Information processing method, program, and information processor

Abstract

To provide an information processing method, a program, and an information processor that enable a care worker to output a message that should be communicated to a care requiring person or the like.SOLUTION: In a care support system, a tablet displays a screen requiring generation of a first message, based on an instruction of a care worker, and transmits an inputted care worker ID, a care requiring person ID, date information, and a notification requiring generation of a first message to a server. The server receives the care worker ID, the care requiring person ID, the date information, and the notification requiring generation of a first message transmitted by the tablet, and refers to anamnesis rows of a care requiring person DB corresponding to the received care requiring person ID, counseling rows and remark rows and the like, in order to allocate an attribute of the care worker, an attribute of the care requiring person, a selection support content item, and a generation instruction of a first message to a template, whereby a prompt is generated, and the generated prompt is inputted to a language model so that a first message that should be communicated by a care worker is acquired and transmitted to the tablet.SELECTED DRAWING: Figure 9

Description

[Technical Field]

[0001] The present disclosure relates to an information processing method, a program, and an information processing device. [Background technology]

[0002] Patent Document 1 discloses an information processing system that enables a doctor and a patient to conduct an online interactive consultation. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Application No. 2004-225272 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the invention of Patent Document 1 has a problem in that it is not possible to output a message that the caregiver should convey to the person in need of care during counseling.

[0005] In one aspect, an object is to provide an information processing method or the like that allows a caregiver to output a message to be conveyed to a person requiring care or the like. [Means for solving the problem]

[0006] An information processing method according to one aspect acquires attributes of a caregiver, attributes of a person requiring care, and a plurality of support content items selected from a total plurality of support content items for the person requiring care, generates a first message to be conveyed by the caregiver using a language model that utilizes the acquired attributes of the caregiver, attributes of the person requiring care, and the selected plurality of support content items, and outputs the generated first message. [Effects of the Invention]

[0007] In one aspect, the burden on the caregiver can be reduced by outputting a message that the caregiver should convey to the person requiring care. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is an explanatory diagram showing an overview of a care support system. [Figure 2] FIG. 2 is a block diagram illustrating an example of the configuration of a server. [Figure 3] FIG. 10 is an explanatory diagram showing the record layout of a caregiver DB. [Figure 4] FIG. 10 is an explanatory diagram showing the record layout of a support content item DB. [Figure 5] FIG. 10 is an explanatory diagram showing the record layout of a care-requiring person DB. [Figure 6] FIG. 2 is a block diagram showing an example of the configuration of a tablet. [Figure 7] FIG. 10 is an explanatory diagram showing a display example of a first first message. [Figure 8] FIG. 10 is an explanatory diagram showing a display example of a second first message. [Figure 9] 10 is a flowchart illustrating an example of a processing procedure of the care support system. [Figure 10] FIG. 10 is an explanatory diagram showing a display example of a first message according to the second embodiment. [Figure 11] FIG. 10 is an explanatory diagram showing a display example of a second first message according to the second embodiment. [Figure 12] 10 is a flowchart showing an example of a processing procedure of the care support system according to the second embodiment. [Figure 13] FIG. 2 is a block diagram showing an example of the configuration of a tablet. [Figure 14] FIG. 10 is an explanatory diagram showing an example of a display of a second message. [Figure 15] 11 is a flowchart showing an example of a processing procedure of the care support system according to the third embodiment. [Figure 16] FIG. 10 is an explanatory diagram showing the record layout of a support content item DB. [Figure 17]10A and 10B are explanatory diagrams showing examples of displaying a first message and precautions. [Figure 18] 10 is a flowchart showing an example of a processing procedure of the care support system according to the fourth embodiment. [Figure 19] FIG. 10 is an explanatory diagram showing a display example of a first message and video data. [Figure 20] 11 is a flowchart showing an example of a processing procedure of the care support system according to the fifth embodiment. [Figure 21] FIG. 10 is an explanatory diagram showing an example of a display of a third message. [Figure 22] 13 is a flowchart showing an example of a processing procedure of the care support system according to the sixth embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] (Embodiment 1) FIG. 1 is an explanatory diagram showing an overview of a care support system. The care support system acquires a prompt including attributes of a caregiver, attributes of a care recipient, and multiple support content items selected from all multiple support content items for the care recipient, as well as an instruction to generate a first message, and outputs a first message to be conveyed by the caregiver based on the acquired prompt. The care support system includes an information processing device 1 and an information processing device 2. The care support system transmits and receives information via a network N. Caregivers include care managers, interviewers, etc.

[0010] The information processing device 1 is an information processing device that processes, stores, and transmits / receives various types of information. The information processing device 1 is, for example, a server device, a personal computer, or a general-purpose tablet PC (personal computer). The information processing device 1 may also be a cloud server device that provides functions included in the information processing device 1 as cloud services. In this embodiment, the information processing device 1 will be described as a server 1.

[0011] The information processing device 2 is, for example, a server device, a smartphone, a tablet, a personal computer, a general-purpose tablet PC (personal computer), etc. In this embodiment, the information processing device 2 will be described as a tablet 2.

[0012] 2 is a block diagram showing an example of the configuration of a server. The server 1 includes a control unit 11, a storage unit 12, a communication unit 13, a mass storage unit 14, and a reading unit 15. The above-mentioned units are connected to each other via a bus. The control unit 11 is configured using one or more processors such as a CPU (Central Processing Unit), an MPU (Micro-Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-purpose computing on graphics processing units), a TPU (Tensor Processing Unit), or an AI chip (semiconductor for AI).

[0013] The storage unit 12 includes a RAM (Random Access Memory) or a ROM (Read Only Memory), etc. The storage unit 12 stores various data and the like necessary for a control program 12P (program product) executed by the control unit 11. The storage unit 12 also temporarily stores data and the like generated when the control program 12P is executed. The control unit 11 executes the control program 12P stored in the storage unit 12 as appropriate, thereby performing various information processing and control processing related to the server 1.

[0014] Furthermore, the storage unit 12 stores a language model 12M. The language model 12M is, for example, a general-purpose large-scale language model (LLM) constructed by performing unsupervised pre-learning using a large set of sentences and images. The language model 12M can be configured using, for example, an algorithm such as GPT (Generative Pre-trained Transformer)-3, GPT-3.5, GPT-4, RWKV (Receptance Weighted Key Value), PaLM2, or LLaMa (Large Language Model Meta AI). When a prompt including attributes of a caregiver, attributes of a care recipient, multiple support content items selected from multiple support content items for the care recipient (hereinafter referred to as selected support content items), and an instruction to generate a first message is input, the language model 12M performs a calculation to generate a first message to be conveyed to the caregiver in response to the prompt, and outputs the generated first message. The first message may be a conversational sentence including a response from the care recipient in addition to the message to be conveyed by the caregiver.

[0015] The attributes of the caregiver include, for example, age, gender, years of experience, qualifications held, supervisor experience, training received, and level of proficiency. The attributes of the care recipient include, for example, name, age, gender, level of care required, household, living arrangement, medical history, level of independence in daily living, level of independence in cognitive function, motivation, and understanding of the disease and its needs. The total support content items, for example, describe matters for an inexperienced caregiver to provide a certain level of care management, and are divided into 44 support content items. The total support content items systematically describe the knowledge of experienced caregivers. The selected support content items are, for example, multiple support content items selected from the total support content items. The total support content items and the selected support content items will be described later. In this embodiment, an example in which the total support content items are divided into 44 support content items will be described, but is not limited to this.

[0016] The language model 12M may be configured by combining a plurality of algorithms. Instead of storing the language model 12M in the storage unit 12, the control unit 11 may access a language processing server that stores the language model 12M and input a prompt to the language model 12M.

[0017] The communication unit 13 is a communication module that transmits and receives information to and from the tablet 2 via the network N. The mass storage unit 14 includes RAM (Random Access Memory) or ROM (Read Only Memory). The mass storage unit 14 stores templates for creating prompts including attributes of the caregiver, attributes of the care recipient, selected support content items, and instructions for generating a message. The templates include a first template and a second template. The first template is used when counseling is conducted between the caregiver and the care recipient. The second template is used when counseling is conducted between the caregiver, the care recipient, and a third party who may be a key person. Counseling includes assessment, monitoring, interviews, etc. The mass storage unit 14 also stores a caregiver DB 141, a support content item DB 142, and a care recipient DB 143, which will be described later.

