Information processing device, processing method for information processing device, and program

Abstract

To enable dialogue between the client and the interlocutor in accordance with the client's mental state. [Solution] The information processing device includes a first dialogue means for engaging in dialogue with a client, and an estimation means for inferring the mental state of the client by analyzing the dialogue between the client and the first dialogue means. Based on the inferred mental state, the first dialogue means engages in dialogue with the client to change the dialogue between the client and the first dialogue means into a dialogue between the client and the dialogue partner.

Description

Technical Field

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

Background Art

[0002] There are many people with mental problems, including those who are potential patients. However, since the number of counselors is small, it is difficult to counsel all people with mental problems. Therefore, mental health tech (digital tools and apps that support mental health) has emerged.

[0003] Patent Document 1 discloses an information processing apparatus that receives emotion information related to emotions from a user and presents a response sentence to an input sentence from the user based on the emotion information and input environment information.

[0004] Patent Document 2 discloses a stress management support system including a video sensor having a function of reading expressions, an audio sensor for listening to voices, and a humanoid robot equipped with artificial intelligence, and the artificial intelligence has a function of performing stress coping, cognitive behavioral therapy, or assertion in an interactive format.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0006] However, the information processing apparatus of Patent Document 1 only presents a response sentence to an input sentence from the user, and thereby does not show a countermeasure when the symptoms of a person with a mental problem are not improved.

[0007] The stress management support system described in Patent Document 2 performs stress coping, cognitive behavioral therapy, or assertion in a dialogue format, but it does not provide countermeasures for cases where the symptoms of a person with mental health problems do not improve through the dialogue format of artificial intelligence.

[0008] The purpose of this disclosure is to enable dialogue between the client and the interlocutor to be conducted in accordance with the client's mental state. [Means for solving the problem]

[0009] The information processing device includes a first dialogue means for engaging in dialogue with a client, and an estimation means for inferring the mental state of the client by analyzing the dialogue between the client and the first dialogue means, wherein the first dialogue means engages in dialogue with the client to change the dialogue between the client and the first dialogue means into a dialogue between the client and the dialogue partner, based on the inferred mental state. [Effects of the Invention]

[0010] According to this disclosure, it is possible to conduct dialogue between the client and the interlocutor in accordance with the client's mental state. [Brief explanation of the drawing]

[0011] [Figure 1] This figure shows an example configuration of an information processing system according to the first embodiment. [Figure 2] This diagram shows an example of the hardware configuration for a user information terminal and a counselor information terminal. [Figure 3] This diagram shows an example of a server hardware configuration. [Figure 4] This figure shows an example of the server's functional configuration. [Figure 5] This figure shows an example of a dialogue screen displayed on the user information terminal's display unit. [Figure 6]It is a flowchart showing a server processing method. [Figure 7] It is a diagram showing an example of a dialogue screen displayed on the display unit of the user information terminal. [Figure 8] It is a diagram showing an example of a dialogue screen displayed on the display unit of the user information terminal. [Figure 9] It is a diagram showing an example of a dialogue screen displayed on the display unit of the counselor information terminal. [Figure 10] It is a flowchart showing a server processing method. [Figure 11] It is a diagram showing an example of a dialogue screen displayed on the display unit of the user information terminal. [Figure 12] It is a diagram showing an example of a dialogue screen displayed on the display unit of the counselor information terminal. [Figure 13] It is a diagram showing an example of a medical record. [Figure 14] It is a diagram showing an example of another medical record. [Figure 15] It is a diagram for explaining the server follow-up process. [Figure 16] It is a flowchart showing a server processing method according to the second embodiment. [Figure 17] It is a diagram for explaining the processing of a specific dialogue model.

Embodiments for Carrying Out the Invention

[0012] (First Embodiment) FIG. 1 is a diagram showing a configuration example of an information processing system 100 according to the first embodiment. The information processing system includes a server 101, a user information terminal 102, a counselor information terminal 103, a network 104, and a wearable device 109.

[0013] The user information terminal 102 and the counselor information terminal 103 are, for example, smartphones. The server 101 is an example of an information processing device and is, for example, a computer.

[0014] The user information terminal 102 and the server 101 can communicate with each other via the network 104. The counselor information terminal 103 and the server 101 can communicate with each other via the network 104. The user information terminal 102 and the counselor information terminal 103 can communicate with each other via the network 104 and the server 101.

[0015] The user information terminal 102 is an information terminal used by the user. The user, for example, is a patient suspected of having a mental illness and can request counseling regarding the mental illness using the user counseling application 107 on the user information terminal 102. The user is a client of counseling. The user counseling application 107 is a computer program that allows the user to utilize the counseling service provided by the server 101.

[0016] The counselor information terminal 103 is an information terminal used by counselors. Counselors are healthcare professionals specializing in mental illness, such as "clinical psychologists" or "certified psychologists" who perform psychological counseling (psychotherapy). Counselors can accept counseling assignments related to mental illness using the counselor's counseling application 108 on the counselor information terminal 103. Counselors are the interlocutors for counseling users. The counselor's counseling application 108 is a computer program that allows counselors to use the counseling service provided by the server 101.

[0017] Server 101 provides counseling services to user information terminal 102 and counselor information terminal 103 using a dialogue model.

