Information processing equipment, methods, programs and systems

Abstract

To streamline the process of creating documents. [Solution] The program causes the processor to perform the following steps: receiving a designation from a first user of a child or a group to which a child belongs as the target for generating draft text; identifying one or more images from among the images stored in the database that record child welfare activities as images related to the designated child or group as the target for generation; generating a prompt that includes instructions to read the situation of the identified images and to generate draft text about the designated child or group based on the read situation; inputting the prompt containing information about the images into a large-scale language model to obtain draft text from the large-scale language model; and presenting the obtained draft text to the first user.

Description

Technical Field

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

Background Art

[0002] Staff such as childcare workers and preschool educators in childcare facilities not only provide support for children's lives and education but also engage in document creation tasks such as communicating with parents, recording work content, and promoting activities. Since document creation tasks require time and effort, there is a demand for efficiency improvement.

[0003] Patent Document 1 describes an image utilization system 10. The image utilization system 10 in Patent Document 1 acquires images including children 80 taken by cameras in a childcare worker terminal 30 or a parent terminal 20, and performs predetermined processing including adding data, deleting data, and / or modifying data on the acquired images to generate processed images. Using the data incorporating the processed images, service data for use in various services is generated, thereby creating processed images suitable for each communication tool and automatically selecting and applying processed images suitable for the purpose of each communication tool to create a communication tool. As a result, images can be appropriately applied to communication tools created by a nursery school 110 or the like to improve work efficiency.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] The image utilization system described in Patent Document 1 was insufficiently efficient in creating text to be included in documents.

[0006] The purpose of this disclosure is to streamline the process of creating documents. [Means for solving the problem]

[0007] The program causes the processor to perform the following steps: receive a designation from a first user of a child or a group to which a child belongs as the target for generating draft text; identify one or more images from the database of images recording child welfare activities as images related to the designated child or group; generate a prompt that includes instructions to read the context of the identified images and to generate draft text about the designated child or group based on the context read; input the prompt containing information about the images into a large-scale language model to obtain draft text from the large-scale language model; and present the obtained draft text to the first user. [Effects of the Invention]

[0008] According to this disclosure, the process of creating documents can be streamlined. [Brief explanation of the drawing]

[0009] [Figure 1] This is a block diagram showing the functional configuration of System 1. [Figure 2] This is a block diagram showing the functional configuration of the terminal device 10. [Figure 3] This is a block diagram showing the functional configuration of Server 20. [Figure 4] This figure shows an example of the configuration of prompt P. [Figure 5] This diagram shows the data structure of User Table 2021. [Figure 6] This diagram shows the data structure of the children's table 2022. [Figure 7] It is a diagram showing the data structure of the image table 2023. [Figure 8] It is a diagram showing the data structure of the document type table 2024. [Figure 9] It is a diagram showing the data structure of the output format table 2025. [Figure 10] It is a diagram showing the data structure of the text table 2026. [Figure 11] It is a diagram showing the data structure of the log management table 2027. [Figure 12] It is a flowchart of the document generation process of this embodiment. [Figure 13] It is a schematic diagram showing an example of the input screen D1. [Figure 14] It is a schematic diagram showing an example of the text draft display screen D2. [Figure 15] It is a flowchart of the evaluation process of this embodiment. [Figure 16] It is a schematic diagram showing an example of the condition setting screen D3. [Figure 17] It is a schematic diagram showing an example of the result display screen D4. [Figure 18] It is a block diagram showing the basic hardware configuration of the computer 90.

Embodiments for Carrying Out the Invention

[0010] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In all the drawings for describing the embodiments, common components are denoted by the same reference numerals, and repeated descriptions are omitted. Note that the following embodiments do not unduly limit the content of the present disclosure described in the claims. Also, not all the components shown in the embodiments are essential components of the present disclosure. Further, each figure is a schematic diagram and is not necessarily drawn precisely.

[0011] Also, in the following description, a "processor" is one or more processors. At least one processor is typically a microprocessor such as a CPU (Central Processing Unit), but may also be another type of processor such as a GPU (Graphics Processing Unit). At least one processor may be single-core or multi-core.

[0012] Also, at least one processor may be a processor in a broad sense, such as a hardware circuit (e.g., FPGA (Field-Programmable Gate Array) or ASIC (Application Specific Integrated Circuit)) that performs part or all of the processing.

[0013] Also, in the following description, expressions such as "xxx table" may be used to describe information from which an output is obtained for an input. This information may be data of any structure or a learning model such as a neural network that generates an output for an input. Therefore, "xxx table" can be referred to as "xxx information".

[0014] Also, in the following description, the configuration of each table is an example. One table may be divided into two or more tables, or all or part of two or more tables may be one table.

[0015] Also, in the following description, when the processing is described with "program" as the subject, since the program is executed by a processor to perform the defined processing while appropriately using a storage unit and / or an interface unit, etc., the subject of the processing may be the processor (or a device such as a controller having that processor).

[0016] The program may be installed on a device such as a computer, or it may reside on a program distribution server or a computer-readable (e.g., non-temporary) recording medium. Furthermore, in the following description, two or more programs may be implemented as a single program, or one program may be implemented as two or more programs.

[0017] Furthermore, in the following explanation, identification numbers are used as identification information for various objects, but other types of identification information (for example, identifiers that include letters or symbols) may also be used.

[0018] Furthermore, in the following explanations, when describing similar elements without distinction, a reference code (or a common code among reference codes) may be used, and when describing similar elements with distinction, the element's identification number (or reference code) may be used.

[0019] Furthermore, in the following explanation, only control lines and information lines deemed necessary for the explanation are shown, and not all control lines and information lines in the product are necessarily shown. All components may be interconnected.

[0020] Each information processing device consists of a computer equipped with an arithmetic unit and a memory device. The basic hardware configuration of the computer and the basic functional configuration of the computer realized by said hardware configuration will be described later. For each of the terminal device 10 and the server 20, explanations that overlap with the basic hardware configuration and basic functional configuration of the computer described later will be omitted.

[0021] <Definition> In this disclosure, "documents" refers to documents related to activities concerning child welfare. These documents include, for example, documents used by persons working at child welfare facilities (such as childcare workers) for record-keeping and reporting (diaries, reports, work records, plans, etc.), and documents used for reporting to persons belonging to child welfare facilities (children, etc.) or their guardians (communication notebooks, letters, etc.).

[0022] In this disclosure, "child" refers to a child who belongs to (attends) a child welfare facility. "Group to which a child belongs" refers to a group that includes a child within a child welfare facility, and includes, for example, the child welfare facility, grade level, class, group, etc. In the following explanation, "child or group to which a child belongs" will be simply referred to as "children, etc."

