Information input support mechanism and information input support program

Abstract

To implement an information input support mechanism and an information input support program which generate an instruction command for instructing a text generative AI so that a generated instructive comment is made into a text including more natural and necessary contents.SOLUTION: An information input support mechanism is provided to cause a text generative AI to generate an instructive comment about a management item directed to a management object person and comprises: a data acquisition unit 10 including an object person information acquisition unit 12 which acquires management object person data specific to the management object person, and an instructive information acquisition unit 13 which acquires data about an instructive item representing instructive contents of the instructive comment; a command generation unit 20 which generates an instruction command for text generation to the text generative AI on the basis of the management object person data and the data about the instructive item acquired by the data acquisition unit; and an instruction unit 80 which gives the instruction command generated by the command generation unit to the text generative AI.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an information input support mechanism and information input support program used in creating guidance comments for persons to be managed, such as patients and practitioners of health management programs in the medical and health field, and in particular to an information input support mechanism and information input support program that creates instruction commands to instruct a sentence generation AI on the conditions for creating sentences, so that the guidance comments created by the sentence generation AI are the desired ones, when creating guidance comments using a sentence generation AI. [Background technology]

[0002] For those who are patients with lifestyle-related diseases such as diabetes, or those who are participating in health management programs, doctors and health management advisors who act as administrators provide regular lifestyle guidance to guide them toward the goal of treating their illness and improving their health.

[0003] In such efforts, by explaining to the target managed person what they should work on daily, and by providing written comments praising the managed person when results appear as a result of their efforts over a certain period of time, it is expected that this will deepen the managed person's understanding, encourage their motivation to continue, and lead to greater results.

[0004] However, creating documents for multiple managed individuals that take into account their different current issues and the results of their efforts places a heavy burden on the administrators, such as doctors and health management advisors.

[0005] Conventionally, in a health management program that manages the health of a managed individual, when an expert sends an advice message to the managed individual, a health management server has been proposed in which the expert selects a preferred one from a number of illustrations or sample text prepared in advance, and then creates an advice message based on the selection of the sample text or illustration, and sends the advice message together with the selected illustration to a terminal held by the managed individual (Patent Document 1).

[0006] Furthermore, a medical image report input system has been proposed as a system for creating findings that are stored together with images acquired in the medical field. In this system, the interpretation points of the images and the display order of answer candidates are saved as templates, and the doctor can then select images and answer candidates according to the template to create the final findings (Patent Document 2). [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Patent No. 6261548 specification [Patent Document 2] JP 2004-157815 A Summary of the Invention [Problem to be solved by the invention]

[0008] According to the conventional writing support technology described above, advice messages and observations are created by combining multiple sample sentences, which significantly reduces the burden of writing sentences compared to when a doctor or health care advisor creates a sentence from scratch.

[0009] However, simply combining short, standard sentences will result in a lack of coherence as a whole sentence, and will leave the text lacking in some areas and sounding unnatural compared to a sentence created by a human writer from scratch.

[0010] On the other hand, a technology that has recently evolved rapidly and become easily available is text generation AI, which creates text such as human conversational text, called chat pod. Text created by such text generation AI has a more natural overall flow rather than a patchwork of simple sentences, so it is expected that by having a text generation AI create guidance comments to be conveyed by a doctor or health management advisor, who is the administrator, to managed persons such as patients in a treatment program or practitioners who are practicing a health management program, it will be possible to create natural comments that the managed persons can accept without feeling uncomfortable. However, in order to create accurate and natural guidance comments for each situation, it is essential that the input command to the text generation AI contains necessary and sufficient information.

