series data of atmospheric pressure) (Figure 6(2)).
[0078] When a chat takes place between the female employee UA's user terminal 30-1 and the generating AI 210 (Figure 6 (3)), the chat history is accumulated. The generating AI 210 then creates a report 127-1 summarizing the female employee's health issues, lifestyle issues, dissatisfaction with systems and policies, career aspirations, and / or workplace dissatisfaction (Figure 6 (4)), and sends it to the support server 10.
[0079] Figure 7 is a diagram illustrating the general content of the questionnaire. For example, the questionnaire is about promoting women's advancement and includes questions about attributes, attitudes towards careers, health issues specific to women, workplace atmosphere, things that women have given up due to health issues specific to women, company policies and systems, and what is felt is necessary to promote women's advancement.
[0080] Figure 8 illustrates the input data for the generating AI 210. The generating AI 210 receives survey responses and past chat conversation data from the support server 10. The generating AI 210 is then instructed by the support server 10 to ask various in-depth questions, as exemplified in Figure 8, while providing support for the survey responses. The questions, as exemplified in Figure 8, ask for details about health issues, lifestyle issues, systems and policies, views on careers, and dissatisfaction with the workplace.
[0081] Figure 9 shows an example of a prompt set in the generating AI 210. Prompt 126A in Figure 9 instructs the AI to ask follow-up questions while being attentive to the answers to the questionnaire.
[0082] Prompt 126A instructs you to create a character that is approachable, to use polite language to convey trustworthiness, and to conduct a survey in the chat first. Prompt 126A instructs you to ask follow-up questions about the survey answers, to repeat the answers and include words of sympathy (e.g., "That's tough," "That must be hard"), and to ask further questions. Prompt 126A instructs you to end with words of appreciation for the answers, such as "Thank you for teaching me so much." Prompt 126A instructs you to shorten long sentences by dividing them into speech bubbles for easier reading, and to slightly delay the appearance of speech bubbles so as not to rush the responses.
[0083] Figure 10 shows an example of a prompt set in the generating AI 210. Prompt 126B shown in Figure 10 instructs the AI to create a report based on subjective information from female employees obtained in the chat.
[0084] Prompt 126B instructs Generator AI 210 to act as a specialist consultant in the field of human resources and welfare. Based on the survey results and in-depth responses from chat, it instructs AI 210 to extract information from all respondents regarding their health concerns, lifestyle challenges, dissatisfaction with systems and policies, areas for improvement, career aspirations, and workplace dissatisfaction (literacy, work style, and work environment), and to present this information concisely in bullet points. Furthermore, it instructs AI 210 to present a report outlining hidden issues, necessary countermeasures, and specific solutions for each segment.
[0085] Figures 11 to 14 show examples of reports generated by the generation AI 210.
[0086] The report in Figure 11 outlines the hidden challenges, necessary measures, and specific solutions regarding the health issues of female employees. Proposed necessary measures and specific solutions include workplace health management and support systems, development of a workplace environment that considers women's health, raising awareness of and making accessible menstrual leave and infertility treatment leave systems, flexible work arrangements such as telecommuting, emergency measures such as providing free sanitary napkins in restrooms, and development of health management programs in collaboration with doctors.
[0087] The report in Figure 12 outlines hidden challenges, necessary measures, and specific solutions regarding work-life balance support initiatives. The report in Figure 12 identifies hidden challenges such as reduced working hours impacting performance evaluations, leading to a decline in evaluations, the need for a change in mindset among men, concerns about appointing female managers to meet numerical targets, and insufficient awareness of support systems.
[0088] The report in Figure 12 indicates that necessary measures include revising the evaluation system to ensure fair evaluation even for employees working reduced hours, implementing measures to change the mindset of male employees, and providing support systems in an easily understandable way for new and young employees. The report in Figure 12 indicates that specific solutions include clarifying evaluation criteria, improving the transparency of the evaluation system so that it applies to employees working reduced hours, encouraging male employees to take childcare leave, and creating a centralized and easily accessible portal for information on support systems.
[0089] Figure 13 shows the hidden challenges, necessary measures, and specific solutions regarding the health issues of female employees. Figure 14 shows the hidden challenges, necessary measures, and specific solutions regarding work-life balance support initiatives.
[0090] By reviewing the reports in Figures 11 and 13, Company A's Welfare Officer H1 can consider measures to address and resolve health issues among female employees, such as encouraging teleworking, improving the company environment to make leave systems more accessible and easier to use, and introducing health management services in collaboration with doctors.