[0018] The wearable device 109 and the user information terminal 102 can communicate with each other. The wearable device 109 is, for example, an Apple Watch and is worn on the user's body. The wearable device 109 detects ambient light, acceleration, gyroscope, heart rate, atmospheric pressure, compass, and blood oxygen, and transmits them to the server 101 via the user information terminal 102.

[0019] Figure 2 shows an example of the hardware configuration of the user information terminal 102 and counselor information terminal 103 shown in Figure 1. The user information terminal 102 and counselor information terminal 103 are, for example, smartphones.

[0020] The user information terminal 102 and the counselor information terminal 103 each have a controller unit 200, a display unit 207, an operation unit 209, a call unit 213, an audio antenna 216, a wireless LAN antenna 218, and an imaging unit 220, respectively.

[0021] The controller unit 200 includes a CPU 201, RAM 202, ROM 203, storage device 204, power control unit 205, display unit controller 206, operation unit controller 208, call unit controller 212, wireless LAN interface 217, and imaging unit controller 219. The operation unit 209 includes a touch panel 210 and keys 211. The call unit 213 includes a microphone 214 and speaker 215.

[0022] The CPU 201 is a central processing unit that controls the entire user information terminal 102 or counselor information terminal 103.

[0023] RAM202 is the work memory for the CPU201 to operate, and is used for loading various programs and saving the results of calculations.

[0024] ROM203 is the memory that stores the CPU201's startup program and various configuration information.

[0025] The storage device 204 is a non-volatile memory for storing large programs and data, and is an internal flash memory or HDD of the controller unit 200. The storage device 204 stores the user counseling application 107 or the counselor counseling application 108 shown in Figure 1.

[0026] The power control unit 205 is composed of a battery, a battery charging module, and the like, in order to control the power required for the operation of the user information terminal 102 or the counselor information terminal 103.

[0027] The display controller 206 transmits the text of the counseling conversation and the video of the video call to the display unit 207 in accordance with the communication protocol of the display unit 207.

[0028] The control unit controller 208 receives input from the touch panel 210 and the keys 211 and converts it into data that the CPU 201 can understand. The touch panel 210 and the display unit 207 are integrated, and the touch panel 210 can detect touch operations performed on the display surface of the display unit 207.

[0029] The communication unit controller 212 converts the audio signal received from the microphone 214 into radio waves for transmission to the other device and outputs them to the audio antenna 216, or processes the radio waves received from the other device by the audio antenna 216 into audio and outputs them to the speaker 215.

[0030] The imaging unit 220 captures an image of the subject using a camera and generates image data. The imaging unit controller 219 controls the imaging unit 220.

[0031] The wireless LAN interface 217 transmits and receives data to and from the network 104 in Figure 1 via the wireless LAN antenna 218. The wireless LAN interface 217 of the user information terminal 102 also transmits and receives data to and from the wearable device 109 in Figure 1 via the wireless LAN antenna 218.

[0032] The CPU 201 can perform voice calls using the call unit 213 and the wireless LAN I / F 217 by running the user counseling application 107 or the counselor counseling application 108. Voice calls are made between the user information terminal 102 and the counselor information terminal 103 via the server 101.

[0033] Furthermore, the CPU 201 can perform video calls using the communication unit 213, imaging unit 220, and wireless LAN I / F 217 by executing the user counseling application 107 or the counselor counseling application 108. The video call is conducted between the user information terminal 102 and the counselor information terminal 103 via the server 101.

[0034] Figure 3 shows an example of the hardware configuration of server 101 shown in Figure 1. Server 101 includes a controller unit 300, a display unit 307, and an operation unit 309.

[0035] The controller unit 300 includes a CPU 301, RAM 302, ROM 303, storage device 304, power control unit 305, display unit controller 306, operation unit controller 308, and network interface 310. The operation unit 309 includes a keyboard and a pointing device, etc.

[0036] CPU301 is the central processing unit that controls the entire server 101.

[0037] RAM302 is the work memory for the CPU301 to operate, and is used for loading various programs and saving the results of calculations.

[0038] ROM303 is memory where the CPU301's startup program and various configuration information are stored.

[0039] The storage device 304 is a non-volatile memory for storing large programs and data, and is an internal flash memory or HDD of the controller unit 300.

[0040] The power control unit 305 controls the power required for the server 101 to operate.

[0041] The display controller 306 transmits display data to the display unit 307 in accordance with the display unit 307's communication protocol.

[0042] The control unit controller 308 receives input from the control unit 309 and converts it into data that the CPU 301 can understand. The control unit 309 and the display unit 307 may also be combined into a single patch panel.

[0043] Network I / F 310 sends and receives data to and from network 104 in Figure 1.

[0044] Figure 4 shows an example of the functional configuration of server 101 in Figure 1. Server 101 realizes the functional configuration section in Figure 4 when the CPU 301 in Figure 3 executes a program.

[0045] Server 101 includes a predictive dialogue model 401, a specific dialogue model 402, a control unit 403, a prediction unit 404, a medical record generation unit 405, and a presentation unit 406.