[0023] <Structure> The configuration of the information processing system according to the first embodiment will now be described. Figure 1 is a block diagram illustrating the configuration of the information processing system according to the first embodiment.

[0024] As shown in Figure 1, the information processing system 1 includes a terminal device 10, a server 20, and a generating AI 40.

[0025] In Figure 1, the number of terminal devices 10 included in System 1 is not limited to one. There may be two or more terminal devices 10 included in System 1.

[0026] Figure 1 shows an example where System 1 includes one Server 20, but the number of Server 20 included in System 1 is not limited to one. Each Server 20 may consist of multiple servers depending on the functions they have. Also, each Server 20 may, for example, be a collection of multiple devices. The way in which the multiple functions required to implement the Server 20 according to this embodiment are distributed to one or more hardware can be appropriately determined in consideration of the processing capacity of each hardware and / or the specifications required for the Server 20.

[0027] The terminal device 10 shown in Figure 1 can be implemented, for example, by a stationary PC (Personal Computer) or a laptop PC. Alternatively, the terminal device 10 may be implemented by a mobile device such as a smartphone or tablet that supports a mobile communication system. Furthermore, the terminal device 10 may be implemented by a wearable device such as an HMD (Head Mount Display).

[0028] The terminal device 10 comprises a communication interface 12, an input device 13, an output device 14, memory 15, storage 16, and a processor 19. The input device 13 is a device for receiving input operations from the user (e.g., a touch panel, touchpad, etc.). The output device 14 is a device for presenting information to the user (e.g., a display, speaker, etc.).

[0029] The server 20 is implemented, for example, by an information processing device connected to the network 80. As shown in Figure 1, the server 20 includes a communication IF 22, an I / O IF 23, memory 25, storage 26, and a processor 29. The I / O IF 23 functions as an interface for an input device to receive input operations from the user and an output device to present information to the user.

[0030] The Generative AI 40 is an artificial intelligence model executed by one or more information processing devices connected to the network 80. The Generative AI 40 generates output based on user input (prompt text, images, etc.) using a pre-trained model. Examples of Generative AI 40 include GPT, Claude, LLaMA, BERT, and Gemini. The Generative AI 40 may also be a model capable of processing multiple types of data simultaneously, such as text, images, audio, and video. In this disclosure, a large-scale language model that primarily outputs text information as output information is described as a type of Generative AI 40. The Generative AI 40 may also be incorporated into the server 20.

[0031] Each information processing device consists of a computer equipped with an arithmetic unit and a memory device. The basic hardware configuration of the computer and the basic functional configuration of the computer realized by said hardware configuration will be described later. For each of the terminal device 10, server 20, and generation AI 40, explanations that overlap with the basic hardware configuration and basic functional configuration of the computer described later will be omitted.

[0032] <Configuration of terminal device 10> Figure 2 is a block diagram showing an example configuration of the terminal device 10 shown in Figure 1. As shown in Figure 2, the terminal device 10 comprises a communication unit 120, an input device 13, an output device 14, an audio processing unit 17, a microphone 171, a speaker 172, a storage unit 180, and a control unit 190. Each block included in the terminal device 10 is electrically connected, for example, by a bus.

[0033] The communication unit 120 performs processing such as modulation and demodulation processing for the terminal device 10 to communicate with other devices. The communication unit 120 performs transmission processing on the signal generated by the control unit 190 and transmits it to an external source (for example, the server 20). The communication unit 120 performs reception processing on the signal received from an external source and outputs it to the control unit 190.

[0034] The input device 13 is a device for a user operating the terminal device 10 to input instructions or information. The input device 13 can be implemented by a reader, keyboard 131, mouse 132, etc. The input device 13 may also be implemented by a touch-sensitive device that inputs instructions by touching an operating surface. The input device 13 converts the instructions input by the user into electrical signals and outputs the electrical signals to the control unit 190. The input device 13 may also include, for example, a receiving port that accepts electrical signals input from an external input device.

[0035] The output device 14 is a device for presenting information to the user operating the terminal device 10. The output device 14 is implemented, for example, by a display 141. The display 141 displays data according to the control of the control unit 190. The display 141 is implemented, for example, by an LCD (Liquid Crystal Display) or an organic EL (Electro-Luminescence) display.

[0036] The audio processing unit 17 performs, for example, digital-to-analog conversion processing of the audio signal. The audio processing unit 17 converts the signal received from the microphone 171 into a digital signal and provides the converted signal to the control unit 190. The audio processing unit 17 also provides the audio signal to the speaker 172. The audio processing unit 17 is implemented, for example, by an audio processing processor. The microphone 171 receives an audio input and provides the audio signal corresponding to that audio input to the audio processing unit 17. The speaker 172 converts the audio signal received from the audio processing unit 17 into audio and outputs the audio to the outside of the terminal device 10.

[0037] The storage unit 180 is implemented by, for example, memory 15 and storage 16, and stores data and programs used by the terminal device 10. The storage unit 180 stores, for example, user information 181.

[0038] User information 181 stores information about the user performing the operation. User information includes, for example, user ID, name, age, address, date of birth, and date of registration for the service.

[0039] The control unit 190 is realized when the processor 19 reads a program stored in the memory unit 180 and executes instructions contained in the program. The control unit 190 controls the operation of the terminal device 10. By operating according to the program, the control unit 190 performs the functions of an operation reception unit 191, a transmission / reception unit 192, and a presentation control unit 193.

[0040] The operation reception unit 191 performs processing to receive instructions or information input from the input device 13. For example, the operation reception unit 191 receives instructions or information input from the input device 13, etc.

[0041] Furthermore, the operation reception unit 191 receives voice information input from the microphone 171. Specifically, for example, the operation reception unit 191 receives voice data input from the microphone 171 and converted into digital data by the voice processing unit 17.

[0042] The transmitting / receiving unit 192 performs processing to enable the terminal device 10 to send and receive data with an external device such as the server 20 in accordance with a communication protocol. Specifically, for example, the transmitting / receiving unit 192 sends instructions input by the user to the server 20. The transmitting / receiving unit 192 receives information provided by the server 20.

[0043] The presentation control unit 193 controls the output device 14 and other devices in order to present information such as information provided by the server 20 to the user.

[0044] <Functional configuration of Server 20> Figure 3 shows an example of the functional configuration of server 20. As shown in Figure 3, server 20 functions as a communication unit 201, a storage unit 202, and a control unit 203.

[0045] The communications unit 201 performs processing to enable the server 20 to communicate with external devices.