[0091] Furthermore, by reviewing the reports in Figures 12 and 14, Company A's Welfare Officer H1 can consider measures and solutions to achieve work-life balance support, including improving fairness in evaluations of employees working reduced hours, developing measures to change the mindset of male employees, introducing training programs and consultation services for female managers, and introducing a centralized portal service for information on support systems.
[0092] Figure 15 is a diagram illustrating the input and output information of the generating AI 210. Figure 15 illustrates the case where the generating AI 210 outputs a solution to the user terminal 30-1 of a female employee UA.
[0093] The support server 10 receives input from female employee UA, including questionnaire responses 121-1, chat history 124-1, attributes, health management information (menstrual cycle management information, fertility / infertility information, vital signs information, sleep information, dietary history, headache status, etc.), and weather information (time-series data of atmospheric pressure) (Figure 15 (A)). Based on this input information, the support server 10 estimates the causal relationships between the information and, based on these causal relationships, predicts the state of female employee UA's malaise, potential causes of malaise, and the timing of those malaise (Figure 15 (1)).
[0094] The support server 10 displays the visualization data of the analysis results on the user terminal 30-1 of the female employee UA (Figure 15 (2)). The visualization data includes the timing of the onset of the disorder, the state of the disorder that will occur, and the candidate causes.
[0095] Figures 16 and 18 show examples of the screen of user terminal 30-1. As shown in Figure 16, the screen of user terminal 30-1 displays time-series data T11 of health forecast, weather, atmospheric pressure, schedule, and hormone balance, as well as a comment C11 that reads, "Headache and dizziness are forecast for the 16th. Let's prepare now. I'm here to help you decide what to do."
[0096] By checking this screen, female employee UA can confirm that on September 16th, she is expected to experience physical discomfort such as headaches and dizziness due to changes in atmospheric pressure and hormonal balance. Then, when female employee UA selects the chat start button B11 on the generating AI210, the screen on user terminal 30-1 transitions to screen M12 (Figure 18), allowing her to start chatting with the generating AI210.
[0097] Returning to Figure 15, let's continue the explanation. In response to the female employee UA's instruction to start a chat with the generating AI 210, the support server 10 instructs the generating AI 210 to engage in a conversation that narrows down the predicted cause candidates while providing empathetic support (Figure 15 (3)).
[0098] Figure 17 shows an example of a prompt set on the generating AI 210. The support server 10 sets the prompt 126B, as illustrated in Figure 17, on the generating AI 210 in order to have the generating AI 210 perform such an interaction.
[0099] Prompt 126B instructs the user to delve deeper into the current mood (stress), plans for the next few days, things that might be problematic if discomfort occurs, and the coping methods used so far, their effectiveness, and impressions (frame 210-1 in Figure 15), while engaging in conversation based on the predicted state, cause, and timing of the support server 10. Prompt 126B instructs the user to select appropriate coping methods based on the predicted state, cause, timing, and the delved-in content, and to list x of the most likely ones.
[0100] The generating AI 210 then narrows down the cause of the female employee UA's malaise from the list of possible causes through conversation with the female employee UA. The generating AI 210 then creates a list of possible solutions for the narrowed-down causes, for example by referring to the solution information 128 (Figure 15 (4)). The support server 10 then adds more detailed information to the prediction results and proposes the optimal solution to the female employee UA (Figure 15 (5)).
[0101] For example, the support server 10 displays on the user terminal 30-1 via chat an example of self-care by the female employee UA herself, or a screen guiding her to seek medical attention such as online consultations.
[0102] Figure 19 shows an example of a chat on user terminal 30-1. As shown in Figure 19, the generating AI 210 (AI) sends a message of support to the female employee UA and suggests self-care (messages P11-P13 in Figure 19). The generating AI 210 suggests going to bed early, corresponding to the female employee UA's schedule for September 16 (message P13).
[0103] Next, female employee UA enters question Q14 about what to do when she has trouble sleeping. Generator AI210 sends message P15 to user terminal 30-1 in a gentle tone, suggesting what to eat and drink before bed.
[0104] At the same time, the generating AI210 suggests recommended beverages to drink before bedtime (Message P16). Clicking on the product description in Message P16 will take you to the partner's e-commerce site. This allows female employees (UA) to easily purchase recommended products.
[0105] Furthermore, the generating AI210 also suggests ways to regulate the autonomic nervous system (Message P17), and introduces a column that explains the details. Clicking on "Column" in Message P17 will take you to this column. Therefore, female employee UA can learn more about ways to regulate the autonomic nervous system.