[0011] In order to solve the above problems, the present application aims to realize an information input support mechanism and an information input support program that create instruction commands to be given to a text generation AI so that the instruction comments created are more natural and contain the necessary content. [Means for solving the problem]

[0012] The information input support mechanism disclosed in the present application is an information input support mechanism for causing a sentence generation AI to create instruction comments for management items aimed at a managed person, and is characterized in that it comprises a data acquisition unit having a subject information acquisition unit that acquires managed person data specific to the managed person and an instruction information acquisition unit that acquires data related to instruction items that represent the instruction content in the instruction comments, a command creation unit that creates an instruction command for the sentence generation AI to create a sentence based on the managed person data and the data related to the instruction items acquired by the data acquisition unit, and an instruction unit that instructs the sentence generation AI of the instruction command created by the command creation unit.

[0013] In addition, the information input support program disclosed in the present application is an information input support program for causing a sentence generation AI to create instruction comments for management items aimed at the managed person, and is characterized in that it creates an instruction command for the sentence generation AI to create the instruction comment based on managed person data, which is information specific to the managed person, and data related to the instruction items that represent the instruction content in the instruction comment, acquired by a data acquisition unit. Effect of the Invention

[0014] According to the present disclosure, instructional comments containing necessary information can be easily created in a natural style using text generation AI. [Brief description of the drawings]

[0015] [Figure 1] FIG. 2 is a block diagram showing a configuration of an information input support mechanism according to the present embodiment. [Diagram 2] 4 is a flowchart showing a process flow of an information input support program according to the present embodiment. [Diagram 3] 13 is a diagram showing an example of an information input screen that supports input of data acquired by a subject information acquisition unit and a text information acquisition unit. FIG. [Figure 4] 13 is a diagram showing an example of an information input screen that supports the input of data acquired by a training information acquisition unit. FIG. [Diagram 5] FIG. 13 is a diagram showing an example of a medical treatment instruction card created by the Diabetes Association. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0016] The information input support mechanism disclosed in the present application is an information input support mechanism for causing a sentence generation AI to create instruction comments for management items aimed at a managed person, and includes a data acquisition unit having a subject information acquisition unit that acquires managed person data specific to the managed person and an instruction information acquisition unit that acquires data related to instruction items that represent the instruction content in the instruction comments, a command creation unit that creates an instruction command for the sentence generation AI to create a sentence based on the managed person data and the data related to the instruction items acquired by the data acquisition unit, and an instruction unit that instructs the sentence generation AI of the instruction command created by the command creation unit.

[0017] In this way, the information input support mechanism disclosed in this application can take advantage of the characteristics of a sentence generation AI that can create sentences in a natural style, making it easy to create easy-to-understand instruction comments that contain the necessary information.

[0018] It is preferable that the information input support mechanism disclosed in the present application further comprises a storage unit storing data on a plurality of previously prepared training items, and the training information acquisition unit acquires the data on the training items by having the user select one or more of the training items from the plurality of previously prepared training items. In this way, the training content to be included in the training comment can be linked with the training items, thereby deepening the managed person's understanding of the training items.

[0019] In addition, it is preferable that the managed person data has a plurality of items, the storage unit stores a plurality of options prepared in advance for a predetermined item among the plurality of items, and the target person information acquisition unit acquires the managed person data to be acquired for the predetermined item by having the target person select from the plurality of options. In this way, it is possible to simplify input of the managed person data and reduce the burden on the administrator.

[0020] Furthermore, it is preferable that the subject information acquisition unit acquires the managed subject data from a subject information grasping means that measures or manages the managed subject data specific to the managed subject. In this way, it is possible to save the administrator the trouble of inputting the managed subject data acquired by the subject information grasping means that is an external device.

[0021] In addition, it is preferable that the data acquisition unit further includes a text information acquisition unit that acquires text information data, which is data on the format of the instruction comment created by the text generation AI, and the command creation unit creates the instruction command based on the text information data. In this way, the instruction comment created by the text generation AI can be in a more desirable format.

[0022] Furthermore, it is preferable that the command creation unit creates the instruction command by giving priority to data related to the training item acquired by the training information acquisition unit among the data acquired by the data acquisition unit. In this way, a training comment having content closely linked to the management item can be created.

[0023] The person to be managed may be a diabetic patient, and the management item may be diabetes.