[0046] The inference dialogue model 401 and the identification dialogue model 402 are, for example, AI chatbots, which are programs in which an AI that has self-learned based on data and logs responds to user statements, and provides appropriate responses to what the user says in natural language. Here, AI refers to artificial intelligence.

[0047] The inferential dialogue model 401 is used to infer the user's mental state by asking questions to the user and, based on the user's responses, repeatedly asking appropriate questions to the user. Here, mental state includes stress, anxiety, depression, anger, trauma, loss of confidence, loneliness, etc. The inferred mental state may be expressed as a numerical value representing these, or it may not be expressed as a numerical value, such as high stress, slightly high stress, etc.

[0048] The identification dialogue model 402 is used to identify potential symptoms of the user by asking questions to the user and, based on the user's responses, repeatedly asking appropriate questions to the user.

[0049] By instructing the user to run the user counseling application 107 on the user information terminal 102, the user can engage in a dialogue with the inference dialogue model 401 and the counselor, as shown in Figure 5.

[0050] Figure 5 shows example screens of dialogues 501 to 503 displayed on the display unit 307 of the user information terminal 102. Here, the user (patient) is the user (client) using the user information terminal 102. The predictive dialogue model (chatbot) 401 is a dialogue model within the server 101. The counselor is the counselor using the counselor information terminal 103.

[0051] Dialogue 501 is a statement from the user to the predictive dialogue model (chatbot) 401. Dialogue 502 is a dialogue between the predictive dialogue model 401 and the user. Dialogue 503 is a statement from the counselor to the user.

[0052] The inference dialogue model 401 engages in a dialogue 501 with the user. The inference unit 404 analyzes the dialogue 501 between the user and the inference dialogue model 401 to infer the user's mental state.

[0053] The prediction unit 404 may also predict the user's mental state by analyzing the dialogue 501 between the user and the prediction dialogue model 401, and the information detected by the wearable device 109 owned by the user. This information may include ambient light, acceleration, gyroscope, heart rate, atmospheric pressure, compass, or blood oxygen level.

[0054] Based on the mental state inferred by the inference unit 404, the inference dialogue model 401 engages in dialogue 502 with the user to change the dialogue between the user and the inference dialogue model 401 to a dialogue between the user and the counselor.

[0055] In dialogue 502, the predictive dialogue model 401 presents one or more counselor candidates based on the user's information. For example, the predictive dialogue model 401 presents counselor candidates A and B. The user has stated that they prefer counselor A. The user's information mentioned above may include, for example, age or address. The predictive dialogue model 401 can then present counselors who are close to the user's address.

[0056] Furthermore, if the user agrees in dialogue 502 to change the dialogue between the user and the counselor, the inference dialogue model 401 may also ask the counselor whether or not to proceed with the dialogue between the user and the counselor.

[0057] In this case, the inference dialogue model 401 can present the counselor with a summary of the content of the dialogue 501 between the user and the inference dialogue model 401.

[0058] If the counselor accepts the above inquiry, the control unit 403 controls the system to take place a dialogue 503 between the user and the counselor after a dialogue 502 between the user and the predictive dialogue model 401.

[0059] In dialogue 503, the inference dialogue model 401 can present the counselor with proposed responses to the user based on the content of dialogue 501 between the user and the inference dialogue model 401, and / or the content of dialogue 503 between the user and the counselor.

[0060] Here, path 105 in Figure 1 is the path of dialogues 501 and 502 between the user and the inference dialogue model 401. Path 106 in Figure 1 is the path of dialogue 503 between the user and the counselor.

[0061] The dialogues 501 and 502 between the user and the inference dialogue model 401 are text-based dialogues. The dialogue 503 between the user and the counselor is also a text-based dialogue.

[0062] Furthermore, dialogues 501 and 502 between the user and the predictive dialogue model 401 are text-only dialogues, while dialogue 503 between the user and the counselor may include dialogue via voice call or video call.

[0063] As shown in Figure 5, dialogues 501 to 503 take place on the same screen.

[0064] After the dialogue 503, the medical record generation unit 405 generates a medical record based on the content of the dialogue 501 between the user and the predictive dialogue model 401, and / or the dialogue 503 between the user and the counselor. The medical record is, for example, in SOAP format. The medical record may also include findings from the analysis of the above-mentioned voice call or video call. Details will be described later.

[0065] The medical record generation unit 405 then controls the system so that only the counselor can download the medical record, or it presents the medical record only to the counselor.

[0066] The presentation unit 406 receives information detected by the user's wearable device 109 as a periodic follow-up after the interaction between the user and the counselor, and presents the received information and / or the results of the analysis of the received information to the counselor. The above information may include ambient light, acceleration, gyroscope, heart rate, atmospheric pressure, compass, or blood oxygen level.

[0067] In addition, in dialogue 501, the identification dialogue model 402 may identify one or more symptom candidates based on dialogue 501 with the user. In that case, after identifying the one or more symptom candidates, the inference dialogue model 401 or the identification dialogue model 402 will conduct a dialogue to change the dialogue to one between the user and the dialogue provider. Before dialogue 503, the inference dialogue model 401 or the identification dialogue model 402 will present the identified one or more symptom candidates to the counselor.