[0046] The storage unit 202 includes, for example, a user table 2021, a child table 2022, an image table 2023, a document type table 2024, an output format table 2025, a document table 2026, a log management table 2027, and so on. The memory unit 202 may also store time-slot information that associates the start and end times of a specific time period with the activities that take place during that time period (for example, 8:30-9:00: arrival at the nursery, 9:00-10:00: morning meeting, 10:00-11:00: class activities, 11:00-11:30: lunch preparation, 11:30-13:00: lunch, 13:00-14:30: afternoon nap, 14:30-15:00: snack time, 15:00-16:00: outdoor play, 16:00-: departure from the nursery).

[0047] User Table 2021 is a table that stores information about users registered with the service according to this embodiment. In this embodiment, a user is, for example, a person who performs the task of creating documents related to the draft documents generated in this embodiment, and one example is a childcare worker.

[0048] The Children's Table 2022 is a table that stores information about children.

[0049] Image table 2023 is a table that stores image information related to children.

[0050] The Document Type Table 2024 is a table that manages the types of documents created in child welfare facilities.

[0051] Output Format Table 2025 is a table that stores information about the format or output examples of text included in documents created at child welfare facilities.

[0052] Document table 2026 is a table that stores information about documents written in records of child welfare activities. Specifically, document table 2026 stores information used to generate draft documents for records of child welfare activities, information about those draft documents, and information about documents created based on those draft documents.

[0053] Log management table 2027 is a table that stores log information, which is information that associates a child's status with the time (or date without time) when that status was observed or recorded. The log information includes, for example, the following: Information regarding attendance at child welfare facilities (attending daycare, daycare, etc.) Information regarding leaving a child welfare facility (dismissal, withdrawal, etc.) • Information regarding health status (temperature, excretion, etc.) • Information about diet (e.g., food and fluid intake) • Information regarding sleep (falling asleep, waking up, sleep state, etc.)

[0054] The memory unit 202 may have different log management tables 2027 for each type of child's condition, such as sleep status or meal status. For example, the memory unit 202 may store information regarding attendance at and return from child welfare facilities, information regarding sleep, information regarding the child's health status (temperature, excretion, etc.), and information regarding the child's meals in different log management tables 2027. Furthermore, one or more time periods may be associated with tables that manage log information related to those time periods. For example, a log management table 2027A may be defined to manage log information related to meals, associated with the time period "11:30-13:00: Lunch," and a log management table 2027B may be defined to manage log information related to sleep, associated with the time period "13:00-14:30: Nap," etc.

[0055] The control unit 203 is realized when the processor 29 reads a program stored in the memory unit 202 and executes instructions contained in the program. By operating according to the program, the control unit 203 performs functions such as a receive control module 2031, a transmit control module 2032, a prompt generation module 2033, an evaluation module 2034, a log management module 2035, a presentation module 2036, etc.

[0056] The receive control module 2031 controls the process by which the server 20 receives signals from external devices according to a communication protocol. Specifically, for example, the receive control module 2031 receives signals transmitted from the terminal device 10 and the large-scale language model.

[0057] The transmission control module 2032 controls the process by which the server 20 transmits signals to external devices according to a communication protocol. For example, the transmission control module 2032 transmits information about various UIs (described later) to the terminal device 10, which will be displayed on the terminal device 10's display 141. For example, the transmission control module 2032 transmits a request including a prompt to the large-scale language model.

[0058] The prompt generation module 2033 generates prompts for input to a large-scale language model. Prompts are input data used to instruct the large-scale language model to generate output data or to guide the content of the output data. Prompts can contain information in various formats, including text data and image data (including still images and videos). Image data included in a prompt may be in any image format such as JPEG, or it may be encoded in text format, etc. Prompts may also include links such as URLs or file paths to the location where the image data is stored, thereby allowing the large-scale language model to load the image data.

[0059] Figure 4 shows an example of the structure of prompt P generated by prompt generation module 2033. Prompt P is text data and includes text data for each part exemplified by role instruction P1, document-specific instruction P2, child information P3, time zone specification P4, image information P5, output format P6, and log information P7. Note that the order in which role instruction P1 to log information P7 are listed is not particularly limited. Prompt P does not have to contain all of role instruction P1 to log information P7, and may contain strings other than role instruction P1 to log information P7.

[0060] (Role Instructions P1) Role instruction P1 includes sentences that specify the role (position) of the large-scale language model when generating draft text. Role instruction P1 includes sentences that specify the position of the large-scale language model in generating text, such as "You are an excellent childcare worker." The prompt generation module 2033 generates prompt P by combining the following document-specific instructions P2 to log information P7 with role instruction P1.

[0061] (Document specific instructions P2) Document-specific instructions P2 include text that specifies the type of document for which a draft document should be generated. Furthermore, document-specific instructions P2 includes instructions regarding the content to be included in the draft document, which vary depending on the document type. For example, if the document type is "communication notebook," document-specific instructions P2 might include text such as, "Please create a draft document to be included in the communication notebook used for communication with parents. Please output the child's activities, expressions, demeanor, and any special notes estimated from the images, according to the format below." The prompt generation module 2033 generates prompts containing different instructions depending on the type of document specified by the user. For example, based on the type of document specified by the user, the prompt generation module 2033 searches the document type table 2024 (described later), retrieves the text data of the "Document-Specific Instructions" field in the corresponding record, and joins it with the role instruction P1 as document-specific instruction P2.

[0062] (Children's Information P3) Child Information P3 includes text that specifies the child(ren) to be included in the draft text generation. Specifically, Child Information P3 includes information that can identify the child(ren) (such as the child's ID, name, etc., or the ID, name, etc., of the group to which the child belongs). The prompt generation module 2033, for example, searches the child table 2022 (described later) based on the information of the child specified by the user, identifies the identification information of the child specified by the user (child ID, facility ID, class ID, group ID, etc.), and combines the text containing the identified identification information with the role instruction P1 as child information P3.

[0063] (Time slot selection P4) Time zone specification P4 includes instructions regarding the specification of time zones. Specifically, time zone specification P4 includes instructions regarding the inclusion of a description of the specified time zone in the draft document. Furthermore, time zone specification P4 may also include instructions regarding the specification of multiple time zones and the generation of draft documents separated by each specified time zone.