[0024] The information input support program disclosed in the present application is an information input support program for causing a sentence generation AI to create instruction comments for management items aimed at a managed person, and is characterized in that it creates an instruction command for causing the sentence generation AI to create the instruction comment based on managed person data, which is information specific to the managed person, and data related to the instruction items that represent the instruction content in the instruction comment, acquired by a data acquisition unit.

[0025] In this way, the information input support program disclosed in this application makes it possible to easily create instructional comments written in a natural style that contains the necessary information, taking advantage of the characteristics of a text generation AI that can create text in a natural style.

[0026] Hereinafter, an embodiment of an information input support mechanism and an information input support program disclosed in the present application will be described with reference to the drawings.

[0027] In the following embodiment, an example is given of a case where the person to be managed is a diabetic patient, and when the doctor in charge, who is the administrator, provides the diabetic patient with lifestyle guidance for the treatment of diabetes, which is a management item, the doctor in charge uses a text generation AI to create guidance comments to be presented along with the guidance item cards shown at appropriate times.

[0028] (Embodiment) FIG. 1 is a block diagram illustrating the overall configuration of an information input support mechanism disclosed in the present application.

[0029] The block diagram in Figure 1 shows the components of the information input support mechanism divided into blocks based on the functions they fulfill, and does not show physical components as blocks. For this reason, one block may be composed of multiple electronic components, and multiple circuits shown as blocks may be mounted on one circuit board.

[0030] As shown in FIG. 1, the information input support mechanism 100 of this embodiment includes a data acquisition unit 10, a command creation unit 20, a control unit 30, a memory unit 40, a display unit 50, an input unit 60, an output unit 70, and a transmission / reception unit 80 which is an instruction unit.

[0031] The data acquisition unit 10 of the information input support mechanism 100 described in this embodiment is a part that acquires various information that will be the contents of the instruction command that causes the text generation AI to create a desired instruction comment. The data acquisition unit 10 of the information input support mechanism 100 shown in this embodiment has a text information acquisition unit 11 that acquires text information data that is format data that indicates the expressions for the instruction comment that the text generation AI is to create and the formal content of the text, a subject information acquisition unit 12 that acquires managed subject data that is information specific to the managed subject such as a diabetic patient to whom the instruction comment is presented, and an instruction information acquisition unit 13 that acquires data related to the instruction items that indicate the instruction contents in the instruction comment.

[0032] The text information acquisition unit 11 acquires text information data, which is data on the formal aspects of the instruction comments to be created by the text generation AI, such as the type of language, the total number of characters, the tone of the sentence, and whether or not a greeting or self-introduction is required.

[0033] The subject information acquisition unit 12 acquires managed subject data, which is various information that is useful for making the content of the guidance comment more specific to the managed subject, such as the name of the managed subject to whom the guidance comment is presented, essential information for identifying the person to whom the guidance comment is being made, such as an identification number such as an ID, various biometric information of the managed subject, and information indicating the current status of the managed subject in the management program.

[0034] The instruction information acquiring unit 13 acquires data on instruction items, which are information on instruction contents classified into instruction items such as instruction sentence examples included in the instruction comments, and specific instruction contents associated with the instruction items.

[0035] In addition, the information input support mechanism 100 of this embodiment provides support by presenting options for the items to be input and acquiring necessary information from external devices and applications so that the doctor in charge, who is the administrator who creates the guidance comments, can easily and reliably input the text information data, managed person data, and data related to guidance items acquired by the data acquisition unit 10.

[0036] More specifically, a method can be adopted in which first the diabetic patient to be managed is identified by his / her name and / or identification ID, and then an information input screen is displayed on the display screen of display unit 50 (described later) showing options for items classifying the characteristics of the managed person to be obtained as managed person data, the managed person's biometric information, and the language, number of characters, style, etc. of the instructional comments to be obtained as text information data, and the administrator, the doctor in charge, selects one of the options.