[0018] In this embodiment, the storage unit 12 and the mass storage unit 14 may be configured as an integrated storage device. The mass storage unit 14 may be configured by a plurality of storage devices. The mass storage unit 14 may be an external storage device connected to the server 1.

[0019] The reading unit 15 reads information stored in the portable storage medium 1a. The portable storage medium 1a is, for example, a CD (Compact Disc), a DVD (Digital Versatile Disc), a USB (Universal Serial Bus) memory, or an SD (Secure Digital). The reading unit 15 reads the control program 12P from the portable storage medium 1a. The control unit 11 stores the read control program 12P in the storage unit 12. The control unit 11 may download the control program 12P from another computer via the network N. In this case, the control unit 11 stores the downloaded control program 12P in the storage unit 12. The control unit 11 may store the read control program 12P in the mass storage unit 14.

[0020] In this embodiment, the server 1 may be configured with multiple servers. The server 1 may be a virtual machine virtually constructed by software within a single device. The server 1 may be a local server installed within the facility where the server 1 is located. The server 1 may be a cloud server connected for communication via a network N. Furthermore, the control program 12P may be executed on a single server, or may be distributed and executed on multiple servers interconnected via the network N.

[0021] FIG. 3 is an explanatory diagram showing the record layout of the caregiver DB. The caregiver DB 141 includes a caregiver ID column, an age column, a gender column, years of experience column, qualifications column, supervisor experience column, training taken column, and proficiency column. The caregiver ID column stores a caregiver ID for identifying the caregiver. The age column stores the age of the caregiver. The gender column stores the gender of the caregiver. The years of experience column stores the years of experience of the caregiver. The qualifications column stores the qualifications held by the caregiver (e.g., nursing welfare worker, nurse, or public health nurse). The supervisor experience column stores whether the caregiver has supervisor experience. The training taken column stores training taken by the caregiver (e.g., specialist I and specialist II). The proficiency column stores the proficiency of the caregiver. The proficiency of the caregiver may be, for example, expert, intermediate, or beginner. The caregiver DB 141 may store the name, password, etc. of the caregiver in association with the caregiver ID.

[0022] FIG. 4 is an explanatory diagram showing the record layout of the support content item DB. The support content item DB 142 includes a support content item column and an assessment item column. The support content item column stores 44 support content items. The assessment item column stores assessment items. The assessment items are subdivisions of the support content items and are associated with all 44 support content items. In FIG. 4, "1. Support for understanding disease management" in the total support content item column is associated with "degree of understanding of the patient / family, etc. regarding the disease," "daily food intake status," "daily water intake status," etc. in the assessment item column and are stored.

[0023] FIG. 5 is an explanatory diagram showing the record layout of the care recipient DB. The care recipient DB 143 includes a care recipient ID column, a name column, an age column, a gender column, a level of care required column, a household column, a living arrangement column, a medical history column, a level of independence in daily living column, a level of independence in cognitive function column, a motivational status column, a disease understanding / need understanding column, a counseling column, a remarks column, and a column of selected support content items. The care recipient ID column stores a care recipient ID for identifying the care recipient. The name column stores the name of the care recipient. The age column stores the age of the care recipient. The gender column stores the gender of the care recipient. The level of care required column stores the level of care required by the care recipient. The level of care required by the care recipient is classified into five levels, for example, from 1 to 5. The household column stores the household of the care recipient. The household of the care recipient is classified into, for example, "living alone," "elderly only," and "with cohabitants other than elderly." The living arrangement column stores the living arrangement of the care recipient. Living arrangements are categorized, for example, into "home (detached house)," "home (apartment building)," "paid nursing home," "housing for the elderly with services," "group home," "nursing facility," and "other." The medical history column stores the medical history of the care recipient. Medical history is categorized, for example, into "cerebrovascular disease," "heart disease," "fracture / fall," "cancer," "respiratory disease," "joint disease," "diabetes," "dementia," "Parkinson's disease," "spinal cord injury," "visual / hearing impairment," and "speech disorder."

[0024] The degree of independence in daily living column stores the degree of independence in daily living of the care recipient. The degree of independence in daily living of the care recipient is a scale that evaluates how independently the care recipient is able to live an independent life, and is classified, for example, as "J1," "J2," "A1," "A2," "B1," "B2," "C1," and "C2." The degree of cognitive independence column stores the degree of independence in daily living of the care recipient with dementia. The degree of independence in daily living of the care recipient with dementia is classified, for example, as "Not applicable (not dementia)," "Ia," "Ib," "IIa," "IIb," "IIIa," "IIIb," and "M." The motivation status column stores the activity motivation of the care recipient. The activity motivation of the care recipient is classified, for example, as "high," "normal," and "low." The disease understanding / need understanding column stores the level of interest of the care recipient in the disease, etc. The level of interest of the care recipient in the disease, etc. is classified, for example, as "high," "normal," and "low." The counseling column stores information regarding counseling. If the person requiring care receives counseling alone, a "○" is stored in the counseling column. If the person requiring care does not receive counseling alone, an "×" and a "third party who could be a key person" are stored in the counseling column. Third parties who could be key persons include the family of the person requiring care, or supporters other than family members. The notes column stores additional information regarding the level of care required. If there are no additional information regarding the person requiring care, no information may be entered. The selected support content item column stores the selected support content items shown in Figure 4. For example, the selected support content item column stores selected support content items that more than 50% of multiple experienced caregivers recognize as necessary for the person requiring care.

[0025] A method for determining the support content items to be selected for a care recipient will be described. The control unit 21 displays a screen (not shown) for determining the support content items to be selected based on instructions from a care recipient using the care support system. The experienced caregiver inputs the care recipient ID of the relevant care recipient via the input unit 24. The experienced caregiver may input the name of the care recipient instead of the care recipient ID. The control unit 21 transmits the care recipient ID and a notification that 44 support content items will be disclosed to the server 1. The control unit 11 receives the care recipient ID and the notification that 44 support content items will be disclosed transmitted from the tablet 2. The control unit 11 reads out the 44 support content items from the support content item DB 142. The control unit 11 transmits the read support content items to the tablet 2. The control unit 21 receives the support content items transmitted from the server 1. The control unit 21 displays the received support content items on the display unit 25. The experienced caregiver selects the support content items necessary for the relevant care recipient from the displayed support content items. A plurality of experienced caregivers select the above-mentioned support content items. The control unit 21 transmits the selected support content items to the server 1. The control unit 11 receives the support content items transmitted from the tablet 2. For example, the control unit 11 determines, as the selected support content items, support content items that 50% or more of the plurality of experienced caregivers recognize as necessary for the care recipient out of 44 support content items.

[0026] The record for the care recipient ID Y211 in Figure 5 contains the following information: name "A," age "70s," gender "male," level of care required "1," household status "living alone," living arrangement "room (detached house)," medical history "cerebrovascular disease," level of independence in daily living "A2 (mild hemiplegia)," level of cognitive independence "not applicable (not dementia)," motivation status "high," understanding of disease and necessity "high," counseling status "yes," notes "older sister and niece come to check on the patient about once a month," "uses day care services," "does medical visits alone," "caregiver has recently changed," and optional support content items "support for understanding disease management," "support for understanding fluid intake," and "support for ongoing medical examinations and treatment."

[0027] 6 is a block diagram showing an example of the configuration of a tablet. The tablet 2 includes a control unit 21, a storage unit 22, a communication unit 23, an input unit 24, and a display unit 25. The above-mentioned units are connected to each other via a bus. The control unit 21 is configured using one or more processors such as a CPU, an MPU, or a GPU.

[0028] The storage unit 22 includes RAM, ROM, etc. The storage unit 22 stores various data, etc. required for the control program 22P (program product) executed by the control unit 21. The storage unit 22 temporarily stores data, etc. generated when the control program 22P is executed. The control unit 21 executes the control program 22P stored in the storage unit 22 as needed to perform information processing and control processing related to the tablet 2.