[0106] The generating AI 210 may refer to the chat history of other female employees and present one or more methods that other female employees with similar symptoms to female employee UA have actually used. Furthermore, if female employee UA's symptoms are severe, the generating AI 210 may display a message on the user terminal 30-1 recommending that she seek medical attention rather than self-care, and may direct her to an online medical consultation site.
[0107] In addition to providing specific solutions, the support server 10 may also display messages about health challenges faced by female employees with similar career paths to the female employee UA. The support server 10 may also display messages specifically about the career paths female employees should take based on their menstrual symptoms. The support server 10 may also display messages about the life events (marriage, pregnancy, childbirth) of other female employees and how they are progressing in their careers. By providing such messages to the female employee UA, the support server 10 enables her to learn about the situations of other female employees and to resolve her own questions and anxieties.
[0108] Furthermore, the support server 10 may display a curation site specifically for female employees on the screen of the PC (user terminal 30-2) used by the female employee UA, providing the female employee UA with flexible options for addressing their health issues.
[0109] Figure 20 shows an example of the screen of user terminal 30-2. As shown in Figure 20, from screen M13, users can proceed to answer a questionnaire about women's working styles. When a female employee selects the chat start button A13 on the generating AI 210, the chat screen C13 is displayed, and she can start chatting with the generating AI 210 (Figure 20 (1)). By engaging in empathetic conversation on chat screen C13, the generating AI 210 narrows down the possible causes of the female employee's predicted distress.
[0110] Furthermore, the curation site allows users to view websites and videos related to solutions for the causes of their ailments, and to transition to online medical consultation sites. Female employee UA can move cursor K1 and click to view desired websites and videos, transition to online medical consultation sites, and consider specific solutions for her own ailments.
[0111] [Example Dashboard] Next, we will explain the visualization data displayed on the company terminal 20 of company A. Figures 21 to 28 show examples of the screen of the company terminal 20.
[0112] Screens M21 to M28 in Figures 21 to 28 show the results (anonymized) of a survey of female employees working at company A regarding women's working styles. On each screen, you can select the attributes of the female employees you want to visualize by selecting the survey time, department, position, age, and pregnancy / childbirth experience.
[0113] Screen M21 in Figure 21 shows the distribution of female employees' job titles. Screen M22 in Figure 22 shows the results of responses regarding satisfaction with company initiatives. Screen M23 in Figure 23 shows the results of responses regarding the work environment. Screen M24 in Figure 24 shows the results of responses regarding vacation time taken. Screen M25 in Figure 25 shows the results of responses regarding awareness and experience of using company A's policies and services. Screen M26 in Figure 26 shows the results of responses regarding career aspirations. Screen M27 in Figure 27 shows the results of responses regarding concerns about health issues specific to women. Screen M28 in Figure 28 shows the results of responses regarding things that women have given up on due to health issues specific to women.
[0114] Below screens M23 to M28, the report contents R23 to R28 generated by the generating AI210 for each item are displayed. By reviewing the report contents R23 to R28, Company A's welfare officer H1 can recognize hidden issues, necessary countermeasures, and specific solutions collected from the inner thoughts of female employees.
[0115] The report may include not only statistics and summaries, but also more in-depth content based on the voices of women collected through chat. For example, it may include information showing the career paths of female employees based on their menstrual symptoms, or information showing the actual career experiences of female employees, categorized by their age and life events (marriage, pregnancy, childbirth).
[0116] Therefore, in company A, even if employee H1 in charge of employee welfare is unsure of what to do or unable to speak with many female employees, they can identify the hidden challenges faced by female employees, the necessary measures, and specific solutions, enabling them to efficiently plan future policies and introduce services. Furthermore, employee H1 in charge of employee welfare may also create a platform for female employees to resolve their concerns by matching them with other female employees who share similar health conditions, circumstances, or career paths, forming mentor groups, and creating communication circles.
[0117] [Effects of the embodiment] The support server 10 in this embodiment gathers the inner thoughts (current subjective information) of female employees through a chatbot using a generative AI that can engage in empathetic conversations, based on health management information specific to women and responses to a questionnaire for female employees regarding health issues specific to women and women's work styles. Then, based on the inner thoughts of the female employees, the support server 10 narrows down the causes of the female employees' ailments and proposes one or more solutions to the female employees. In this way, the support server 10 can facilitate the smooth resolution of health issues specific to women by providing concrete solutions to the health issues of female employees.