[0037] Furthermore, the data acquisition unit 10 can link the data input unit 60 described below with this information input screen to read information identifying the managed person, and receive managed person data such as the managed person's biometric information and current condition as data from a target person information acquisition means (not shown) such as an external measuring device that has acquired the managed person's biometric information and information indicating the current condition, or a mobile terminal that runs a management application, thereby eliminating the need for the responsible doctor, who is the administrator, to enter data using the information input screen.

[0038] Furthermore, by linking and storing the managed person and the managed person data of the managed person and the contents of previously created instruction commands in the memory unit 40 described below, when the managed person is identified, the data stored in the memory unit 40 can be called up and displayed on the display unit 50, or can be displayed as a selected item by default on the information input screen.

[0039] The command creation unit 20 creates an instruction command that instructs the sentence generation AI on what kind of instruction comment to create, based on the managed person data, data related to the training items, and sentence information data acquired by the data acquisition unit 10 described above.

[0040] The instruction command created by the command creation unit 20 is in a form that matches the command input form of the sentence generation AI that creates the instruction comment, and creates and outputs the necessary and sufficient content for creating the desired instruction comment.

[0041] The command creation unit 20 creates a command that gives importance to the data related to the training items acquired by the training information acquisition unit 13 among the managed person data, data related to training items, and text information data acquired by the data acquisition unit 10, and instructs that the entire data related to the training items be reflected in the training comments. For this reason, the instruction command created by the command creation unit 20 is output as a command that instructs the text generation AI to appropriately reflect the designated content in the created text.

[0042] The control unit 30 controls the overall operation of the information input support mechanism 100 .

[0043] The control unit 30 can be configured with a known microcomputer or electronic logic circuit. When the information input support mechanism 100 according to the present embodiment is realized as software installed in a personal computer or as an application for a smartphone or tablet terminal, the central processing unit (CPU) of the personal computer, tablet terminal, or smartphone operates as the control unit 30.

[0044] The memory unit 40 stores various data such as the managed person data, data related to training items, and text information data acquired by the data acquisition unit 10, as well as a portion of the managed person data and a portion of the text information data linked to the managed person used for the above-mentioned input support information presented by the data acquisition unit 10 to reduce data input by the doctor in charge who is the administrator, instruction commands created by the command creation unit 20, and programs for operating the information input support mechanism 100.

[0045] In addition, by storing in the memory unit 40 the instruction comments created by the sentence generation AI together with an evaluation of the instruction comments and the instruction commands used when creating the comments, it is expected that the instruction commands can be analyzed based on the obtained evaluation of the instruction comments, which can be useful in creating better instruction commands.

[0046] The memory unit 40 can be constructed from various known memory circuits, and when the information input support mechanism 100 is realized as software for a personal computer or an application for a smartphone or tablet terminal, similar to the control unit 30 described above, the memory unit of each electronic device operates as the memory unit 40.

[0047] The display unit 50 is composed of a display device (display terminal) equipped with a display screen such as a liquid crystal panel or an EL panel, and a control unit for displaying a desired image on the display device, and displays various information (data) acquired by the data acquisition unit 10, an information input screen for the above-mentioned input support when the administrator inputs necessary information into the data acquisition unit 10, and the like. In addition, the display unit 50 can display the created instruction commands and the instruction comments created by the text generation AI, so that the administrator, who is the user, can check the contents of each.

[0048] When the information input support mechanism 100 is realized as software for a personal computer or an application for a smartphone or tablet terminal, the display screen and image display driver of each electronic device operate as the display unit 50.

[0049] The input unit 60 is a part where data held by a subject information grasping means, which is an external measuring device such as a weighing scale or a body composition scale, or a mobile terminal such as a smartphone or a tablet terminal running software or applications that operate other health management programs, is input to the information input support mechanism 100 described in this embodiment.