[0068] Figure 6 is a flowchart showing the processing method of server 101. When the user instructs the user to run the user counseling application 107 on the user information terminal 102, server 101 initiates a dialogue between the user and the inference dialogue model 401 shown in Figure 6.

[0069] In step S601, the inference unit 404 uses the inference dialogue model 401 to conduct a dialogue 701 with the user, as shown in Figure 7, and receives the user's statements. Figure 7 shows an example of the dialogue 701 screen displayed on the display unit 307 of the user information terminal 102. Dialogue 701 corresponds to dialogue 501 in Figure 5 and is a dialogue between the inference dialogue model (mental doctor) 401 and the user. The user is responding to the questions posed by the inference dialogue model 401.

[0070] Furthermore, user responses are not limited to open-ended input; they may also be able to select from a list of suggested answers. For example, the inference unit 404 can use the inference dialogue model 401 to present multiple suggested answers (e.g., "I feel dizzy," "I'm tired") after the question, "How are you feeling?"

[0071] In step S602, the inference unit 404 uses the inference dialogue model 401 to analyze the dialogue 701 between the user and the inference dialogue model 401 and infers the user's mental state.

[0072] The prediction unit 404 may also predict the user's mental state by analyzing the dialogue 701 between the user and the prediction dialogue model 401, and the information detected by the wearable device 109 owned by the user. This information may include ambient light, acceleration, gyroscope, heart rate, atmospheric pressure, compass, or blood oxygen level.

[0073] Here, ambient light is used to analyze sleep patterns and day-night rhythms. Accelerometer and gyroscope are used to analyze activity levels and exercise patterns. Heart rate is used to analyze stress levels derived from heart rate variability.

[0074] In step S603, the inference unit 404 determines whether or not there is a suspicion of a mental illness based on the user's mental state. If it is determined that there is no suspicion of a mental illness, the process returns to step S601 and the above process is repeated. If it is determined that there is a suspicion of a mental illness, the process proceeds to step S605.

[0075] In step S605, the inference dialogue model 401 performs a dialogue 801 with the user to change the dialogue between the user and the inference dialogue model 401 to a dialogue between the user and the counselor, as shown in Figure 8. Figure 8 shows an example of the dialogue screen 800 displayed on the display unit 307 of the user information terminal 102. Dialogue 801 corresponds to dialogue 502 in Figure 5 and is a dialogue between the inference dialogue model (mental doctor) 401 and the user, and is a text-only dialogue.

[0076] In dialogue 801, the predictive dialogue model 401 presents candidate counselors based on the user's information. User information may include, for example, age or address. The predictive dialogue model 401 can present counselors who are close to the user's address. Alternatively, as shown in dialogue 502 in Figure 5, the predictive dialogue model 401 may present multiple candidate counselors based on the user's information.

[0077] In step S606, the inference dialogue model 401 determines whether the user has agreed to the change in dialogue 801. In the example dialogue 801 in Figure 8, the user has agreed. If it is determined that the user has not agreed, the process returns to step S601. If it is determined that the user has agreed, the process proceeds to step S607.

[0078] Additionally, the user counseling app 107 may be configured to either "Change automatically" or "Ask for permission when changing." If "Change automatically" is selected, the process will skip step S606 from step S605 and proceed to step S607.

[0079] In step S607, the control unit 403 changes the dialogue mode from between the inference dialogue model 401 and the user to between the inference dialogue model 401 and the counselor.

[0080] In step S608, the inference dialogue model 401 presents the counselor with a summary of the content of the dialogue between the user and the inference dialogue model 401, as shown in dialogue 901 in Figure 9, and asks the counselor whether or not to have a dialogue between the user and the counselor. Figure 9 shows an example of the dialogue screen 900 displayed on the display unit 307 of the counselor information terminal 103. Dialogues 901 and 902 are dialogues between the inference dialogue model (mental doctor) 401 and the counselor. If the counselor agrees to the above inquiry, the inference dialogue model 401 terminates the dialogue mode between the inference dialogue model 401 and the counselor and proceeds to the process shown in Figure 10.

[0081] Figure 10 is a flowchart showing the processing method of server 101 following the processing in Figure 6. The control unit 403 displays notifications 802 and 803 to initiate the dialogue mode between the user and the counselor, as shown in Figures 8 and 9. The control unit 403 then controls the dialogue 803 and 904 between the user and the counselor, as shown in Figures 8 and 9. Dialogue 803 and 904 are statements from the counselor to the user.

[0082] On the screen 800 of the user information terminal 102, the dialogue continues on the same screen 800 even when the dialogue changes from dialogue 801 with the inference dialogue model 401 to dialogue 803 with the counselor.

[0083] Similarly, on the screen 900 of the counselor information terminal 103, the dialogue continues on the same screen 900 even when it changes from dialogues 901-902 with the predictive dialogue model 401 to dialogue 904 with the user.

[0084] Furthermore, since the screen 900 of the counselor information terminal 103 displays dialogue 901, which is a summary of the conversation between the user and the predictive dialogue model 401, the counselor does not need to ask the same question again.

[0085] Figure 11 shows an example of a dialogue screen 1100 displayed on the display unit 307 of the user information terminal 102, following screen 800 in Figure 8. Figure 12 shows an example of a dialogue screen 1200 displayed on the display unit 307 of the counselor information terminal 103, following screen 900 in Figure 9.