[0029] The communication unit 23 is a communication module that transmits and receives information to and from the server 1 via the network N. The input unit 24 accepts operational inputs from the caregiver. The input unit 24 includes, for example, a keyboard, a mouse, and a microphone for voice input. The display unit 25 is a liquid crystal display or an organic EL display. The display unit 25 outputs various information in accordance with instructions from the control unit 21. The input unit 24 and the display unit 25 may be a touch panel configured as an integrated unit.

[0030] The processing of the care support system of embodiment 1 will be described. The control unit 21 displays a screen (not shown) requesting the generation of a first message based on an instruction from the caregiver. The screen requesting the generation of the first message includes fields for inputting a caregiver ID, a care recipient ID, and date and time information. Before performing counseling (monitoring), the caregiver inputs their own caregiver ID, the care recipient ID of the care recipient they are responsible for, and date and time information via the input unit 24. The date and time information is, for example, the date and time when counseling will be conducted. The caregiver may input their name and the name of the care recipient instead of the caregiver ID and the care recipient ID. The control unit 21 transmits the input caregiver ID, the care recipient ID, the date and time information, and a notification requesting the generation of the first message to the server 1.

[0031] The control unit 11 receives the caregiver ID, the care recipient ID, the date and time information, and a notification requesting the generation of a first message transmitted from the tablet 2. The control unit 11 determines whether or not two-person counseling between the caregiver and the care recipient is possible by referring to the medical history column, the counseling column, the remarks column, etc. in the care recipient DB 143 corresponding to the received care recipient ID. In FIG. 5, "cerebrovascular disease" is stored in the medical history column corresponding to the care recipient ID (Y211), "〇" is stored in the counseling column, and "the care recipient will go to the hospital alone" is stored in the remarks column. In this case, the control unit 11 determines that two-person counseling between the caregiver and the care recipient is possible based on the contents of the medical history column, the counseling column, the remarks column, etc. Also, in FIG. 5, "dementia / diabetes" is stored in the medical history column corresponding to the care recipient ID (Y212), "× (daughter)" is stored in the counseling column, and "the daughter is in charge of housework, etc." is stored in the remarks column. In this case, the control unit 11 determines that two-way counseling between the caregiver and the care recipient is not possible based on the contents of the medical history column, the counseling column, the remarks column, etc. (e.g., the state of dementia, etc.). When the control unit 11 determines that two-way counseling between the caregiver and the care recipient is possible, it reads a first template from the mass storage unit 14. When the control unit 11 determines that two-way counseling between the caregiver and the care recipient is not possible, it reads a second template from the mass storage unit 14. Furthermore, the control unit 11 refers to the counseling column, etc. in the care recipient DB 143 corresponding to the received care recipient ID, and identifies a third party (the daughter) who could be a key person. The subsequent processing is the same whether it is determined that two-way counseling between the caregiver and the care recipient is possible or not.

[0032] The control unit 11 refers to the caregiver DB 141 and reads out the attributes of the caregiver corresponding to the caregiver ID. The control unit 11 refers to the care recipient DB 143 and reads out the attributes of the care recipient corresponding to the care recipient ID and the selected support content items. The control unit 11 generates a prompt to be input into the language model 12M by applying the read attributes of the caregiver, the attributes of the care recipient, the selected support content items, and the instruction to generate the first message to the first template or the second template.

[0033] For example, if the caregiver ID K111 and the care recipient ID Y211 are entered, the caregiver's attributes are "female in her 50s, one year of experience, certified care worker, no experience as a supervisor, no training, and beginner level of proficiency." The care recipient's attributes are "Mr. A, male in his 70s, level of care required: level 1, lives alone, lives in a detached house, has a history of cerebrovascular disease, has a daily living independence level of A2 (with mild hemiplegia), has a cognitive independence level of not applicable (no dementia), is highly motivated, has a high understanding of the disease and its needs, conducts counseling with the caregiver and the care recipient, has an older sister and niece visit approximately once a month to check on the patient, uses day care services, makes medical visits alone, and has recently had a change in caregiver." The selected support content items are "support for understanding disease management, support for understanding fluid intake, and support for ongoing medical examinations and treatment." The instruction for generating the first message is, "Generate a message that you should convey to the care recipient." In this case, the generated prompt is, "You are a caregiver for a woman in her 50s, and have a certified care worker license and over one year of experience. Your level of proficiency as a caregiver is beginner level. The care recipient is Mr. A, a man in his 70s with a care level of "1." He lives alone in his room (detached house), has a history of cerebrovascular disease, has a daily living independence level of A2 (with mild hemiplegia), is highly motivated, has a high understanding of his disease and needs, uses day care services, and makes medical visits independently. You and the care recipient have only met a few times due to changes in caregivers. The care recipient's sister and niece come to check on him about once a month. Counseling is also conducted between the caregiver and the care recipient. Generate a message that you should convey to the care recipient. When generating the message, consider supporting his understanding of disease management, helping him understand his hydration status, and supporting his ongoing medical examinations and treatment."

[0034] For example, if the caregiver ID K112 and the care recipient ID Y212 are entered, the caregiver's attributes are "female in her 40s, 10 years of experience, licensed nurse, experienced supervisor, no training, and skilled." The care recipient's attributes are "Ms. B, female in her 80s, level of care required "3," living with non-elderly people (daughter and family), living in a separate home, history of dementia and diabetes, level of independence in daily living J2, level of cognitive independence IIIb, low motivation, low understanding of disease and needs, counseling between the caregiver and care recipient alone is deemed ineffective, so counseling will be conducted between the caregiver, care recipient, and a third party (daughter) who could be a key person, housework, etc., and the daughter is in charge of household chores, etc., and the caregiver has recently been replaced." The selected support content items are "support for understanding disease management, support for understanding fluid intake, and support for ongoing medical examinations and treatment." The instruction for generating the first message is, "Generate a message to convey to the daughter of the care recipient." In this case, the generated prompt is, "You are a caregiver for a woman in her 40s, with a nursing license and over 10 years of experience. Your level of proficiency as a caregiver is expert. The care recipient is Ms. B, a woman in her 80s with a care level of "3." She lives with her daughter and her family in a private home. She has a history of dementia and diabetes, a daily living independence level of J2, a cognitive independence level of IIIb, low motivation, and a low understanding of her illness and needs. Her daughter handles housework and other tasks. You and the care recipient have only met a few times due to changes in caregivers. Furthermore, because a two-way counseling session between the caregiver and the care recipient is deemed ineffective, counseling will be conducted between the caregiver, the care recipient, and a third party (the daughter) who could be a key person. Generate a message to convey to the care recipient's daughter." When creating messages, consider supporting understanding of disease management, helping to understand fluid intake, and supporting ongoing medical consultations and treatment.

[0035] The control unit 11 inputs the generated prompt into the language model 12M to acquire a first message to be conveyed by the caregiver. The control unit 11 stores the first message in the mass storage unit 14 in association with the caregiver ID, the care-requiring person ID, and date and time information. The control unit 11 transmits the acquired first message to the tablet 2.

[0036] The control unit 21 receives the first message sent from the server 1. The control unit 21 displays the received first message on the display unit 25. FIG. 7 is an explanatory diagram showing an example of the display of the first first message. The screen d01 in FIG. 7 includes an order column d11 and a first message column d12. The order column d11 displays the order in which the first messages are to be transmitted. The first message column d12 displays the first messages output by the language model 12M. The first messages are displayed in the order in which the first messages are to be transmitted. There are a total of eight first messages displayed.

[0037] The first message is, "Hello, Mr. A. We've met a few times before, and I hope we can continue to work together." The second message is, "Thank you. First, Mr. A, I'd like you to learn a little more about cerebrovascular disease. Do you know the basics of the disease, treatment options, and things to keep in mind in daily life?" The third message is, "Cerebrovascular disease refers to a condition in which there is a problem with the blood vessels in the brain. Specifically, this includes stroke and cerebral infarction. Mr. A has a history of cerebrovascular disease, so he needs to see a doctor regularly. By following his doctor's instructions, we can increase the effectiveness of treatment." The fourth message is, "Thank you." The fifth message is, "Next, please tell me about your fluid intake. How much fluid do you drink per day, Mr. A? Also, what is your favorite drink?" The sixth message is, "Thank you. Please make sure to drink fluids frequently to prevent dehydration." The seventh first message is, "Finally, please tell us about ongoing medical treatment and treatment. Is Mr. A in a position to regularly visit a hospital or other facility?" The eighth first message is, "I understand. We will continue to monitor changes in Mr. A's condition and, as necessary, reset goals and adjust response methods."