[0118] The support server 10 then uses a chatbot to gather the unspoken thoughts of female employees that cannot be collected through surveys, extracting the information that female employees truly need from the company and providing it to the company. This allows the company to concretely consider developing measures and introducing services that directly address the concerns and requests of female employees, including health issues specific to female employees. In this way, the support server 10 supports companies in developing concrete measures and introducing services to create a workplace where female employees can work comfortably.
[0119] Therefore, the support server 10 can provide appropriate support for the health issues of female employees and bridge the gap between the measures implemented by the company and the needs that female employees expect from the company.
[0120] [System configuration of the embodiment] The support server 10 and various terminals are conceptual in function and do not necessarily need to be physically configured as shown in the diagram. In other words, the specific forms of distribution and integration of the functions of the support server 10 and various terminals are not limited to those shown in the diagram, and all or part of them can be configured by functionally or physically distributing or integrating them in any unit according to various loads and usage conditions.
[0121] Furthermore, each process performed on the support server 10 and various terminals may be implemented, in whole or in part, by a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), and a program that is analyzed and executed by the CPU and GPU. Alternatively, each process performed on the support server 10 and various terminals may be implemented as hardware using wired logic.
[0122] Furthermore, among the processes described in the embodiments, all or part of the processes described as being performed automatically can be performed manually. Alternatively, all or part of the processes described as being performed manually can be performed automatically by known methods. In addition, the processing procedures, control procedures, specific names, and information including various data and parameters described above and illustrated may be changed as appropriate unless otherwise specified.
[0123] [program] Figure 29 shows an example of a computer in which a support server 10 and various terminals are realized when a program is executed. Computer 1000 has, for example, memory 1010 and CPU 1020. Computer 1000 also has a hard disk drive interface 1030, a disk drive interface 1040, a serial port interface 1050, a video adapter 1060, and a network interface 1070. These components are connected by a bus 1080.
[0124] Memory 1010 includes ROM 1011 and RAM 1012. ROM 1011 stores, for example, a boot program such as the BIOS (Basic Input Output System). The hard disk drive interface 1030 is connected to the hard disk drive 1090. The disk drive interface 1040 is connected to the disk drive 1100. For example, a removable storage medium such as a magnetic disk or optical disk is inserted into the disk drive 1100. The serial port interface 1050 is connected to, for example, the mouse 1110 and the keyboard 1120. The video adapter 1060 is connected to, for example, the display 1130.
[0125] The hard disk drive 1090 stores, for example, an OS (Operating System) 1091, an application program 1092, a program module 1093, and program data 1094. That is, the programs defining the various processes of the support server 10 and various terminals are implemented as program modules 1093 containing code executable by the computer 1000. The program modules 1093 are stored, for example, on the hard disk drive 1090. For example, a program module 1093 for performing the same processes as those in the support server 10 and various terminals is stored on the hard disk drive 1090. Note that the hard disk drive 1090 may be replaced by an SSD (Solid State Drive).
[0126] Furthermore, the configuration data used in the processing of the above-described embodiment is stored as program data 1094 in, for example, memory 1010 or hard disk drive 1090. The CPU 1020 then reads the program module 1093 and program data 1094 stored in memory 1010 or hard disk drive 1090 into RAM 1012 as needed and executes them.
[0127] Furthermore, the program module 1093 and program data 1094 are not limited to being stored in the hard disk drive 1090; for example, they may be stored in a removable storage medium and read by the CPU 1020 via a disk drive 1100 or the like. Alternatively, the program module 1093 and program data 1094 may be stored in another computer connected via a network (LAN (Local Area Network), WAN (Wide Area Network), etc.). The program module 1093 and program data 1094 may then be read by the CPU 1020 from the other computer via a network interface 1070.