[0086] Dialogue 1101 in Figure 11 corresponds to dialogue 1201 in Figure 12. Dialogue 1102 in Figure 11 corresponds to dialogue 1202 in Figure 12. Dialogues 1101 and 1201 between the user and the counselor are text-based. Dialogues 1102 and 1202 between the user and the counselor are voice or video calls. When the control unit 403 enters the dialogue mode between the user and the counselor, it allows the use of voice and video calls.

[0087] On the user information terminal 102, the screen following the dialogue 701 screen in Figure 7 is screen 800 in Figure 8, and the screen following screen 800 in Figure 8 is screen 1100 in Figure 11. The dialogue 701 screen, screen 800, and screen 1100 are all the same screen.

[0088] Similarly, on the counselor information terminal 103, the screen following screen 900 in Figure 9 is screen 1200 in Figure 12. Screens 900 and 1200 are the same screen.

[0089] In step S1001, the inference dialogue model 401 receives the user dialogue (statement) shown in Figure 11.

[0090] In step S1002, the inference dialogue model 401 analyzes the content of the user dialogue in Figure 7 and / or the user dialogue in Figure 11.

[0091] In step S1003, the inference dialogue model 401 presents (outputs) suggested responses to the user to the counselor based on the results of the analysis in step S1001. These suggested responses are, for example, candidate questions. The counselor can refer to these suggested responses and then make a response.

[0092] In step S1004, the inference dialogue model 401 receives the dialogue of the counselor's responses shown in Figure 12.

[0093] In step S1005, the inference dialogue model 401 sends the counselor's response dialogue to the user. As a result, the counselor's response dialogue is displayed on the screen 1100 of the user information terminal 102 in Figure 11.

[0094] In step S1006, the inference dialogue model 401 determines whether the dialogue between the user and the counselor has ended. If it is determined that the dialogue has not ended, the process returns to step S1001. If it is determined that the dialogue has ended, the process proceeds to step S1007.

[0095] In step S1007, the inference dialogue model 401 sends a dialogue 1203 to the counselor requesting input for <counselor's findings>, and receives a dialogue 1204 from the counselor containing the content of <counselor's findings>.

[0096] The medical record generation unit 405 generates a medical record 1300, as shown in Figure 13, based on the content of the dialogue between the user and the predictive dialogue model 401, the dialogue between the user and the counselor, and the dialogue between the counselor and the predictive dialogue model 401.

[0097] In step S1008, the medical record generation unit 405 controls the system so that only the counselor can download the medical record, or it presents the medical record only to the counselor and uses the inference dialogue model 401 to send a dialogue 1205 to the counselor to that effect. Users cannot view the medical record. Counselors can view the medical record.

[0098] Subsequently, the control unit 403 terminates the dialogue mode between the counselor and the inference dialogue model 401.

[0099] Figure 13 shows an example of a medical record 1300 generated in step S1007 of Figure 10. Medical record 1300 is, for example, a SOAP-format medical record and includes S (subjective information) 1301, O (objective information) 1302, A (assessment) 1303, and P (plan) 1304.

[0100] S (subjective information) 1301 includes not only the content of the dialogue between the user and the counselor, but also the content of the dialogue between the user and the inference dialogue model 401.

[0101] O (objective information) 1302 includes findings 1311 based on the analysis results of the verbal dialogue 1101 in Figure 11, and findings 1312 based on the analysis results of the voice call or video call dialogue 1202.

[0102] Figure 14 shows an example of another medical record 1400 generated in step S1007 of Figure 10. Medical record 1400 is a SOAP-format medical record and includes S (subjective information) 1401. S (subjective information) 1401 separately describes the findings 1402 of the inferential dialogue model 401 and the counselor's findings 1403. The findings 1402 of the inferential dialogue model 401 are findings based on the analysis of the content of the dialogue with the inferential dialogue model 401. The counselor's findings 1403 are findings based on the counselor's dialogue 1204 in Figure 12.

[0103] O, A, and P are described in the same way as S (subjective information) 1401.

[0104] Figure 15 is a diagram illustrating the follow-up process of server 101 after the processing shown in the flowchart of Figure 10.

[0105] The wearable device 109 is worn by the user and detects information such as ambient light, acceleration, gyroscope, heart rate, atmospheric pressure, compass, and blood oxygen level. The user information terminal 102 periodically receives the information detected by the wearable device 109 and transmits that information to the server 101.

[0106] The display unit 406 of the server 101 receives information detected by the wearable device 109 owned by the user, and, as shown in screen 1500, uses the predictive dialogue model 401 to present the received information and / or the results of the analysis of the received information to the counselor. Screen 1500 is the screen displayed on the display unit 307 of the counselor information terminal 103.

[0107] This allows counselors to recognize user information as part of follow-up observations after dialogues between the user and the counselor.

[0108] (Second embodiment) Figure 16 is a flowchart showing the processing method of server 101 according to the second embodiment. Figure 16 is obtained by adding steps S1601 to S1604 to Figure 6. The differences between the second embodiment and the first embodiment will be explained below.

[0109] If a suspected mental illness is determined in step S603, the process proceeds to step S1601.