[0038] Although an example in which only the first message to be conveyed by the caregiver is generated has been described in FIG. 7, the present invention is not limited to this. The language model 12M may output an expected response from the care recipient in a conversational format in addition to the first message to be conveyed by the caregiver. In this case, the language model 12M generates, for example, the response of the care recipient (Mr. A) between the first first message (Hello, Mr. A. ~ Nice to meet you.) and the second first message (Thank you. First of all...), "Yes, I understand. I look forward to working with you."

[0039] Fig. 8 is an explanatory diagram showing an example of the display of the second first message. Screen d02 in Fig. 8 includes a sequence column d21 and a first message column d22. The display format in Fig. 8 is the same as that in Fig. 7. There are a total of seven first messages displayed.

[0040] The first message reads, "Hello, Mr. B's daughter. I've met you several times before, and I look forward to working with you in the future. I've spoken with Mr. B, but there were some parts of the situation that I didn't fully understand, so could your daughter please explain?" The second message reads, "Thank you. First, let me confirm Mr. B's condition. Mr. B has a history of dementia and diabetes, and his daily living independence level is J2 and his cognitive independence level is IIIb. He seems to have a low level of understanding of disease management and its necessity." The third message reads, "I understand. Let's help Mr. B deepen his understanding of disease management. I would like your daughter to inform Mr. B of the importance of regular hospital visits. I would also like your daughter to support Mr. B in following his doctor's instructions." The fourth message reads, "Next, please tell me about Mr. B's fluid intake. How much fluid does Mr. B consume per day? Also, what is his favorite drink?" The fifth message reads, "Thank you. I would like your daughter to tell Mr. B to drink fluids frequently." The sixth first message is, "Finally, please tell us about ongoing medical treatment and treatment. Is Mr. B in a position to regularly visit a hospital or other facility?" The seventh first message is, "I understand. We will continue to monitor changes in Mr. B's condition and, as necessary, reset goals and adjust response methods."

[0041] 9 is a flowchart showing an example of a processing procedure of the care support system. Based on an instruction from a caregiver, the control unit 21 displays a screen requesting the creation of a first message (step S101). The control unit 21 transmits the input caregiver ID, care-requiring person ID, date and time information, and a notification requesting the creation of a first message to the server 1 (step S102).

[0042] The control unit 11 receives the caregiver ID, the care recipient ID, the date and time information, and a notification requesting the generation of a first message transmitted from the tablet 2 (step S201). The control unit 11 refers to the medical history column, the counseling column, the remarks column, etc. in the care recipient DB 143 corresponding to the received care recipient ID, and determines whether counseling between the caregiver and the care recipient is possible (step S202). If the control unit 11 determines that the care recipient will receive counseling (step S202: YES), it reads out a first template from the mass storage unit 14 (step S203). The control unit 11 refers to the caregiver DB 141 and reads out the attributes of the caregiver corresponding to the caregiver ID (step S206). Alternatively, if the control unit 11 determines that the care recipient will not receive counseling (step S202: NO), it reads out a second template from the mass storage unit 14 (step S204). The control unit 11 refers to the counseling sequence in the care recipient DB 143 corresponding to the received care recipient ID, and identifies a third party who could be a key person (step S205). The control unit 11 proceeds to the process at step S206. The control unit 11 refers to the care recipient DB 143 and reads out the attributes of the care recipient corresponding to the care recipient ID and the selected support content items (step S207). The control unit 11 generates a prompt to be input to the language model 12M by applying the attributes of the caregiver, the attributes of the care recipient, the selected support content items, and the instruction to generate the first message to the first template or the second template (step S208). The control unit 11 inputs the generated prompt into the language model 12M to acquire the first message to be conveyed by the caregiver (step S209). The control unit 11 stores the first message in the mass storage unit 14 in association with the caregiver ID, the care recipient ID, and date and time information (step S210). The control unit 11 transmits the acquired first message to the tablet 2 (step S211).

[0043] The control unit 21 receives the first message transmitted from the server 1 (step S103). The control unit 21 displays the received first message on the display unit 25 (step S104).

[0044] According to the first embodiment, the care support system can output a first message to be conveyed by the caregiver based on the attributes of the caregiver, the attributes of the care recipient, and the selected support content items.

[0045] According to the first embodiment, the caregiver can efficiently provide counseling to the person requiring care, etc. by using the first message output by the care support system.

[0046] According to the first embodiment, the care support system can reduce the burden on the caregiver by outputting the first message that the caregiver should convey.

[0047] (Embodiment 2) In the second embodiment, a method will be described in which the care support system outputs a first message to be conveyed by the caregiver using a plurality of assessment items in addition to the attributes of the caregiver, the attributes of the care recipient, and the selected support content items.

[0048] The control unit 21 transmits the caregiver ID, the care recipient ID, date and time information, and a notification requesting the generation of a first message to the server 1 in the same procedure as in the first embodiment. The control unit 11 receives the caregiver ID, the care recipient ID, date and time information, and a notification requesting the generation of a first message transmitted from the tablet 2 in the same procedure as in the first embodiment. The control unit 11 references the medical history column, the counseling column, the remarks column, etc. in the care recipient DB 143 corresponding to the received care recipient ID and determines whether counseling between the caregiver and the care recipient is possible. If the control unit 11 determines that counseling between the caregiver and the care recipient is possible, it reads a first template from the mass storage unit 14. If the control unit 11 determines that counseling between the caregiver and the care recipient is not possible, it reads a second template from the mass storage unit 14. Furthermore, the control unit 11 references the counseling column, etc. in the care recipient DB 143 corresponding to the received care recipient ID to identify a third party who could be a key person. The subsequent processing is the same whether it is determined that two-way counseling between the caregiver and the person requiring care is possible or not.

[0049] The control unit 11 refers to the caregiver DB 141 and reads out the attributes of the caregiver corresponding to the caregiver ID. The control unit 11 refers to the care recipient DB 143 and reads out the attributes of the care recipient corresponding to the care recipient ID and the selected support content item. The control unit 11 acquires an assessment item corresponding to the read selected support content item from the support content item DB 142. For example, when the selected support content item "support for understanding disease management" is read out from the care recipient DB 143, the control unit 21 acquires the assessment item "daily fluid intake status" corresponding to "support for understanding disease management" from the support content item DB 142. In this embodiment, an example in which the control unit 21 acquires one assessment item from one selected support content item will be described, but this is not limiting. The control unit 21 may acquire multiple assessment items from one selected support content item.

[0050] The control unit 11 generates a prompt to be input to the language model 12M based on the read attributes of the caregiver, the attributes of the care recipient, the selected support content items, the acquired assessment items, and the instruction to generate the first message. The prompt to be input to the language model 12M is generated, for example, by applying the attributes of the caregiver, the attributes of the care recipient, the selected support content items, the acquired assessment items, and the instruction to generate the first message to a first template or a second template.

[0051] For example, if the caregiver ID K111 and the care recipient ID Y211 are entered, the caregiver's attributes are "female in her 50s, one year of experience, certified care worker, no experience as a supervisor, no training, and beginner level of proficiency." The care recipient's attributes are "Mr. A, male in his 70s, level of care required: level 1, lives alone, lives in a detached house, has a history of cerebrovascular disease, has a daily living independence level of A2 (with mild hemiplegia), has a cognitive independence level of not applicable (no dementia), is highly motivated, has a high understanding of the disease and its needs, conducts counseling with the caregiver and the care recipient, has an older sister and niece visit approximately once a month to check on the patient, uses day care services, makes medical visits alone, and has recently had a change in caregiver." The selected support content items are "support for understanding disease management, support for understanding fluid intake, and support for ongoing medical examinations and treatment." The assessment items are "daily fluid intake status, level of understanding of the patient and family regarding fluid and salt amounts, and daily food intake status," which correspond to the selected support content items "support for understanding disease management, support for understanding fluid intake status, and support for ongoing medical examinations and treatment." The instruction for generating the first message is "Generate the message you would like to convey to the person requiring care."