[0064] An example of a time zone specification P4 is shown below. Note that the fields enclosed in curly braces are dynamically entered by the prompt generation module 2033. (Example of time-slot specification P4) The following is a breakdown of the facility's daily schedule by time of day. The draft document should describe the children's behavior during {Time Slot 1} and {Time Slot 2}, and output it separately for each time slot. Time zone 1 corresponds to image 1. Time zone 2 corresponds to image 2. 8:30-9:00: Arrival at daycare • 9:00-10:00: Morning meeting 10:00-11:00: Class activities 11:00-11:30: Lunch preparation (toilet, handwashing, gargling) • 11:30-13:00: Lunch ·13:00-14:30: Nap 14:30-15:00: Snack time 15:00-16:00: Outdoor play 16:00- : Dismissal

[0065] When a user specifies one or more images and one or more time periods, the prompt generation module 2033 refers to the one or more time periods and the time the one or more images were taken to determine which of the one or more images corresponds to the time period specified by the user. For example, it determines which images have a time taken within the time period specified by the user as corresponding to the time period specified by the user.

[0066] It is not mandatory for the user to specify a time period. In that case, the prompt generation module 2033 may refer to the time period information stored in the memory unit 202 to identify the time period that includes the time the user specified the image was taken, and associate the image with the time period.

[0067] Then, the prompt generation module 2033 reads the image conditions corresponding to each time period and generates text containing instructions for the large-scale language model to generate draft sentences containing sentences related to each time period, separated by time period, and combines this text with the role instruction P1 as a time period specification P4.

[0068] (Image information P5) Image information P5 includes text that specifies the image to be read by the large-scale language model. For example, image information P5 may include image data converted to any format, or information that can identify the image (image identifier, image URL, etc.). The prompt generation module 2033, for example, searches the image table 2023 (described later) based on the information of the image specified by the user, identifies the storage location information of the image specified by the user, and combines the text containing the storage location information as image information P5 with the role instruction P1.

[0069] (Output format P6) Output format P6 includes information about the output format (template) for the draft document and examples of draft document generation. In this embodiment, the contents of output format P6 correspond to the document type. The prompt generation module 2033 generates prompts containing information about the output format of the draft document, which varies depending on the type of document specified by the user. For example, based on the type of document specified by the user, the prompt generation module 2033 searches the output format table 2025 (described later), retrieves the text data of the "Output Format" field in the corresponding record, and combines it with role instruction P1 as output format P6.

[0070] (Log information P7) Log information P7 contains log information that is referenced by the large-scale language model when generating draft text. The prompt generation module 2033, for example, retrieves log information for a child specified by the user on a date specified by the user from the log management table 2027, and combines it with the role instruction P1 as log information P7, along with an instruction to generate a draft text by referring to the log information.

[0071] The prompt generation module 2033 may extract log information recorded for the child specified by the user from the log management table 2027 during the time period associated with the image, and include it in log information P7. Then, when generating a draft text for the image specified by the user, it may include an instruction in log information P7 to include the content of the extracted log information.

[0072] Alternatively, the prompt generation module 2033 may extract log information for the child specified by the user from a log management table 2027 of the type corresponding to the time period associated with the image (for example, a log management table 2027B that manages log information related to sleep, corresponding to the time period "13:00-14:30: afternoon nap"), and include it in log information P7. Then, when generating a draft text for the image specified by the user, an instruction to include the content of the extracted log information may be included in log information P7.

[0073] The following is an example of log information P7. In this example, "13:00-14:30: Nap" is specified as time zone 1, and an image of a sleeping child specified by the user is specified as image 1. (Example from log information P7) Since {Time Zone 1} corresponds to {Image 1}, the text for {Image 1} should also include the log information recorded during {Time Zone 1}. #{Time Zone 1} Log Information 13:05: Fall asleep 13:10: Lying on my back 13:15: Awakening 13:30: Fall asleep 13:35: Sideways 14:00: Awakening

[0074] A large-scale language model that receives a prompt with inserted log information can refer to the log information to include details that could not be gleaned from the image in the draft text. Therefore, it becomes possible to output a draft text that provides a more detailed description of the children in child welfare facilities. For example, in the above example, since Image 1 is an image of a child sleeping, the large-scale language model cannot infer from Image 1 that the child woke up in the middle of a nap. On the other hand, by referring to the log information, the large-scale language model can output a draft text about the child waking up in the middle of a nap.

[0075] As described above, the prompt generation module 2033 generates prompt P by combining role instructions P1 to log information P7.

[0076] The evaluation module 2034 evaluates children by analyzing text data (draft texts) generated by a large-scale language model regarding the children being evaluated. Specifically, evaluation module 2034 evaluates designated children by performing text mining on text data using any text mining method. For example, evaluation module 2034 vectorizes text data using any method. Then, evaluation module 2034 extracts characteristic words related to any indicator (specifically, indicators related to children's qualities and growth established in child welfare activities, such as the so-called "five domains of childcare" represented by "health," "relationships," "environment," "language," and "expression"). Evaluation module 2034 evaluates the degree of each child's characteristics in each indicator by counting the frequency of occurrence of the extracted words or evaluating the importance of those words in the text data using methods such as TF-IDF. Evaluation module 2034 may also use any sentiment analysis to evaluate the overall tone of the text data.

[0077] Furthermore, the evaluation module 2034 may present the evaluation results to the user using images such as radar charts or graphs, showing the degree of evaluation for each indicator. Alternatively, the evaluation module 2034 may present the degree of evaluation for each indicator to the user in text. For example, it may generate text based on the importance of words extracted from the text data of the draft text and present it to the user. For example, if the importance score of words associated with "human relations" is high, it may be determined that the child being evaluated has strengths in human relations, and text of the evaluation result comments may be generated.

[0078] Furthermore, the evaluation module 2034 may also input text data or data processed into any format, such as vectorization, into a large-scale language model, and output the children's evaluations for any indicator as numerical values ​​or text indicating the degree of evaluation for each indicator, in the same manner as described above.

[0079] The log management module 2035 stores log information in the storage unit 202. Specifically, the log management module 2035 processes the information regarding the child's status, received by the receiving control module 2031 from an external device, into an arbitrary format and stores it in the storage unit 202 (particularly the log management table 2027). For example, the log management module 2035 stores log information in the storage unit 202 as follows. • The system receives information regarding the child's condition from a terminal device operated by the user that records the child's condition, processes it in an arbitrary format, and stores it in the storage unit 202 as log information. • An information processing device (any device such as a camera or sensor) that monitors the child's condition takes images at predetermined times (at times arbitrarily chosen by the user, or at periodic intervals). The log management module 2035 processes the information regarding the child's condition received by the server 20 from the information processing device into an arbitrary format and stores it as log information in the storage unit 202.

[0080] The presentation module 2036 generates a control signal to cause the output device 14 of the terminal device 10 to output predetermined information and transmits it to the terminal device 10. In particular, the presentation module 2036 causes the output device 14 of the terminal device 10 to output the text drafts obtained by the server 20 from the large-scale language model.