[0110] In step S1601, the control unit 403 changes the dialogue mode from between the user and the inference dialogue model 401 to between the user and the identification dialogue model 402. This change does not require the user's consent. The identification dialogue model 402 is a dialogue model for identifying the user's symptoms.

[0111] In step S1602, the identification dialogue model 402 engages in a dialogue (questions and answers) with the user to identify symptoms and receives the user's statements (answers).

[0112] In step S1603, the specific dialogue model 402 analyzes the dialogue between the user and the specific dialogue model 402.

[0113] In step S1604, the identification dialogue model 402 identifies one or more candidate symptoms for the user based on the results of the analysis in step S1603. These symptoms may include, for example, schizophrenia, mood disorders, epilepsy, etc.

[0114] The Specific Dialogue Model 402 determines that a symptom has been identified when its confidence level reaches 70% or higher. It is preferable for the Specific Dialogue Model 402 to identify multiple symptom candidates with a confidence level of 70% or higher, rather than identifying just one symptom. Identifying only one symptom increases the likelihood of error. Identifying multiple symptom candidates increases the likelihood of the correct answer being among them. The final judgment on the symptoms should preferably be left to the counselor.

[0115] The identification dialogue model 402 determines whether one or more candidate symptoms have been identified. If it is determined that no symptoms have been identified, the confidence level of the symptoms is low, so the process returns to step S1602 and asks the necessary questions to identify the symptoms. If it is determined that symptoms have been identified, the process proceeds to step S605.

[0116] The processing from step S605 onward is the same as in Figure 6. However, in step S608 of Figure 6, the inference dialogue model 401 or the identification dialogue model 402 presents the counselor with one or more candidate symptoms that have been identified. The candidate symptoms are displayed on the display unit 307 of the counselor information terminal 103, making it easier for the counselor to make a final judgment on the user's symptoms.

[0117] Figure 17(A) is a diagram illustrating the training process of the specific dialogue model 402. The neural network 1702 is the neural network of the specific dialogue model 402. During training, the neural network 1702 takes input data 1701 as input and outputs output data 1703.

[0118] Input data 1701 is data on typical findings, such as "frequently complains of hallucinations or delusions." Output data 1703 is data on symptoms.

[0119] The neural network 1702 repeatedly adjusts its parameters so that the output data 1703 approaches the correct data 1704. The correct data 1704 is the correct data (training data) for the input data 1701, for example, "schizophrenia".

[0120] First, let's explain an example of the combination of input data 1701 and correct answer data 1704 when the symptoms are "schizophrenia". Input data 1701 includes phrases such as "frequently complains of hallucinations and delusions," "difficulty organizing thoughts and jumping between topics," "shows emotional flattening and poor emotional expression," "has weak social relationships and tends to avoid social interaction," and "believes to be special and often feels monitored or attacked by others." The correct answer data 1704 is "schizophrenia."

[0121] Next, we will explain an example of a combination of input data 1701 and correct answer data 1704 when the symptoms are "mood disorder (depression)". Input data 1701 includes: "persistent depressed mood with a significant decrease in interest and pleasure," "constantly feeling tired and lacking energy," "low self-esteem and feeling worthless," "sleep disturbances (hypersleeping or insomnia)," and "changes in appetite (overeating or loss of appetite) and weight fluctuations." The correct answer data 1704 is "mood disorder (depression)."

[0122] Next, we will explain an example of the combination of input data 1701 and correct answer data 1704 when the symptom is "epilepsy". Input data 1701 includes statements such as, "Sudden loss of consciousness or seizures occur repeatedly," "The body may stiffen or muscles may spasm during a seizure," "Severe fatigue and confusion are often felt after a seizure," "Short-term memory loss may occur," and "Certain stimuli (e.g., light or sound) may trigger a seizure." The correct answer data 1704 is "epilepsy."

[0123] Figure 17(B) is a diagram illustrating the inference process of the specific dialogue model 402. The neural network 1712 is the neural network of the specific dialogue model 402, and is a trained neural network that was trained using the neural network 1702 in Figure 17(A). The inference process corresponds to the process in step S1604 of Figure 16. During inference, the neural network 1712 takes input data 1711 as input and outputs output data 1713.

[0124] Input data 1711 is data of findings obtained by analyzing the dialogue between the user and the specific dialogue model 402. Output data 1713 is a list of candidate symptoms.

[0125] Output data 1713 preferably contains multiple (e.g., three) symptom candidates. The neural network 1712 should output multiple symptom candidates with a confidence level of 70% or higher, rather than outputting a single symptom. Outputting only one symptom increases the likelihood of error. Outputting multiple symptom candidates increases the likelihood of the correct answer being among them. The final decision on the symptoms should preferably be left to a counselor.

[0126] As described above, according to the first and second embodiments, the server 101 analyzes the content of the dialogue between the user and the predictive dialogue model 401, and if there is an abnormality in the user's mental state, the dialogue between the user and the counselor can be made without changing the screen. Because the user is provided with an opportunity to interact with a counselor in a way that is easy for the user, it is possible to reduce the number of people with mental health problems.