[0050] Therefore, the input unit 60 can be realized as a data input unit connectable to a subject information grasping means, which is an external device, and as a data reading unit that reads necessary information from a storage medium such as a USB memory. In addition, when the input unit 60 functions as a barcode reader that reads one-dimensional or two-dimensional barcodes, it can read information that identifies the subject to be managed from a barcode provided on the patient's appointment card or chart of the diabetic patient, who is the subject to be managed. Furthermore, when data can be transmitted and received between the subject information grasping means, which is an external device, and the input unit 60 via a short-range communication means such as Bluetooth (registered trademark), the input unit 60 functions as a part that acquires necessary information from communication data received by the transmission / reception unit 80, which will be described later.

[0051] The output unit 70 can output various data to the outside, such as instruction commands created by the information input support mechanism 100 in this embodiment, instruction comments created by a sentence generation AI, and a list of information on managed persons.

[0052] If the output unit 70 has a printer function, it can print directly. In addition, when printing on a printer of another device or displaying the data on a display monitor of another device, or when saving the data on an external device, the output unit 70 can have an output function of converting the data into an appropriate data format and storing it on various storage media such as a USB memory.

[0053] The transmitting / receiving unit 80 has a transmitting function for transmitting data from the information input support mechanism 100 to an external device, and a receiving function for receiving data transmitted from an external device.

[0054] In the most general form, in the information input support mechanism 100 shown in this embodiment, the instruction command created by the command creation unit 20 is transmitted by the transmission function of the transmission / reception unit 80, and the sentence generation AI included in the external server 300 receives the instruction command via the Internet environment 200 shown in FIG. 1 and creates an instruction comment according to the content of the instruction command. Therefore, in the information input support mechanism 100 according to this embodiment, the transmission / reception unit 80 corresponds to an instruction unit that instructs the sentence generation AI to create an instruction comment. Note that in the future, when the sentence generation AI is installed as software in an information device used by an administrator, such as a personal computer, and becomes available for use, an information transmission part that transmits the instruction command created by the information input support mechanism 100 to the sentence generation AI present in the same information device and instructs it to create an instruction comment corresponds to the instruction unit.

[0055] Moreover, the instruction comments created by the sentence generation AI are transmitted from the server 300 and received by the transmission / reception unit 80 of the input support mechanism 100 via the Internet environment 200, and the doctor in charge, who is the administrator, can check the contents on the display terminal of the display unit 50. Moreover, as described above, the instruction comments created by the sentence generation AI can be stored in the storage unit 40 of the input support mechanism 100.

[0056] When the above-mentioned Internet environment 200 is used, the transmission / reception unit 80 utilizes various carrier communications, and when the data acquisition unit 10 receives information about the subject from other management software or applications, or measurement results from various bio-information measuring devices such as a weight scale or a blood pressure monitor, it is desirable for the transmission / reception unit 80 to be compatible with multiple communication means so as to be able to utilize various short-range communication means used by the subject information grasping means, which is an external device such as Bluetooth (registered trademark).

[0057] Next, a specific description will be given of a procedure in which a doctor in charge, who is an administrator, creates guidance comments for a diabetic patient, who is a person to be managed, using the information input support program executed by the information input support mechanism according to this embodiment.

[0058] FIG. 2 is a flowchart for explaining the operation of the information input support program explained in this embodiment.

[0059] 2, the information input support program described in this embodiment acquires information about the managed person and the contents of the instruction comment (managed person data, data about instruction items, text information data) by the data acquisition unit 10 of the information input support mechanism 100 (step S10), and the command creation unit 20 creates an instruction command based on the information acquired by the data acquisition unit 10 (managed person data, data about instruction items, text information data) (step S102), and instructs the server 300 equipped with a text generation AI to create a comment text by sending the instruction command by the transmission / reception unit 80, which is an instruction unit (step S103). The text generation AI that receives the instruction command creates a teaching comment according to the instruction command (step S104).

[0060] FIG. 3 is an example of a first information input screen displayed on the display screen of the display unit 50 to assist the attending physician in inputting data acquired by the subject information acquisition unit and the text information acquisition unit when creating instructional comments for a diabetic patient using the information input support mechanism shown in this embodiment.