[0081] <Data structure> The data structure used in this embodiment will be described below.

[0082] Figure 5 shows the data structure of User Table 2021. User Table 2021 is a table with User ID as the primary key and columns such as User Name and Assigned Class.

[0083] The "User ID" field stores the user ID that identifies the user.

[0084] The "Username" field stores the user's name. The username can be any string, such as the user's full name or nickname.

[0085] The "Assigned Class" field stores information about the tasks the user is responsible for in a child welfare facility. Specifically, the "Assigned Class" field stores identification information for the group the user is responsible for in a child welfare facility.

[0086] Figure 6 shows the data structure of the child table 2022. The child table 2022 is a table with child ID as the primary key and columns such as name, facility ID, class ID, group ID, etc.

[0087] The item "Child ID" is an item that stores the child ID used to identify a child.

[0088] The "Name" field is used to store the child's name. The "Name" field can also be set to any string of characters, such as the child's full name.

[0089] The "Facility ID" field stores identification information for the child welfare facility to which the child belongs. The memory unit 202 may also store information about the facility (such as its name) associated with each Facility ID.

[0090] The item "Class ID" is an item that stores identification information of the class to which a child belongs at the child welfare facility to which the child belongs. In addition, the memory unit 202 may also store information related to the class (class name, assigned caregiver, etc.) associated with each Class ID.

[0091] The item "Group ID" is an item that stores the identification information of the group to which a child belongs within the class of the child welfare facility to which the child belongs. The memory unit 202 may also store information about the group (such as the group name) associated with each Group ID.

[0092] Note that child welfare facilities, classes, and groups are just examples of groups to which children belong. In addition to child welfare facilities, classes, and groups, the child table 2022 may also store identification information for any group to which a child belongs.

[0093] Figure 7 shows the data structure of image table 2023. Image table 2023 is a table with image ID as the primary key and columns such as storage location and shooting time.

[0094] The "Image ID" field stores the image ID that identifies the image.

[0095] The item "Storage Location" stores storage location information for images related to an image ID. The storage location information may be, for example, address information in the storage unit 202, or a URL (Uniform Resource Locator) for referencing the image.

[0096] The "Shooting Time" field stores information about the time the image associated with the image ID was taken. For example, the "Shooting Time" field stores the timestamp information of when the image was taken.

[0097] Figure 8 shows the data structure of the document type table 2024. The document type table 2024 is a table that has columns such as document type, document-specific instructions, etc., with document type ID as the primary key.

[0098] The "Document Type ID" field stores the document type ID that identifies the type of document.

[0099] The "Document Type" field stores the name of the document type. For example, the "Document Type" field stores strings such as "Childcare Log," "Communication Notebook," and "Newsletter."

[0100] The "Document-Specific Instructions" field stores information about the parts of the prompts that differ depending on the document type, which are input into the large-scale language model. Specifically, the "Document-Specific Instructions" item stores text data related to instructions for generating draft documents, which vary depending on the type of document. For example, the item "Document-Specific Instructions" stores text data related to instructions for the type of document to be used for generating draft documents, as well as instructions for the structure and style of the draft documents corresponding to that document type.

[0101] Figure 9 shows the data structure of the output format table 2025. The output format table 2025 is a table with output format ID as the primary key and columns such as document type ID and output format.

[0102] The "Output Format ID" field stores the output format ID that identifies the output format.

[0103] The "Document Type ID" field stores the Document Type ID for the document type corresponding to the Output Format ID.

[0104] The "Output Format" field stores information about the format (template) used when generating draft documents. Specifically, the "Output Format" item stores information about rules and writing styles used to organize draft documents into a consistent format. For example, the "Output Format" field stores text data where dynamic parts are indicated by blank fields. An example of this text data is shown below. Fields enclosed in curly braces are dynamically entered by the prompt generation module 2033. Fields enclosed in square brackets are entered by the large-scale language model. (Output format) Today's {children / students} will [describe today's activities]. The format was [Describe the details of the initiative] and [Describe the situation regarding the initiative]. Also, [Special Notes].

[0105] The "Output Format" field may store either text data of specific example sentences, or text data indicating the content to be included in the draft document using bullet points or other methods.

[0106] Figure 10 shows the data structure of document table 2026. Document table 2026 is a table with document ID as the primary key and columns such as user ID, child ID, document type ID, creation time, image ID, prompt, draft document, and finalized document.

[0107] The "Document ID" field stores the document ID that identifies the document.

[0108] The "User ID" field stores the user ID that identifies the user.

[0109] The "Child ID" field stores the child ID that identifies the child. Note that document table 2026 may also store group identification information (facility ID, class ID, group ID, etc.) in addition to the child ID. This allows the system to store information about documents generated for a given group, in addition to documents generated for a given child.

[0110] The "Document Type ID" field stores the document type ID that identifies the type of document.

[0111] The "Creation Time" field stores the time when the document associated with the document ID was created. Specifically, for example, the "Creation Time" field stores the time when the content of the document was finalized.

[0112] The "Image ID" field stores the image ID that identifies the image.

[0113] The item "Draft Text" stores the text data of the draft text generated by the large-scale language model in the draft text generation process described later, for the text associated with the text ID.

[0114] The item "Prompt" stores the text data of the prompt used when generating the draft text stored in the item "Draft Text" in the draft text generation process described later.

[0115] The "Confirmed Text" field stores the text data of the document whose content has been finalized by the user, for the document associated with the document ID. Specifically, for example, the "Confirmed Text" field stores the text data of the document whose content has been finalized after the user has reviewed and made appropriate revisions to the text data in the "Draft Text" field.

[0116] Figure 11 shows the data structure of the log management table 2027. The log management table 2027 is a table with log ID as the primary key and columns such as child ID, time information, log information, etc.

[0117] The "Log ID" field stores the log ID that identifies the log information.

[0118] The "Child ID" field stores the child ID that identifies the child. Note that log management table 2027 may also store group identification information (facility ID, class ID, group ID, etc.) in addition to the child ID. This allows the system to store log information related to a given group, in addition to log information related to a given child.

[0119] The "Time Information" field stores information about the time the log was recorded. Specifically, for example, the "Time Information" field stores the timestamp information of when the log was recorded.

[0120] The "Log Information" field stores log information related to the child's condition, recorded at the time specified in the "Time Information" field. Specifically, the "Log Information" item stores identification information and strings that indicate the child's status. For example, if the item "Log Information" stores information about the child's sleep state as log information, it would store strings that can identify the child's sleep state, such as "falling asleep" or "lying on their back."

[0121] <Operation> (Text draft generation process) The operation of System 1 in this embodiment will now be described. Figure 12 is a flowchart of the document draft generation process in this embodiment.