[0127] (Other embodiments) Furthermore, the disclosed technology can also be realized by performing the following process: that is, the disclosed technology can also be realized by supplying software (programs) that implement one or more functions of the various embodiments described above to a system or device via a network or storage medium, and the computer (or CPU, MPU, etc.) of that system or device reads and executes the program. The computer may have one or more processors or circuits and may include a network of separate computers or separate processors or circuits for reading and executing computer executable instructions. In this case, the processor or circuit may include a central processing unit (CPU), a microprocessing unit (MPU), a graphics processing unit (GPU), an application-specific integrated circuit (ASIC), or a field-programmable gateway (FPGA). The processor or circuit may also include a digital signal processor (DSP), a dataflow processor (DFP), or a neural processing unit (NPU).

[0128] Furthermore, the embodiments described above are merely examples illustrating how to implement this disclosure, and they should not be interpreted as limiting the technical scope of this disclosure. In other words, this disclosure can be implemented in various ways without departing from its technical concept or its main features.

[0129] This embodiment includes the following configuration. (Item 1) A first means of dialogue for engaging in conversation with the client, The system includes an inference means for inferring the mental state of the client by analyzing the dialogue between the client and the first dialogue means, The first dialogue means is an information processing device that, based on the inferred mental state, engages in dialogue with the client to change the dialogue between the client and the first dialogue means into a dialogue between the client and the dialogue partner. (Item 2) The first dialogue means is an information processing device according to item 1 that presents one or more candidates for dialogue partners based on the client information. (Item 3) The client information is the information processing device described in item 2, which is either age or address. (Item 4) The information processing device described in any one of items 1 to 3, wherein the dialogue between the client and the first dialogue means is a text-based dialogue. (Item 5) An information processing apparatus according to any one of items 1 to 4, further comprising control means for controlling the conduction of a dialogue between the client and the dialogue participant after a dialogue by the first dialogue means. (Item 6) The dialogue between the client and the first dialogue means is a text-based dialogue. The information processing device described in item 5, wherein the dialogue between the client and the interlocutor is a text-based dialogue. (Item 7) The dialogue between the client and the first dialogue means is a text-only dialogue. The information processing device described in item 5 includes a conversation between the client and the interlocutor via voice call or video call. (Item 8) The information processing device described in any one of items 1 to 7, wherein the dialogue between the client and the first dialogue means and the dialogue for modification are performed on the same screen. (Item 9) The information processing apparatus according to any one of items 5 to 7, wherein the dialogue between the client and the first dialogue means, the dialogue for modification, and the dialogue between the client and the dialogue participant are all performed on the same screen. (Item 10) An information processing apparatus according to any one of items 5 to 7, 9, further comprising a medical record generation means that generates a medical record based on the content of the dialogue between the client and the first dialogue means, or the content of the dialogue between the client and the dialogue means, after the dialogue between the client and the dialogue person. (Item 11) The medical record generation means is an information processing device according to item 10 that controls the state so that only the interlocutor can download the medical record, or presents the medical record only to the interlocutor. (Item 12) The inference means is an information processing device according to any one of items 1 to 11, which infers the mental state of the client by analyzing the dialogue between the client and the first dialogue means and the information detected by a device owned by the client. (Item 13) The information processing device described in item 12 is such that the information is ambient light, acceleration, gyroscope, heart rate, atmospheric pressure, compass, or blood oxygen level. (Item 14) An information processing apparatus according to any one of items 5 to 7, 9 to 11, further comprising a means for receiving information detected by a device owned by the client after a conversation between the client and the interlocutor, and for presenting the received information and / or the results of analysis of the received information to the interlocutor. (Item 15) The information processing device described in item 14 is such that the information is ambient light, acceleration, gyroscope, heart rate, atmospheric pressure, compass, or blood oxygen. (Item 16) The aforementioned medical record is a medical record in SOAP format, as described in item 10 or 11, and is an information processing device. (Item 17) The conversation between the client and the interlocutor includes conversations conducted via voice call or video call. The medical record is an information processing device described in any one of items 10, 11, or 16, which includes findings based on the analysis results of the voice call or video call. (Item 18) If the client agrees in the dialogue for modification, the first dialogue means will ask the dialogue partner whether or not to have a dialogue between the client and the dialogue partner. The control means controls the information processing device according to any one of items 5-7, 9-11, or 14-17 to take place when the interlocutor accepts the inquiry. (Item 19) The first dialogue means is an information processing device according to any one of items 5-7, 9-11, or 14-18, which presents the dialogue participant with a summary of the content of the dialogue between the client and the first dialogue means. (Item 20) The first dialogue means is an information processing device according to any one of items 5-7, 9-11, or 14-19, which presents a proposed response to the client to the dialogue provider based on the content of the dialogue between the client and the first dialogue means, or the content of the dialogue between the client and the dialogue provider. (Item 21) The system further includes a second dialogue means for identifying one or more candidate symptoms based on the dialogue with the client, The information processing device according to any one of items 1 to 20, wherein the first dialogue means or the second dialogue means performs a dialogue to change the dialogue between the client and the dialoguer after identifying the candidate for one or more symptoms. (Item 22) The first dialogue means or the second dialogue means is an information processing device according to item 21 that presents the identified one or more candidate symptoms to the dialoguer. (Item 23) The first dialogue step involves engaging in a dialogue with the client using the first dialogue model, An inference step of inferring the mental state of the client by analyzing the dialogue between the client and the first dialogue model, Based on the inferred mental state, a second dialogue step is taken to have a dialogue with the client to change the dialogue between the client and the first dialogue model to a dialogue between the client and the dialogue partner. Processing method for a fingertip information processing device. (Item 24) A program that causes a computer to function as an information processing device as described in any one of items 1 through 22. [Explanation of symbols]