[0061] As shown in Figure 3, the first information input screen shows multiple options for each input item, and the doctor in charge can reduce the effort required to input the required information by selecting the appropriate option on the screen by clicking, or touching it if using a touch panel.

[0062] The first information input screen shown in Fig. 3 displays a "Destination Name" field for inputting the patient name, patient ID, etc., in order to identify the diabetic patient to whom the guidance comment is to be presented first. The doctor in charge inputs the patient name and patient ID into the boxes on the screen. As described above, when information identifying the diabetic patient is input from the input unit 60, which is a barcode reader, the input patient information is displayed.

[0063] As shown in FIG. 3, in the information input support mechanism 100 described in this embodiment, various data such as a "patient classification" which classifies the diabetes, a "blood glucose level" which classifies the blood glucose level based on the measurement data, a "BMI" which classifies and identifies the degree of obesity of the patient, and a "weight change" which indicates the change in weight required to understand the current state of the patient are input by the attending physician by selecting the appropriate data from options displayed on the information input screen as unique information of the diabetic patient that is necessary to be reflected in the instruction comments.

[0064] As described above, when a diabetic patient as a managed person uses an external measuring device as a means for grasping information about the person, and the patient's biological information such as weight and blood glucose level is input from the external means for grasping information about the person via the input unit 60, or when the memory unit 40 stores the patient's past input data, the corresponding option is selected based on the managed person data obtained from the input unit 60 or the memory unit 40 and displayed as the default, allowing the doctor in charge to confirm the input contents and eliminating the need for the doctor to select an option again.

[0065] In addition, the first information input screen shown in Figure 3 displays text information data, which is superficial data about the textual aspects of the instruction comments created by the text generation AI, as a choice of options, to assist the attending physician in selecting and inputting the text information data in the same manner as the managed person data described above.

[0066] To assist with the input of text information data, the system displays items such as a "Language" item for selecting the language in which to write the instruction comment, or for directly inputting the language, as shown in the example of Figure 3, a "Character count" item for specifying the number of characters in the instruction comment, and a "Texture" item for specifying the text tone of the instruction comment, allowing the attending physician to select the appropriate option from each item.

[0067] In addition, the first information input screen shown in Figure 3 has a "creation command" item for instructing the text generation AI on the type of person expected to create the instruction comment, allowing the doctor in charge to select or input the type of person who will create the instruction comment.

[0068] Furthermore, with regard to inputting the above-mentioned text information data and data on the image of the person for whom the instructional comments are to be made, if the memory unit 40 has stored past data, that content can be displayed as the default on the information input screen, thereby reducing the input burden on the attending physician.

[0069] FIG. 4 is a diagram illustrating a second information input screen displayed when data related to training items is acquired in the information input support mechanism according to this embodiment.

[0070] As shown in Figure 4, the second information input screen, which supports the input of guidance items that the doctor in charge will give to the diabetic patient in the guidance comment, displays a list of guidance items defined by the Diabetes Association, and the doctor in charge selects from the list the guidance item that corresponds to the guidance comment to be created this time. In order to prevent the doctor in charge from selecting the wrong guidance item, for example, in the case of guidance item "1-1 Blood Glucose", only the summary of the content to be explained in the guidance item, such as "Blood glucose level is the concentration of sugar (glucose) in the blood," and the specific explanation of guidance item "1-1 Blood Glucose," such as "How food becomes bloodline" and "Blood glucose level of a healthy person," are displayed, making the content easier to understand compared to when only the guidance item is displayed.

[0071] The memory unit 40 stores, as text, the specific explanation content to be given to the diabetic patient for each instruction item selected by the attending physician on the second information input screen shown in Figure 4. When the data acquisition unit 10 grasps the selected instruction item, the specific explanation content for that instruction item is transmitted as data from the memory unit 40. The data acquisition unit 10 can then transmit this specific instruction content to the command creation unit 20, so that it can be reflected in the instruction command for the text generation AI.

[0072] Figure 5 shows a medical treatment instruction card containing instruction items created by the Diabetes Association.