[0122] First, the terminal device 10 runs an application for generating a draft document. Note that document creation may also be performed via a web browser. For example, the user operates the terminal device 10, selects an application for generating a draft document, and has the terminal device 10 run it. Once the terminal device 10 runs the application, the control unit 190 displays a login screen for user authentication on the display 141.

[0123] On the login screen, the user enters, for example, a login ID and password. Once user authentication is complete, the control unit 190 accesses the server 20, associating it with the user ID.

[0124] When the server 20 receives access from the terminal device 10, it generates an input UI for inputting information necessary to generate a draft document, based on information about children stored in the child table 2022, information about document types stored in the document type table 2024, etc., and presents it to the user via the terminal device 10. For example, the server 20 displays the information input screen on the display 141 of the terminal device 10.

[0125] In step S11, the terminal device 10 displays an input screen D1 for generating a draft document on the display 141. Figure 13 is a schematic diagram showing an example of the input screen D1. The input screen D1 includes operation objects D11, D12, D13, D14, D15, and D16.

[0126] Operation object D11 accepts an operation to specify the date for which the draft document will be generated. For example, when an operation is received by operation object D11, terminal device 10 displays a list of dates that can be selected. Then, terminal device 10 accepts the date specification from the user.

[0127] Operation object D12 accepts an operation to specify the document type. For example, when operation object D12 is received, terminal device 10 displays a list of document types received from server 20 in a selectable format. Then, terminal device 10 accepts the specification of the document type from the user.

[0128] Operation object D13 accepts an operation to specify the child for whom draft text will be generated. For example, when operation object D13 receives an operation, terminal device 10 displays a list of children received from server 20 in a selectable format. Terminal device 10 then accepts the child selection from the user. Operation object D13 may also be an object that allows children to be specified by free input into an input form.

[0129] Furthermore, in addition to being able to specify children, operation object D13 may also be able to specify groups (facilities, classes, teams, etc.). For example, when operation object D13 receives an operation from the user, it may display a list of groups received from server 20 in a selectable format and accept the specification of the group to be used for generating draft documents.

[0130] Operation object D14 accepts an operation to specify the time period for which draft text should be generated. For example, when an operation is received on operation object D14, terminal device 10 refers to time zone information and displays a list of time zones on display 141 for the user to select.

[0131] The terminal device 10 accepts from the user the specification of one or more time periods to be used for generating draft documents. However, the user is not required to specify a time period.

[0132] Operation object D15 accepts an operation to specify an image to be used for generating a draft document. For example, when operation object D15 accepts an operation, terminal device 10 accesses a database where images are stored, located in the storage unit 202 of server 20, and displays a list of images recording activities related to child welfare on display 141. Then, terminal device 10 accepts from the user the specification of one or more images to be used for generating a draft document. Note that the specified image does not have to show the child specified in operation object D13 (for example, it may be an image without people, such as a swimming pool or a tree).

[0133] Operation object D16 accepts input for instructions to generate a draft document. For example, when terminal device 10 receives an operation to operation object D16, terminal device 10 sends an instruction to generate a draft document to server 20, along with the information based on the operations entered to operation objects D11 to D15 on input screen D1. The information that terminal device 10 sends to server 20 includes, for example, the following information. • Date information related to the date specified by the user • Identification information for the document type specified by the user (e.g., document type ID) • Identification information of children, etc., specified by the user (e.g., child ID, facility ID, class ID, group ID) • Information that can identify the time period specified by the user (for example, a string that can identify the start and end points of the time period) • Identification information of the image specified by the user (e.g., image ID)

[0134] The following explanation describes an example where a designated child is specified as a child, etc. If a designated group other than a child is specified, you may appropriately replace "child" with "group" in the following explanation.

[0135] In step S12, the server 20 receives a date specification from the user. Specifically, for example, 20 receives date information sent from the terminal device 10.

[0136] Furthermore, server 20 accepts the specification of the document type. Specifically, for example, server 20 accepts the document type ID sent from terminal device 10.

[0137] Furthermore, the server 20 accepts requests from the user to specify children or other individuals as targets for generating draft texts. Specifically, for example, the server 20 accepts child IDs transmitted from the terminal device 10.

[0138] Furthermore, the server 20 identifies one or more images from the database that record child welfare activities as images related to the child(ren) designated as the target for generation. Specifically, for example, the server 20 identifies an image specified by the user as an image related to the child(ren) designated as the target for generation by searching the image table 2023 based on the image ID transmitted from the terminal device 10. The server 20 then obtains information on the items "storage location" and "time taken" of the identified image.

[0139] Furthermore, the server 20 accepts one or more time slots from the user. Specifically, for example, the server 20 accepts time-specific information transmitted from the terminal device 10.

[0140] Furthermore, the server 20 determines the image corresponding to the time period specified by the user, based on the time period specified by the user and the time information associated with the identified image. For example, server 20 determines that images whose shooting time falls within the time period specified by the user are corresponding to the time period specified by the user. Furthermore, the user may operate the terminal device 10 to select images from among multiple images specified by the user that correspond to a time period specified by the user.

[0141] In step S13, information is obtained regarding instructions for the large-scale language model, which vary depending on the document type. Specifically, for example, server 20 searches the document type table 2024 based on the document type ID received in step S12 and retrieves the text data stored in the "Document-Specific Instructions" field of the corresponding record.

[0142] Furthermore, server 20 obtains information regarding the output format of the draft document. Specifically, for example, server 20 searches the output format table 2025 based on the document type ID received in step S12 and retrieves the text data stored in the "Output Format" field of the corresponding record.

[0143] Furthermore, server 20 retrieves information about the child specified by the user. Specifically, for example, server 20 searches child table 2022 based on the child ID received in step S12 and retrieves the information stored in the "Name" field.

[0144] Additionally, server 20 retrieves log information about the child specified by the user. Specifically, for example, when server 20 receives date information and child ID, it searches log management table 2027 based on the date information and child ID, and extracts the corresponding records. In this case, if there are multiple log management tables 2027 depending on the type of log information, records may be extracted from all types of log management tables 2027, or records may be extracted from any type of log management table 2027 based on the type of log information specified by the user.

[0145] In step S14, the server 20 generates a prompt that includes reading the image status identified in step S12 and generating a draft text about the child(ren) designated as the target for generation based on the read status. Specifically, for example, the server 20 generates a prompt to be input to the large-scale language model based on the information acquired in steps S12 and S13. Furthermore, the server 20 may include instructions in the prompt to select and present an appropriate attachment image for the generated draft text from the image identified in step S12.