[0052] In this case, the prompt that is generated is, "You are a caregiver for a woman in her 50s, and have a certified care worker license and over one year of experience. Your level of proficiency as a caregiver is beginner level. The person requiring care is Mr. A, a man in his 70s with a care level of "1." He lives alone in his room (detached house). He has a history of cerebrovascular disease. His level of independence in daily living is A2 (he has mild hemiplegia). He is highly motivated, has a good understanding of his illness and needs, uses day care services, and goes to the hospital alone. You and the person requiring care have only met a few times due to changes in caregivers. The person's sister and niece come to check on him about once a month. Counseling is also conducted between the caregiver and the person requiring care." Generate a message that you should convey to the person requiring care. When generating this message, consider supporting understanding of disease management, helping to understand his hydration status, and supporting ongoing medical examinations and treatment. Furthermore, regarding supporting understanding of disease management, include a message to check his daily hydration status. When providing support for understanding fluid intake, please include a message to check the level of understanding of the individual and their family regarding fluid and salt intake. When providing support for ongoing medical examinations and treatment, please include a message to check the status of daily food intake.

[0053] For example, if the caregiver ID K112 and the care recipient ID Y212 are entered, the caregiver's attributes are "female in her 40s, 10 years of experience, licensed nurse, experienced supervisor, no training, and skilled." The care recipient's attributes are "Ms. B, female in her 80s, level of care required "3," living with non-elderly family members (daughter and family), living in a separate home, history of dementia and diabetes, level of independence in daily living J2, level of cognitive independence IIIb, low motivation, low understanding of disease and needs, counseling between the caregiver and care recipient alone is deemed ineffective, so counseling will be conducted between the caregiver, care recipient, and a potential key person (daughter), the daughter is in charge of housework, etc., and the caregiver has recently been replaced." The selected support content items are "support for understanding disease management, support for understanding fluid intake, and support for ongoing medical examinations and treatment." The assessment items are "daily fluid intake status, level of understanding of the patient and family regarding fluid and salt amounts, and daily food intake status," which correspond to the selected support content items "support for understanding disease management, support for understanding fluid intake status, and support for ongoing medical examinations and treatment." The instruction for generating the first message is "Generate a message that you would like to convey to the daughter of the person requiring care."

[0054] In this case, the prompt generated would be, "You are a caregiver for a woman in her 40s, with a nursing license and over 10 years of experience. Your level of proficiency as a caregiver is expert. The care recipient is Ms. B, a woman in her 80s with a care level of "3." She lives in a detached house with her daughter and her family. She has a history of dementia and diabetes. Her level of independence in daily living is J2, her level of cognitive independence is IIIb, her motivation is low, and she has a low understanding of her illness and needs. Her daughter is responsible for housework, etc. You and the care recipient have only met a few times due to changes in caregivers. Furthermore, because counseling between the caregiver and the care recipient alone is deemed ineffective, counseling will be conducted between the caregiver, the care recipient, and a third party (your daughter) who could be a key person. Generate a message to convey to the care recipient through their daughter. When generating this message, consider supporting understanding of disease management, helping to understand hydration status, and supporting ongoing medical visits and treatment. Furthermore, for support in understanding disease management, please include a message to check the status of daily fluid intake. For support in grasping fluid intake, please include a message to check the level of understanding of the patient and family regarding fluid and salt intake. For support in ongoing medical examinations and treatment, please include a message to check the status of daily food intake.

[0055] The control unit 11 inputs the generated prompt into the language model 12M to acquire a first message to be conveyed by the caregiver. The control unit 11 stores the first message in the mass storage unit 14 in association with the caregiver ID, the care-requiring person ID, and date and time information. The control unit 11 transmits the acquired first message to the tablet 2.

[0056] The control unit 21 receives the first message transmitted from the server 1. The control unit 21 displays the received first message on the display unit 25.

[0057] Fig. 10 is an explanatory diagram showing an example of the display of the first first message according to embodiment 2. A screen d03 in Fig. 10 includes a sequence column d31 and a first message column d32. The display format in Fig. 10 is the same as that in Fig. 7. A total of 12 first messages are displayed.

[0058] The first message is, "Mr. A, today I'd like to learn more about fluids. Some of what I've said to you before may overlap, but please let me know. Thank you." The second message is, "Has the doctor at the hospital told you anything about fluids or diet?" The third message is, "I see. The doctor hasn't given you any specific instructions, has he?" The fourth message is, "It's getting hotter, so I'm worried about heatstroke. How much fluid do you usually drink?" The fifth message is, "Do you usually use a cup or drink from a plastic bottle?" The sixth message is, "I'll borrow some. It holds about 350 ml. When you drink at one time, do you fill it to the brim and drink it all?" The seventh message is, "I see. So you're getting about 250 ml of fluid at a time. What do you usually drink from it?" The eighth message is, "You drink a lot of water. Do you drink water with meals as well?" The ninth first message is, "How often do you drink outside of meals?" The tenth first message is, "So it looks like you're drinking about 1700ml nine times a day. You're getting plenty of fluids." The eleventh first message is, "How do you prepare your meals?" The twelfth first message is, "So you use a meal delivery service. That sounds like you're getting a well-balanced diet. If you have a menu, can I see it next time?"

[0059] Fig. 11 is an explanatory diagram showing an example of the display of a second first message according to embodiment 2. A screen d04 in Fig. 11 includes a sequence column d41 and a first message column d42. The display format in Fig. 11 is the same as that in Fig. 7. A total of 10 first messages are displayed.

[0060] The first message is, "I spoke with B-san, but there were some parts of the situation that were difficult to understand, so could your daughter please explain?" The second message is, "Today, I'd like to ask you about your usual fluid intake. Some of what I've said may overlap with what the previous doctor told me, but please let me know. Thank you." The third message is, "Has the doctor at the hospital said anything about fluid intake or diet?" The fourth message is, "I see." The fifth message is, "You mentioned sweets, but does your daughter prepare your daily meals?" The sixth message is, "Are you concerned about your daughter's weight gain or loss recently? I'll keep an eye on her weight in the future, in cooperation with the day care center." The seventh message is, "As the hot weather approaches, I'm worried about heatstroke. Does your family know how much fluid your daughter is consuming on a daily basis?" The eighth message is, "How much does she drink per day?" The ninth first message is, "Are you taking in any other fluids besides tea? Do you drink tea with meals? How about miso soup?" The tenth first message is, "I apologize for the inconvenience, but please confirm your approximate daily intake."

[0061] Fig. 12 is a flowchart showing an example of a processing procedure of the care support system according to the second embodiment. In the flowchart shown in Fig. 12, step S208 in the processing shown in Fig. 9 is replaced with steps S301 and S302. Description of the same steps as in Fig. 9 will be omitted. The control unit 11 acquires assessment items corresponding to the read selected support content items from the support content item DB 142 (step S301). The control unit 11 generates a prompt to be input to the language model 12M by applying the read attributes of the caregiver, the attributes of the care-requiring person, the selected support content items, the acquired assessment items, and an instruction to generate a first message to a first template or a second template (step S302). The control unit 21 proceeds to the processing of step S209.

[0062] According to the second embodiment, the care support system can output a first message that the caregiver should convey based on the attributes of the caregiver, the attributes of the care recipient, the selected support content items, and the assessment items.

[0063] According to the second embodiment, the care support system can output a first message according to a specific situation by using assessment items in addition to attributes of the caregiver, attributes of the care recipient, and selected support content items.

[0064] (Embodiment 3) In the third embodiment, a method will be described in which the care support system outputs a second message based on image data of a care-requiring person and a first message. The second message indicates the emotion or understanding of the care-requiring person.

[0065] In the third embodiment, a care recipient will be described as an example, but the present invention is not limited to this. The third embodiment can also be applied to a relative (daughter) of the care recipient. In the third embodiment, the second message will be described as an example showing the understanding level of the care recipient.