[0073] FIG. 5 shows an example of the medical treatment instruction card for "1-1 Blood Sugar" explained in FIG. 4, with the front side of the card shown at the top and the back side of the card shown at the bottom.

[0074] The guidance comments for diabetes treatment created with the assistance of the information input support mechanism 100 using the information input support program described in this embodiment are handed over to the diabetic patient, who is the person to be managed, along with the treatment guidance card shown in Figure 5. The diabetic patient can refer to the guidance comments to deepen his / her understanding of the guidance items and put the contents into practice, thereby progressing with the diabetes treatment guidance.

[0075] In addition, on the back of the medical treatment guidance card shown in the lower part of Figure 5, there is a check box for recording the date on which guidance was received, allowing both the attending physician and the diabetic patient to check the guidance history.

[0076] Based on the conditions for creating an instruction comment specified by the attending physician as a result of the input support provided by the data acquisition unit 10 as described above, the command creation unit 20 creates an instruction command to instruct the sentence generation AI that creates the instruction comment.

[0077] For example, if the text generation AI is Chat GPT, this instruction command will be in a format that includes a "creation command" that indicates the premise for creating the document and the goal to be achieved by the created text, a string of one or more commands beginning with "♯♯" such as "♯♯ Language: Japanese" indicating the conditions for creating the text, a string of restrictions such as "♯♯ Restriction: Number of characters in the created text must be 300 or less" indicating prohibitions on text creation, and a required item field beginning with "♯♯ context:" that describes the guidance content for the selected guidance item.

[0078] Here, in the case of an instruction command to Chat GPT, the instruction content for the guidance item is entered as text in the required information field beginning with "♯♯context:". In this way, the specific explanatory content for the guidance item will be reflected in the guidance comment created by Chat GPT, which is a text-generating AI, and this will be the most important content in the created guidance comment.

[0079] The instruction command created by the command creation unit 20 is transmitted from the transmission / reception unit 80 as an instruction unit via the Internet environment 200 to the sentence generation AI in the server 300, and the instruction comment is created by the sentence generation AI.

[0080] As mentioned above, the instruction comments created by the text generation AI are sent back to the information input support mechanism 100 via the Internet environment 200 and displayed on the display unit 50. The doctor in charge can then check whether there are any serious defects in the content and send it as data to the diabetic patient's smartphone or other device, or print it out and communicate it to the diabetic patient on paper.

[0081] As described above, in the information input support mechanism 100 according to the present embodiment, when the doctor in charge, who is the administrator, inputs information identifying the diabetic patient who is the person to be managed when creating a diabetes care instruction comment, the information is displayed as a selectable option on the information input screen in order to support the input of information that the text information acquisition unit 11, the subject information acquisition unit 12, and the instruction information acquisition unit 13 need to acquire. Therefore, the doctor in charge can input information required for an instruction command to make the text generation AI create an appropriate instruction comment by selecting the appropriate option from the displayed options. Note that it is not essential that all input items are displayed as options on the information input screen, and as shown as an example of the selection of "language" in FIG. 3, when there are many options, the doctor in charge can directly input some of them.

[0082] Furthermore, in the information input support mechanism 100 according to this embodiment, the instruction command created by the command creation unit 20 is transmitted by the transmission / reception unit 80 to the server 300 having the text generation AI via the Internet environment 200, and the guidance comment created by the text generation AI is received by the transmission / reception unit 80 and its content can be displayed on the display unit 50 to the doctor in charge, who is the administrator. Therefore, the doctor in charge can communicate the guidance comment to the diabetic patient, who is the person to be managed, after confirming that it is appropriate.

[0083] In the above embodiment, an example was shown in which the data acquisition unit 10 includes the text information acquisition unit 11 that acquires text information data, which is format data about the text of the instruction comment to be created. However, since a general text generation AI can automatically determine and create text based on the content of the input instruction command, it is not essential that the data acquisition unit 10 is equipped with the text information acquisition unit 11 that acquires text information data such as language, number of characters, and tone.