[0130] 101 Server 401 Dialogue Model for Inference 402 Specific Dialogue Model 403 Control Unit 404 Guessing Department 405 Medical Record Generation Department 406 Presentation section

Claims

1. A first means of dialogue for engaging in conversation with the client, The system includes an inference means for inferring the mental state of the client by analyzing the dialogue between the client and the first dialogue means, The first dialogue means is an information processing device that, based on the inferred mental state, engages in dialogue with the client to change the dialogue between the client and the first dialogue means into a dialogue between the client and the dialogue partner.

2. The information processing apparatus according to claim 1, wherein the first dialogue means presents one or more candidates for dialogue partners based on the client information.

3. The information processing device according to claim 2, wherein the client information is age or address.

4. The information processing apparatus according to claim 1, wherein the dialogue between the client and the first dialogue means is a text-based dialogue.

5. The information processing apparatus according to claim 1, further comprising control means for controlling the conduction of a dialogue between the client and the dialogue participant after a dialogue by the first dialogue means.

6. The dialogue between the client and the first dialogue means is a text-based dialogue. The information processing apparatus according to claim 5, wherein the dialogue between the client and the interlocutor is a text-based dialogue.

7. The dialogue between the client and the first dialogue means is a dialogue conducted solely through text. The information processing apparatus according to claim 5, wherein the dialogue between the client and the interlocutor includes a dialogue via voice call or video call.

8. The information processing apparatus according to claim 1, wherein the dialogue between the client and the first dialogue means and the dialogue for modification are performed on the same screen.

9. The information processing apparatus according to claim 5, wherein the dialogue between the client and the first dialogue means, the dialogue for modification, and the dialogue between the client and the dialogue participant are all performed on the same screen.

10. The information processing apparatus according to claim 5, further comprising a medical record generation means that generates a medical record based on the content of the dialogue between the client and the first dialogue means, or the content of the dialogue between the client and the dialogue means, after the dialogue between the client and the dialogue person.

11. The information processing apparatus according to claim 10, wherein the medical record generation means is controlled to a state in which only the interlocutor can download the medical record, or the medical record is presented only to the interlocutor.

12. The information processing apparatus according to claim 1, wherein the estimation means estimates the mental state of the client by analyzing the dialogue between the client and the first dialogue means and the information detected by a device owned by the client.

13. The information processing apparatus according to claim 12, wherein the information is ambient light, acceleration, gyroscope, heart rate, atmospheric pressure, compass, or blood oxygen.

14. The information processing apparatus according to claim 5, further comprising a presentation means for receiving information detected by a device owned by the client after a conversation between the client and the interlocutor, and for presenting the received information and / or the results of an analysis of the received information to the interlocutor.

15. The information processing apparatus according to claim 14, wherein the information is ambient light, acceleration, gyroscope, heart rate, atmospheric pressure, compass, or blood oxygen.

16. The information processing device according to claim 10, wherein the medical record is a medical record in SOAP format.

17. The conversation between the client and the interlocutor includes conversations conducted via voice call or video call. The information processing device according to claim 10, wherein the medical record includes findings based on the analysis results of the voice call or video call.

18. If the client agrees in the dialogue for modification, the first dialogue means will ask the dialogue partner whether or not to have a dialogue between the client and the dialogue partner. The information processing apparatus according to claim 5, wherein the control means controls the interaction between the client and the interaction person to take place when the interaction person accepts the inquiry.

19. The information processing apparatus according to claim 5, wherein the first dialogue means presents the dialogue participant with a summary of the content of the dialogue between the client and the first dialogue means.

20. The information processing apparatus according to claim 5, wherein the first dialogue means presents a proposed response to the client to the dialogue provider based on the content of a dialogue between the client and the first dialogue means, or a dialogue between the client and the dialogue provider.

21. The system further includes a second dialogue means for identifying one or more candidate symptoms based on the dialogue with the client, The information processing apparatus according to claim 1, wherein the first dialogue means or the second dialogue means performs a dialogue to change the dialogue between the client and the dialoguer after identifying the one or more candidate symptoms.

22. The information processing apparatus according to claim 21, wherein the first dialogue means or the second dialogue means presents the dialoguer with one or more specified symptom candidates.

23. The first dialogue step involves conducting a dialogue with the client using the first dialogue model, An inference step of inferring the mental state of the client by analyzing the dialogue between the client and the first dialogue model, Based on the inferred mental state, a second dialogue step is performed to have a dialogue with the client to change the dialogue between the client and the first dialogue model to a dialogue between the client and the dialogue partner. Processing method for a fingertip information processing device.

24. A program for causing a computer to function as an information processing device according to any one of claims 1 to 22.

Images