[0066] FIG. 13 is a block diagram showing an example of the configuration of a tablet. The tablet 2 includes an imaging device 26. The imaging device 26 includes a camera 26a and a microphone 26b. The camera 26a is an imaging device such as a CCD (Charge Coupled Device) camera or a CMOS (Complementary Metal Oxide Semiconductor) camera, and captures an image of the care-requiring person during counseling. The microphone 26b collects audio during counseling. The imaging device 26 acquires video data of the care-requiring person during counseling using the camera 26a and the microphone 26b. The video data includes image data and audio data. The imaging device 26 does not have to be built into the tablet 2, but may be connected externally via short-range wireless communication such as Bluetooth (registered trademark) or Wi-Fi (registered trademark).

[0067] The caregiver places the tablet 2 in a desired position. The desired position is, for example, a position where an image of the person requiring care can be captured during counseling. The caregiver may hold the tablet 2 in their hand. The control unit 21 acquires video data using the imaging device 26. The control unit 21 acquires date and time information in addition to the video data. After the counseling session ends, the control unit 21 transmits the acquired video data and date and time information to the server 1.

[0068] The control unit 11 receives the video data and date and time information transmitted from the tablet 2. The control unit 11 reads out a first message corresponding to the received date and time information from the mass storage unit 14. The control unit 11 inputs the read first message, video data, and second message generation instruction to the language model 12M. The second message generation instruction is, for example, "Please generate the level of understanding of the care recipient using the first message and video data." The language model 12M separates the input video data into image data and audio data. The language model 12M extracts features for each of the first message, image data, audio data, and second message generation instruction. The language model 12M extracts features such as the facial expression, gaze, and pupils of the care recipient from the image data. The language model 12M extracts features such as the voice intonation, voice volume, and speaking speed of the care recipient from the audio data.

[0069] The method by which the language model 12M extracts features from image data and audio data is not limited to the above example. The language model 12M may extract features from image data and audio data by using multiple convolutional neural networks (CNNs) equipped with convolutional layers and pooling layers. In this case, the language model 12M extracts features related to the image data by inputting image data for each frame into the convolutional neural network. The language model 12M also converts audio data into text data and converts the converted text data into language vectors. A method such as BOW (Bag of Words) or distributed representation is used to convert text data into language vectors. The language model 12M extracts features related to the audio data by inputting the converted language vectors into the convolutional neural network.

[0070] The language model 12M combines the extracted features to synthesize a single sequence of the first message, image data, audio data, and an instruction to generate a second message. The language model 12M performs a calculation to generate a second message from the combined sequence and outputs the generated second message. For example, when the first message shown in FIG. 7, video data of a care recipient undergoing counseling, and an instruction to generate a second message are input, the language model 12M outputs a second message stating, "Mr. A does not fully understand the content of the counseling." The control unit 11 acquires the second message output by the language model 12M. The control unit 11 stores the video data and the second message in the mass storage unit 14, linking them to the caregiver ID, the care recipient ID, date and time information, and the first message. The control unit 11 transmits the acquired second message to the tablet 2.

[0071] The control unit 21 receives the second message transmitted from the server 1. The control unit 21 displays the received second message on the display unit 25. FIG. 14 is an explanatory diagram showing an example of the display of the second message. The screen d05 in FIG. 14 includes a second message field d51. The second message output by the language model 12M is displayed in the second message field d51. The second message field d51 in FIG. 14 displays the second message, "Mr. A does not fully understand the content of the counseling." In the present embodiment, an example has been described in which the screen d05 includes only the second message field d51, but this is not limiting. The screen d05 may also include an order field, a first message field, a notes field, a video data field, etc. In this case, the second message field d51 is displayed around the order field, the first message field, the notes field, the video data field, etc.

[0072] In the third embodiment, the control unit 11 outputs the second message using the language model 12M (LLM), but the present invention is not limited to this. The control unit 11 may also use an emotion estimation model or an understanding estimation model using CNN or the like. In this case, the emotion estimation model or the understanding estimation model is configured as a multimodal model including multiple input layers and intermediate layers for extracting features from various input data such as the first message, image data, and audio data. The emotion estimation model or the understanding estimation model outputs the second message by combining the features extracted from the multiple intermediate layers in a fully connected layer before the output layer.

[0073] 15 is a flowchart showing an example of a processing procedure of the care support system according to embodiment 3. The control unit 21 acquires video data using the imaging device 26 (step S301). The control unit 21 acquires date and time information in addition to the video data (step S302). After counseling ends, the control unit 21 transmits the acquired video data and date and time information to the server 1 (step S303).

[0074] The control unit 11 receives the video data and date and time information transmitted from the tablet 2 (step S401). The control unit 11 reads out the first message corresponding to the received date and time information from the mass storage unit 14 (step S402). The control unit 11 inputs the read out first message, video data, and instruction to generate a second message to the language model 12M (step S403). The language model 12M separates the input video data into image data and audio data (step S404). The language model 12M extracts features for each of the first message, image data, audio data, and instruction to generate a second message (step S405). The language model 12M combines the extracted features to synthesize a single sequence related to the first message, image data, audio data, and instruction to generate a second message (step S406). The control unit 11 acquires the second message output based on the sequence synthesized by the language model 12M (step S407). The control unit 11 stores the video data and the second message in association with the caregiver ID, the care-requiring person ID, the date and time information, and the first message in the mass storage unit 14 (step S408). The control unit 11 transmits the acquired second message to the tablet 2 (step S409).

[0075] The control unit 21 receives the second message transmitted from the server 1 (step S304). The control unit 21 displays the received second message on the display unit 25 (step S305).

[0076] According to the third embodiment, the care support system can output the second message based on the image data of the care-requiring person and the first message.

[0077] According to the third embodiment, the caregiver can check the level of understanding of the care-requiring person regarding the counseling based on the second message.

[0078] (Embodiment 4) In the fourth embodiment, a method will be described in which the care support system outputs a note regarding a selected support content item in association with a first message.

[0079] FIG. 16 is an explanatory diagram showing the record layout of the support content item DB. The support content item DB 142 includes a cautions column. The cautions column stores matters that caregivers should be aware of when asking questions about support content items or assessment items. The cautions are set corresponding to the support content items or assessment items, and there may be overlaps. In addition, the caution "explanation of the purpose, review of what will be discussed today, etc." is stored in the cautions column independently of the support content items or assessment items.

[0080] The control unit 11 performs the process of acquiring and storing the first message in the same procedure as in the third embodiment. The control unit 11 reads out notes corresponding to the read support content item or assessment item from the support content item DB 142. In FIG. 16, the control unit 11 reads out notes corresponding to the assessment item "daily water intake status, the level of understanding of the person and family regarding water and salt amounts, and daily dietary intake status," respectively, such as "understand with objective numerical values, etc.; understand the specific details of water intake; consider instructions from the family doctor; understand information necessary for evaluating dietary status; consider instructions from the doctor and weight gain / loss status." In addition to the above notes, the control unit 11 also reads out a note "review of the purpose, what to talk about today, etc." If there are duplicate notes, the control unit 11 does not need to read out the same note.

[0081] The control unit 11 separates the first message into an introduction portion and a content portion. Specifically, the control unit 11 separates the first message into an introduction portion and a content portion by using the opening portion of the first message (for example, the first and second first messages) as the introduction portion and the remaining first message as the content portion. The introduction portion of the first message is, for example, the opening portion of the first message, which is a standard message for starting counseling between the caregiver and the person requiring care. The content portion of the first message is, for example, a message related to a support content item or an assessment item. The control unit 11 may separate the first message into the introduction portion and the content portion using a known machine learning model or the like that separates standard phrases from other sentences.

[0082] The control unit 11 associates the introductory portion of the first message with the notes "review of the gist, what will be discussed today, etc." The control unit 11 calculates the similarity between the content portion of the first message and the support content items or assessment items. The similarity between the content portion of the first message and the support content items or assessment items is calculated, for example, using cosine similarity. The control unit 11 associates the content portion of the first message having the highest calculated similarity with the support content items or assessment items. Specifically, the control unit 11 identifies which of the assessment items "daily water intake status, level of understanding of the person and family regarding water and salt amounts, daily food intake status" the content portion of the first message corresponds to based on the calculated similarity. The control unit 11 associates the above-mentioned content portion of the first message with notes related to the support content items or assessment items. The control unit 11 transmits to the tablet 2 the associated introductory part of the first message, the content part of the first message, the support content items or assessment items, and the notes.