[0146] In step S15, the server 20 inputs a prompt to the large-scale language model containing information about the image determined in step S12. Specifically, for example, the server 20 sends a prompt to the large-scale language model's API containing the image data determined in step S12 or information for the large-scale language model to access the image. Based on the received prompt, the large-scale language model generates a response containing a draft sentence and sends the response to the server 20.

[0147] In step S16, the server 20 obtains a response to the prompt from the large-scale language model and retrieves the draft text included in the response. The server 20 also retrieves information about the attached images to the draft text included in the response (e.g., storage location information).

[0148] In step S17, the server 20 generates a display UI for displaying the draft text extracted in step S16 and presents it to the user via the terminal device 10. Specifically, for example, the server 20 displays the draft text display screen on the display 141 of the terminal device 10.

[0149] In step S18, the terminal device 10 presents the draft document to the user by displaying a draft document display screen D2 on the display 141.

[0150] Figure 14 is a schematic diagram showing an example of the draft document display screen D2. The draft document display screen D2 includes display object D21, display object D22, display object D23, operation object D24, and operation object D25.

[0151] Display object D21 displays information about the draft document and documents being generated. For example, display object D21 displays the name of the child, the date, etc., that the server 20 received as a specification in step S12. Display object D21 may also display the child's sleep, eating, and excretion status and the time when that status was recorded, based on the log information acquired by the server 20 in step S13.

[0152] Display object D22 is an object that displays the draft text output by the large-scale language model in response to prompt input.

[0153] Display object D23 displays the attached image for the draft text, which was output by the large-scale language model in response to the prompt input.

[0154] Operation object D24 accepts operations requesting modifications to the draft text displayed on display object D22. For example, when operation object D24 accepts an operation, terminal device 10 makes the text of the draft text displayed on display object D22 editable. The user edits the text of the draft text displayed on display object D22 by operating input device 13. Note that editing of the draft text displayed on display object D22 may be possible without operations on operation object D24.

[0155] Operation object D25 accepts operations to finalize the content of a document. For example, when an operation is received by operation object D25, the draft document displayed on display object D22 at the time the operation was received is finalized as the official final document, and the text data of the finalized document is sent to server 20.

[0156] In step S19, the server 20 receives the text data of the finalized document sent from the terminal device 10. The server 20 stores the user's user ID, the child's ID specified by the user in step S12, the document type ID of the document type specified by the user in step S12, the image ID of the image identified in step S12, the text data of the prompt generated in step S14, the text data of the draft document obtained in step S16, the text data of the finalized document, and the record creation time in the fields of the new record in the document table 2026: document ID, user ID, child ID, document type ID, image ID, prompt, draft document, finalized document, and creation time, respectively.

[0157] Server 20 creates document data of the document type specified in step S12, including the text data of the finalized document, the information displayed on display object D21 of the document draft display screen D2, and the attached image displayed on display object D23, and stores it in the storage unit 202. Server 20 may also send the created document data to a predetermined destination (for example, a contact person associated with a child identified by the child ID).

[0158] (Student evaluation process) Figure 15 is a flowchart of the evaluation process in this embodiment.

[0159] First, the terminal device 10 runs an application for evaluating the children. The evaluation of the children may also be performed via a web browser. For example, the user operates the terminal device 10, selects an application for evaluating the children, and has the terminal device 10 run it. Once the terminal device 10 runs the application, the control unit 190 displays a login screen for user authentication on the display 141.

[0160] On the login screen, the user enters, for example, a login ID and password. Once user authentication is complete, the control unit 190 accesses the server 20, associating it with the user ID.

[0161] When the server 20 receives access from the terminal device 10, it generates an input UI for inputting information necessary to evaluate the children, and presents it to the user via the terminal device 10. For example, the server 20 displays a screen for setting conditions for evaluating the children on the display 141 of the terminal device 10.

[0162] In step S21, the terminal device 10 displays a condition setting screen D3, which is a setting screen for evaluating children, on the display 141. Figure 16 is a schematic diagram showing an example of the condition setting screen D3. It includes operation objects D31, D32, D33, and D34.

[0163] Operation object D31 accepts operations to specify the children to be evaluated. For example, when operation object D31 receives an operation, terminal device 10 displays a list of children received from server 20 in a selectable format. Terminal device 10 then accepts the specification of the children from the user. Operation object D31 may also be an object that allows specifying children through free input in an input form.

[0164] In this embodiment, we will describe the case where a specific child is designated as the child. If a group other than a child is designated, you may appropriately replace "child" with "group" in the following description.

[0165] Operation object D32 accepts an operation to specify the type of document to be analyzed. For example, when operation object D32 is received, terminal device 10 displays a list of document types received from server 20 in a selectable format. Then, terminal device 10 accepts the user's specification of one or more document types.

[0166] Operation object D33 accepts an operation to specify the period to be searched for in the document. For example, when operation object D33 accepts an operation, terminal device 10 displays a list of dates, allowing the user to select the start and end dates of the period. Then, terminal device 10 accepts the specification of the period by accepting the start and end dates of the period.

[0167] Operation object D34 accepts input for instructions to evaluate the child. For example, when terminal device 10 receives an operation to operation object D34, it sends instructions to server 20 to evaluate the child, along with information based on the operations entered to operation objects D31 to D33 on the condition setting screen D3. The information that terminal device 10 sends to server 20 includes, for example, the following information. • Identification information of children, etc., specified by the user (e.g., child ID, facility ID, class ID, group ID) • Identification information for the document type specified by the user (e.g., document type ID) • Period information related to the period specified by the user (for example, information on the start date and end date)

[0168] In step S22, the server 20 receives the designation of children, etc., from the user to be evaluated. Specifically, for example, the server 20 receives the child ID transmitted from the terminal device 10.

[0169] Furthermore, server 20 accepts the specification of document type and period. Specifically, for example, server 20 accepts the document type ID and period information sent from terminal device 10.

[0170] Furthermore, the server 20 extracts draft documents associated with the children designated as subjects for evaluation from the draft documents stored in the database. Specifically, for example, the server 20 searches the "Child ID," "Document Type ID," and "Creation Time" fields in the document table 2026 based on the child ID, document type ID, and period information received from the terminal device 10, and retrieves the information stored in the "Draft Document" field of the corresponding record. This information is the information stored by the server 20 in step S19 of the draft document generation process. This makes it possible to obtain draft documents for the specified document type that were generated by the large-scale language model for the specified child during the period from the start to the end of the specified period.

[0171] In step S23, the server 20 evaluates the children and others designated as subjects for evaluation by analyzing the text data of the extracted draft documents.