[0084] In the above embodiment, the diabetic patient is the person to be managed, the doctor in charge is the administrator who manages the patient and creates the instruction comment, and diabetes is the management item. However, the information input support mechanism and the information input support program disclosed in the present application are not limited to diabetic patients. In addition to diabetes, patients with various diseases, such as hypertension and cardiovascular diseases, for which improvement of the patient's lifestyle is important and effective in treatment, can be managed as the person to be managed, and information input can be supported in creating instruction comments that the doctor in charge, who is the administrator, shows to these patients. In this case, the management items are various diseases suffered by the person to be managed for whom the instruction comment is to be created.

[0085] In addition, the information input support mechanism and the information input support program disclosed in this application can be used not only to create guidance comments that a doctor in charge presents to a patient, but also to support input when a text generation AI is used to create guidance comments that a health management advisor, as a manager, presents at appropriate times to practitioners of lifestyle improvement programs for health promotion and health management, as persons to be managed. In this case, the management item is health management.

[0086] In addition, when creating guidance comments for managed persons other than diabetic patients, instead of the guidance items created by the Diabetes Association shown in the above embodiment, guidance items set by each medical organization and their associated guidance texts, and guidance items that stipulate basic guidance content that is appropriately shown in the health management program will be used. [Industrial Applicability]

[0087] The information input support mechanism and information input support program disclosed in this application are extremely useful in creating instruction commands that include necessary and sufficient conditions when a doctor in charge or a health management advisor, who is an administrator, has a text-generation AI create instruction comments to show to patients and health management program practitioners, who are the subjects of management. [Explanation of symbols]

[0088] 10 Data Acquisition Section 11 Text information acquisition part 12. Target Information Acquisition Department 13 Guidance Information Acquisition Department 20 Command Creation Section 80 Transmitting / receiving unit (instruction unit)

Claims

1. An information input support mechanism for allowing a sentence generation type AI to create guidance comments for management items for a managed person, a data acquisition unit including a subject information acquisition unit that acquires managed subject data unique to the managed subject, and a training information acquisition unit that acquires data related to training items that indicate training contents in the training comments; A command creation unit that creates an instruction command for the sentence generation AI to create a sentence based on the managed person data and the data related to the training item acquired by the data acquisition unit; An information input assistance mechanism comprising: an instruction unit that instructs the sentence generation AI of the instruction command created by the command creation unit.

2. Further comprising a storage unit in which data relating to a plurality of previously prepared teaching items is stored, 2. The information input support mechanism according to claim 1, wherein the training information acquisition unit acquires the data relating to the training items by having the user select one or more of the training items from the data relating to the plurality of training items prepared in advance.

3. The management target data includes a plurality of items, the storage unit stores a plurality of options prepared in advance for a predetermined item among the plurality of items, 3. The information input support mechanism according to claim 2, wherein the target person information acquisition unit acquires the managed person data to be acquired for the predetermined item by having the managed person data selected from the plurality of options.

4. 2. The information input support mechanism according to claim 1, wherein the target person information acquisition unit acquires the managed person data from a target person information grasping means that measures or manages managed person data specific to the managed person.

5. The data acquisition unit further includes a text information acquisition unit that acquires text information data, which is data on the format of the instruction comment created by the text generation AI, The information input support mechanism according to claim 1 , wherein the command creation section creates the instruction command based on the text information data.

6. 2 . The information input support mechanism according to claim 1 , wherein the command creation section creates the instruction command by giving priority to data relating to the training item acquired by the training information acquisition section among the data acquired by the data acquisition section.

7. 2. The input support mechanism according to claim 1, wherein the person to be managed is a diabetic, and the management item is diabetes.

8. An information input support program for causing a sentence generation type AI to create guidance comments for management items for a managed person, An information input support program, characterized in that it creates an instruction command for causing a sentence generation AI to create the instruction comment based on managed person data, which is information specific to the managed person, and data related to training items that represent the training content in the training comment, acquired by a data acquisition unit.

Images