[0083] The control unit 21 receives the introductory part of the first message transmitted from the server 1, the content part of the first message, the support content items or assessment items, and the notes. The control unit 21 displays on the display unit 25 the introduction part of the first message, the content part of the first message, the support content items or assessment items, and the notes.

[0084] FIG. 17 is an explanatory diagram showing an example of the display of a first message and precautions. The screen d06 in FIG. 17 includes a speaker column d61, a first message column d62, and a precaution column d63. The speaker column d61 displays the speaker who will deliver the first message. The speaker may be, for example, a caregiver, a care recipient, or a relative of the care recipient. The first message column d62 displays the first message output by the language model 12M. The first messages in FIG. 17 are in the form of a conversation between the caregiver and the care recipient, and are displayed from top to bottom. In the example of FIG. 17, the introductory parts of the first messages are the first and second first messages, and the content parts of the first messages are the third and subsequent first messages. The precaution column d63 is located to the right of the first message column d62. The precautions d63 are not limited to being located to the right of the first message column d62, and may be located around the first message column d62. The notes field d63 displays the notes associated with the introductory part of the first message and the notes associated with the content part of the first message. The notes field d63 also displays the support content items or assessment items associated with the content part of the first message.

[0085] There are four cautions displayed in total. The first caution, "Explanation of the purpose and introduction of what will be discussed today," is displayed in conjunction with the first messages from the first to second messages. The second caution, "Consider your doctor's instructions," is displayed in conjunction with the first messages from the third to fifth messages. The assessment item related to the second caution, "Understanding of fluid and salt intake," is displayed above the second caution. The third caution, "Understand objective numerical values, etc., and understand the specific details of fluid intake," is displayed in conjunction with the first messages from the sixth to eighteenth messages. The assessment item related to the third caution, "Daily fluid intake status," is displayed above the third caution. The fourth caution, "Understand the information necessary to evaluate dietary status, and consider doctor's instructions and weight gain / loss status," is displayed in conjunction with the first messages from the nineteenth and twenty-second messages. The assessment item related to the fourth caution, "Daily food intake status," is displayed above the fourth caution.

[0086] In the fourth embodiment, an example has been described in which the precautions are stored in the support content item DB 142, but the present invention is not limited to this. The control unit 11 may acquire the precautions using the language model 12M (LLM). In this case, the control unit 11 generates a prompt to be input into the language model 12M based on the attributes of the caregiver, the attributes of the care recipient, the support content items, the assessment items, the first message, and an instruction to generate the precautions. The instruction to generate the precautions is, for example, "Generate the precautions that you should convey to the care recipient along with the first message." The control unit 11 acquires the precautions by inputting the generated prompt into the language model 12M.

[0087] FIG. 18 is a flowchart showing an example of a processing procedure of the care support system according to the fourth embodiment. In the flowchart shown in FIG. 18, step S211 and steps S103 to S104 in the processing shown in FIG. 9 are replaced with steps S601 to S607 and steps 701 to S702. Explanations of steps similar to those in FIG. 9 will be omitted. The control unit 11 reads out, from the support content item DB 142, notes corresponding to the assessment items of the read support content items and the notes "review of the gist, what will be discussed today, etc." (step S601). The control unit 11 separates the first message into an introduction section and a content section (step S602). The control unit 11 associates the introduction section of the first message with the notes "review of the gist, what will be discussed today, etc." (step S603). The control unit 11 calculates the similarity between the content section of the first message and the support content items or assessment items (step S604). The control unit 11 associates the content portion of the first message having the calculated high similarity with the support content item or assessment item (step S605). The control unit 11 associates the content portion of the first message described above with notes related to the support content item or assessment item (step S606). The control unit 11 transmits the associated introductory portion of the first message, the content portion of the first message, the support content item or assessment item, and the notes to the tablet 2 (step S607).

[0088] The control unit 21 receives the introductory part of the first message, the content part of the first message, the support content items or assessment items, and the notes (step S701) transmitted from the server 1. The control unit 21 displays the introductory part of the first message, the content part of the first message, the support content items or assessment items, and the notes (step S702).

[0089] According to the fourth embodiment, the care support system can display the notes regarding the selected support content items in association with the first message.

[0090] According to the fourth embodiment, a caregiver using the care support system can smoothly proceed with counseling by checking the precautions.

[0091] (Embodiment 5) In the fifth embodiment, a method will be described in which the care support system outputs the first message and the video data in association with each other in chronological order.

[0092] Based on an instruction from the caregiver, the control unit 21 displays a screen (not shown) requesting the reading of the first message and video data. The screen requesting the reading of the first message and video data includes fields for inputting a caregiver ID, a care-requiring person ID, and date and time information. The caregiver inputs the caregiver ID, the care-requiring person ID, and the date and time information via the input unit 24. The control unit 21 transmits to the server 1 the caregiver ID, the care-requiring person ID, the date and time information, and a notification requesting the reading of the first message and video data input by the caregiver.

[0093] The control unit 11 receives the caregiver ID, the care recipient ID, the date and time information, and a notification requesting the reading of the first message and video data, all of which are transmitted from the tablet 2. The control unit 11 reads the first message and video data corresponding to the caregiver ID, the care recipient ID, and the date and time information from the mass storage unit 14. The control unit 11 extracts audio data from the video data. The control unit 11 identifies the caregiver's audio data from the extracted audio data using the frequency and intensity of the audio data. The control unit 11 may identify the caregiver's audio data and the care recipient's audio data. The control unit 11 converts the identified caregiver's audio data into text data. The control unit 11 synchronizes the first message and the video data by comparing the first message with the text data derived from the caregiver. Specifically, the control unit 11 aligns the first message and the video data along the same time axis and identifies whether each part of the first message corresponds to each area (time period) of the video data. The method for synchronizing the first message and the video data is not limited to the above example. The control unit 11 may convert the first message into audio data, compare the waveform of the converted audio data with that of the caregiver, and synchronize the first message with the video data. The control unit 11 then transmits the synchronized first message and video data to the tablet 2.

[0094] The control unit 21 receives the first message and the video data transmitted from the server 1. The control unit 21 displays the first message and the video data on the display unit 25.

[0095] FIG. 19 is an explanatory diagram showing a display example of a first message and video data. In FIG. 19, the first message and video data are displayed in association with each other in chronological order. A screen d07 in FIG. 19 includes a speaker column d71, a first message column d72, and a video data column d73. The display format of the speaker column d71 and the first message column d72 is the same as in FIG. 17. The video data column d73 displays thumbnails of video data at predetermined time intervals (for example, 10 to 60 seconds). In FIG. 19, thumbnails of video data are displayed at 40-second intervals. In FIG. 19, the thumbnail of video data at "0:40" is displayed in association with the 11th first message from the top, "I fill it up so much that it doesn't spill (about 80%) and drink it all."

[0096] 20 is a flowchart showing an example of a processing procedure of the care support system according to embodiment 5. Based on an instruction from a caregiver, the control unit 21 displays a screen (not shown) requesting reading of the first message and video data (step S801). The control unit 21 transmits to the server 1 a notice requesting reading of the first message and video data, including the caregiver ID, the care-requiring person ID, date and time information, and the first message and video data (step S802).

[0097] The control unit 11 receives the caregiver ID, the care-requiring person ID, the date and time information, and a notification requesting the reading of the first message and video data, all of which are transmitted from the tablet 2 (step S901). The control unit 11 reads out the first message and video data corresponding to the caregiver ID, the care-requiring person ID, and the date and time information from the mass storage unit 14 (step S902). The control unit 11 extracts audio data from the video data (step S903). The control unit 11 identifies the caregiver's audio data from the audio data using the frequency, intensity, etc. of the audio data (step S904). The control unit 11 converts the identified caregiver's audio data into text data (step S905). The control unit 11 compares the first message with the converted text data to synchronize the first message and the video data (step S906). The control unit 11 transmits the synchronized first message and video data to the tablet 2 (step S907).

[0098] The control unit 21 receives the first message and the video data transmitted from the server 1 (step S803). The control unit 21 displays the first message and the video data on the display unit 25 (step S804).

[0099] According to the fifth embodiment, the care support system can output the generated first message and video data in association with each other in chronological order.

[0100] According to the fifth embodiment, the caregiver can check the condition of the care-requiring person later during the counseling session.