[0128] Although embodiments applying the invention made by the present inventors have been described above, the present invention is not limited by the descriptions and drawings that constitute part of the disclosure of the present invention in these embodiments. That is, all other embodiments, examples, and operational techniques made by those skilled in the art based on these embodiments are included in the scope of the present invention. [Explanation of Symbols]
[0129] 10 Support Servers 11 Communications Department 12 Storage section 13 Control Unit 20 Enterprise terminals 30, 30-1, 30-2 User terminals 121 Survey Results 121-1 Questionnaire Survey Responses 121A Questionnaire Survey 122 User Information 123 Health management information 124,124-1 Chat history 125 Analysis results 126 Prompt data 126A, 126B prompt Reports 127, 127-1, R1, R2 128. Information on countermeasures 129 Visualization Data 131 Survey Data Collection Department 132 Health Management Information Acquisition Department 133 Analysis Department 134 Chat Control Unit 134A, 134B Generated AI Chat 135 Section on Solutions 136 Report Creation Department 137 Visualization section 200 AI Generator Servers 210 Generation AI
Claims
1. A first acquisition unit acquires health management information showing the past health status of a female user as quantitative information of the female user's past health. A second acquisition unit that acquires past subjective information of the aforementioned female user, An analysis unit that estimates causal relationships between the information based on the aforementioned past quantitative information and the aforementioned past subjective information, and predicts the state of discomfort of the female user, as well as the candidate causes and timing of the discomfort, based on the aforementioned causal relationships. A third acquisition unit that uses a generative AI (Artificial Intelligence) model to generate text corresponding to the information input by the female user and engage in dialogue with the female user, and acquires at least the female user's current subjective information regarding the prediction results by the analysis unit, Based on the aforementioned subjective information, the presenting unit narrows down the cause of the female user's discomfort from the candidate causes and presents one or more solutions for the narrowed-down cause to the female user's terminal. A processing apparatus characterized by having
2. The aforementioned past quantitative information includes past weather information, information including past female-specific physical information of the female user, vital information, sleep information, dietary history, and / or pain history. The processing apparatus according to claim 1, characterized in that the subjective information is at least the results of a questionnaire answered by the female user regarding health issues specific to women, and / or the chat history of the female user regarding health issues specific to women.
3. The processing apparatus according to claim 1, characterized in that the third acquisition unit instructs the generating AI model to engage in a dialogue in which it listens to the female user's concerns and instructs it to narrow down the possible causes of the discomfort, generates text that is empathetic and uses language that is considerate of women, and instructs the model to change the timing of its responses in response to the questions.
4. The aforementioned questionnaire concerns health issues specific to women and women's work styles. The subjective information includes the results of the female user's responses to the survey regarding women's specific health issues and women's work styles, and / or the chat history of the female user regarding women's specific health issues and women's work styles. The third acquisition unit acquires chat histories from multiple female users regarding health issues specific to women and the working styles of women, The aforementioned processing apparatus is A generation unit instructs the generation AI model to create a report summarizing health issues, lifestyle issues, dissatisfaction with systems and policies, career views, and / or workplace dissatisfaction, based on the current subjective information of multiple female users acquired by the third acquisition unit. A visualization unit that creates visualization data from the statistical results of the aforementioned survey and the aforementioned report, and transmits the created visualization data to the terminal of the company where the female user works. The apparatus according to claim 2, further comprising the above.
5. The processing apparatus according to claim 4, characterized in that the creation unit instructs the generating AI model to create a report that shows hidden issues, necessary countermeasures, and solutions regarding health issues, lifestyle issues, dissatisfaction with systems and policies, views on careers, and / or dissatisfaction in the workplace.
6. The processing apparatus according to claim 1, characterized in that the display unit presents one or more of the countermeasures to the female user's terminal in a state that allows it to cooperate with a service provider corresponding to the countermeasure.
7. A processing method executed by a processing unit, A step of acquiring health management information showing the past health status of a female user as past quantitative information of the female user, The process of obtaining past subjective information of the aforementioned female user, The process involves estimating causal relationships between the information based on the aforementioned past quantitative information and the aforementioned past subjective information, and predicting the state of discomfort of the female user, as well as potential causes and timing of the discomfort, based on the aforementioned causal relationships. Using a generative AI (Artificial Intelligence) model that generates text corresponding to the information input by the female user and engages in dialogue with the female user, the process includes obtaining the female user's current subjective information regarding the prediction results in at least the prediction process, Based on the aforementioned subjective information, the process involves narrowing down the possible causes of the female user's discomfort from the list of possible causes, and presenting one or more solutions for the narrowed-down causes to the female user's terminal. A processing method characterized by including the following.
8. A step of acquiring health management information showing the past health status of a female user as past quantitative information of the female user, The steps include obtaining past subjective information of the aforementioned female user, The process involves estimating causal relationships between the information based on the aforementioned past quantitative information and subjective information, and predicting the state of the female user's discomfort, potential causes of the discomfort, and the timing of the discomfort based on these causal relationships. Using a generative AI (Artificial Intelligence) model that generates text corresponding to the information input by the female user and engages in dialogue with the female user, the steps include obtaining the female user's current subjective information regarding the prediction results in at least the prediction step, Based on the aforementioned subjective information, the steps include narrowing down the cause of the female user's discomfort from the candidate causes and presenting one or more solutions for the narrowed-down cause to the female user's terminal, A processing program that causes a computer to execute something.