[0101] (Embodiment 6) In the sixth embodiment, a method is described in which the care support system outputs a third message based on the first message, the audio data included in the video data, and the second message. The third message is, for example, a message indicating a countermeasure.

[0102] The control unit 21 displays a screen (not shown) requesting the generation of a third message based on an instruction from the caregiver. The screen requesting the generation of the third message includes fields for inputting a caregiver ID, a care-requiring person ID, and date and time information. The caregiver inputs the caregiver ID, the care-requiring person ID, and the date and time information via the input unit 24. The control unit 21 transmits to the server 1 the caregiver ID, the care-requiring person ID, the date and time information input by the caregiver, and a notification requesting the generation of a third message.

[0103] The control unit 11 receives the caregiver ID, the care recipient ID, the date and time information, and a notification requesting the generation of a third message sent from the tablet 2. The control unit 11 reads out the first message, video data, and second message corresponding to the caregiver ID, the care recipient ID, and the date and time information from the mass storage unit 14. The control unit 11 extracts audio data from the video data. The control unit 11 inputs the read first message, audio data, second message, and third message generation instruction into the language model 12M. The third message generation instruction is, for example, "Based on the first message, audio data, and second message, please generate content that should be kept in mind in the next counseling session."

[0104] The language model 12M extracts features from each of the first message, the voice data, the second message, and the third message generation instruction. The language model 12M combines the extracted features to synthesize a single sequence related to the first message, the voice data, the second message, and the third message generation instruction. The language model 12M performs a calculation to generate a third message from the synthesized sequence and outputs the generated third message. For example, when the first message, the voice data, the second message, and the third message generation instruction shown in FIG. 7 are input, the language model 12M outputs a third message stating, "Thank you for your hard work. It seems that Mr. A did not fully understand the content of the middle part of the counseling session (the conversation about cerebrovascular disease). Please review the middle part at the next counseling session. Also, Mr. A will find it easier to understand the content of the counseling session if you speak more slowly. Please try speaking more slowly at the next counseling session."

[0105] The control unit 11 acquires the third message output by the language model 12M. The control unit 11 stores the third message in the mass storage unit 14 in association with the caregiver ID, the care-requiring person ID, the date and time information, the first message, the second message, and the video data. The control unit 11 transmits the acquired third message to the tablet 2.

[0106] The control unit 21 receives the third message transmitted from the server 1. The control unit 21 displays the third message on the display unit 25.

[0107] FIG. 21 is an explanatory diagram showing an example of the display of a third message. The screen d08 in FIG. 21 includes a third message field d81. The third message field d81 displays the third message output by the language model 12M. The third message field d81 in FIG. 21 displays the third message, "Thank you for your hard work. It seems that Mr. A did not fully understand the content of the middle part of the counseling session (the conversation about cerebrovascular disease). Please review the middle part at the next counseling session. Also, Mr. A will find it easier to understand the content of the counseling session if you speak more slowly. Please try speaking more slowly at the next counseling session." In this embodiment, an example in which the screen d08 includes the first message field d81 has been described, but this is not limiting. The screen d08 may also include an order field, a first message field, a note field, a video data field, etc. In this case, the third message field d81 is displayed around the order field, the first message field, the note field, the video data field, etc.

[0108] 22 is a flowchart showing an example of a processing procedure of the care support system according to embodiment 6. The control unit 21 displays a screen requesting the generation of a third message based on an instruction from the caregiver (step S1001). The control unit 21 transmits the caregiver ID, the care-requiring person ID, date and time information input by the caregiver, and a notification requesting the generation of the third message to the server 1 (step S1002).

[0109] The control unit 11 receives the caregiver ID, the care-requiring person ID, the date and time information, and the notification requesting the generation of a third message, all of which have been transmitted from the tablet 2 (step S1101). The control unit 11 reads out the first message, video data, and second message corresponding to the caregiver ID, the care recipient ID, and the date and time information from the mass storage unit 14 (step S1102). The control unit 11 extracts audio data from the video data (step S1103). The control unit 11 inputs the read out first message, audio data, second message, and third message generation instruction to the language model 12M (step S1104). The language model 12M extracts features for each of the first message, audio data, second message, and third message generation instruction (step S1105). The language model 12M combines the extracted features to synthesize a single sequence related to the first message, audio data, second message, and third message generation instruction (step S1106). The control unit 11 acquires the third message output based on the sequence synthesized by the language model 12M (step S1107). The control unit 11 stores the third message in the mass storage unit 14 in association with the caregiver ID, the care-requiring person ID, the date and time information, the first message, the second message, and the video data (step S1108). The control unit 11 transmits the acquired third message to the tablet 2 (step S1109).

[0110] The control unit 21 receives the third message transmitted from the server 1 (step S1003). The control unit 21 displays the third message on the display unit 25 (step S1004).

[0111] According to the sixth embodiment, the care support system can output the third message based on the first message, the audio data included in the video data, and the second message.

[0112] According to the sixth embodiment, by using the third message, the caregiver can utilize the lessons learned from the previous counseling session in the next counseling session.

[0113] The features described in each of the above embodiments can be combined with each other. Furthermore, the independent claims and dependent claims described in the claims can be combined with each other in any and all combinations, regardless of the reference format. Furthermore, the claims use a format in which a claim references two or more other claims (multiple claim format), but this is not limited to this. Multiple claims (multi-multi claims) that reference at least one other multiple claim may also be used.

[0114] The embodiments disclosed herein are illustrative in all respects and should not be considered limiting. The scope of the present invention is defined by the claims, not by the above meaning, and is intended to include all modifications within the meaning and scope of the claims. [Explanation of symbols]

[0115] 1. Information processing device (server) 11 Control section 12 Storage section 12P control program 12M language models 13 Communications Department 14 Mass storage 141 Caregiver Database 142 Support content item DB 143 Database of people requiring care 15 Reading unit 2. Information processing equipment (computers) 21 Control Unit 22 Memory section 22P control program 23 Communications Department 24 Input section 25 Display section 26 Imaging device 26a Camera 26b Mike 1a Portable storage media

Claims

1. Acquire attributes of a caregiver, attributes of a care recipient, and a plurality of support content items selected from a plurality of all support content items for the care recipient; generating a first message to be conveyed by the caregiver using a language model that utilizes the acquired attributes of the caregiver, the attributes of the care recipient, and the selected plurality of support content items; Output the generated first message. Information processing methods.

2. generating a prompt including the acquired attributes of the caregiver, the attributes of the care-requiring person, the selected plurality of support content items, and an instruction to generate the first message; generating the first message by providing the generated prompt to the language model; The information processing method according to claim 1 .

3. acquiring a plurality of assessment items corresponding to the selected plurality of support content items; generating a prompt including attributes of the caregiver, attributes of the care recipient, the selected plurality of support content items, the acquired plurality of assessment items, and an instruction to generate the first message; generating the first message by providing the generated prompt to the language model; The information processing method according to claim 1 .

4. After generating the first message, acquiring video data of the care-requiring person; generating a second message indicating an emotion or a level of understanding of the care-requiring person by providing the acquired video data and the generated first message to the language model; Output the generated second message.

4. The information processing method according to claim 1.

5. and outputting notes regarding the plurality of support content items in association with the first message.

4. The information processing method according to claim 1.

6. The generated first message and the video data are output in a time-series manner. The information processing method according to claim 4.

7. outputting a third message indicating a countermeasure for the first message using the language model that utilizes the first message, the audio data included in the video data, and the second message; The information processing method according to claim 4.

8. Acquire attributes of a caregiver, attributes of a care recipient, and a plurality of support content items selected from a plurality of all support content items for the care recipient; generating a first message to be conveyed by the caregiver using a language model that utilizes the acquired attributes of the caregiver, the attributes of the care recipient, and the selected plurality of support content items; Output the generated first message. A program that causes a computer to perform a process.

9. An information processing device having a control unit, The control unit Acquire attributes of a caregiver, attributes of a care recipient, and a plurality of support content items selected from a plurality of all support content items for the care recipient; generating a first message to be conveyed by the caregiver using a language model that utilizes the acquired attributes of the caregiver, the attributes of the care recipient, and the selected plurality of support content items; Output the generated first message. Information processing device.

Images