[0172] In step S24, the server 20 generates a results display UI for displaying the analysis results and presents it to the user via the terminal device 10. Specifically, for example, the server 20 displays the analysis results on the display 141 of the terminal device 10.

[0173] In step S25, the terminal device 10 presents the analysis results to the user by displaying the results display screen D4, which displays the analysis results, on the display 141.

[0174] Figure 17 is a schematic diagram showing an example of the results display screen D4. The results display screen D4 includes the display object D41.

[0175] Display object D41 displays the evaluation results output by server 20 in step S23. Specifically, for example, display object D41 displays the degree of evaluation for each indicator quantified in step S23 in a comparable manner. As an example, display object D41 displays the degree of evaluation for each indicator in a radar chart, as shown in Figure 17. Display object D41 also displays text indicating comments regarding the child's evaluation results.

[0176] <Variation> Each of the above steps, the document draft generation process and the child evaluation process, can be executed on either the terminal device 10 or the server 20. The above explanation shows an example where each step in each process is executed in a specific order, but the execution order of each step is not limited to the example described, unless there are dependencies.

[0177] In the above embodiment, one output format was set for each document type in the output format table 2025. However, multiple output formats may be set for each document type. In this case, the output format may be specified by the user in steps S11 and S12, and the output format P6 may be generated based on that specification.

[0178] In the above embodiment, in step S12, the server 20 accepted the designation of both an image and a child. However, the server 20 may accept only the designation of either an image or a child. If only an image is designated, the server 20 may use any facial recognition technology to identify a child from the child's face in the image and consider the identified child as the one designated by the user. If only a child is designated, the server 20 may consider all images taken on the date specified by the user, on which the child's identification information has been added using any facial recognition technology, as the one designated by the user.

[0179] In the above embodiment, in step S12, the server 20 accepted a time range specification from the user. However, the server 20 may accept a time range specification instead of, or in addition to, a time range specification. In that case, the server 20 may determine that an image whose shooting time falls within a predetermined time range before or after the specified time is the image corresponding to the specified time. The server 20 may also include in the log information P7 of prompt P an instruction to generate a sentence about the child's state at the specified time, by referring to log information whose recorded time falls within a predetermined time range before or after the specified time.

[0180] In the above embodiment, in step S12, the server 20 received a time slot specification from the user. However, the server 20 may accept the specification of one or more scenes instead of, or in addition to, a time slot. A scene refers to any scene in child welfare activities, such as going to the nursery, morning assembly, or class activities. The server 20 may then generate an instruction to generate a draft text containing text related to the scene by determining an image corresponding to the scene specified by the user from images related to the child specified by the user, and by reading the situation of the image corresponding to the scene, and configure this as part of prompt P. Alternatively, the server 20 may use any image recognition technology to associate a specific image from among the images related to the child specified by the user with an image recording the scene specified by the user. The server 20 may then refer to the specific image and generate an instruction to generate a draft text containing text about the scene specified by the user, and configure this as part of prompt P.

[0181] In the above embodiment, the prompt generation module 2033 of server 20 generated a prompt into which log information was inserted. However, after inputting a prompt containing an instruction to read an image to the large language model, server 20 may receive a request for necessary log information from the large language model. Server 20 may then extract log information from the log management table 2027 in response to the request from the large language model and provide it to the large language model.

[0182] Specifically, first, the server 20 generates a prompt that includes instructions to generate draft text about a child based on a specific image, similar to the embodiment described above, and inputs it into the large-scale language model. At that time, for example, the server 20 inputs into the large-scale language model information about a function for searching the log management table 2027, which the large-scale language model can call from the server 20.

[0183] The large-scale language model reads the image content based on the input prompt, determines the log information that should be requested from server 20, and then requests it from server 20. For example, if the large-scale language model interprets the input image as a recording of a sleeping child, it will decide to request sleep-related log information from server 20. The large-scale language model then requests server 20 to call a function to retrieve sleep-related log information for the child specified in the prompt.

[0184] Server 20 searches the log management table 2027 based on the request received from the large language model and provides the necessary log information to the large language model. For example, server 20 executes a function received from the large-scale language model to search log management table 2027B, which stores log information related to sleep, based on the child ID of the child specified in the prompt. It then extracts the relevant log information and provides it to the large-scale language model.

[0185] The large-scale language model, while referring to the contents of the received log information, outputs a draft text about the image entered at the prompt. This allows the large-scale language model to include information that could not be gleaned from the image itself in the draft text by referring to the log information. Therefore, it becomes possible to output a draft text that provides a more detailed description of the children in child welfare facilities.

[0186] <Summary>

[0187] As described above, Server 20 receives a designation from the first user for a child or a group to which a child belongs as the target for generating draft documents. From the images stored in the database that record child welfare activities, Server 20 identifies one or more images as images related to the designated child or group. Server 20 generates a prompt that includes instructions to read the context of the identified images and instructions to generate draft documents about the designated child or group based on the read context. Server 20 inputs the prompt containing information about the images into a large-scale language model, retrieves the draft documents from the large-scale language model, and presents them to the first user. This allows the user to create documents by referring to the draft documents output from the large-scale language model. Therefore, the creation of documents can be made more efficient.

[0188] Furthermore, the draft document is a draft of the text to be included in documents that record activities related to child welfare. This will streamline the process of creating the text to be included in documents that record activities related to child welfare.

[0189] Furthermore, the server 20 accepts the specification of the document type and generates prompts containing different instructions depending on the specified type. This allows the server to present different document drafts depending on the type of document specified by the user.

[0190] Furthermore, the server 20 generates prompts containing information about the output format of the draft document, which varies depending on the specified type. This allows the server to present draft documents in different output formats depending on the type of document specified by the user.

[0191] Furthermore, the server 20 accepts the specification of one or more time periods and determines the image corresponding to the time period based on the time information associated with the specified time period and the identified image. By reading the status of the image corresponding to the time period, the server 20 generates a prompt that includes instructions to generate a draft text containing text related to the time period. This allows the server to present the user with a draft text corresponding to the specified time period.

[0192] Furthermore, Server 20 identifies multiple images as images related to children, accepts the specification of multiple time periods, and determines multiple images corresponding to each of the specified time periods. By reading the status of each of the multiple images corresponding to the multiple time periods, Server 20 generates a prompt that includes instructions for generating text divided into multiple time periods. This allows the user to be presented with draft text divided into multiple time periods, enabling the user to create text that clearly describes the children's behavior during each time period.

[0193] Furthermore, the server 20 accepts the specification of one or more scenes, determines the image corresponding to the scene from the images identified as being related to the child, reads the situation in the image corresponding to the scene, and generates a prompt that includes generating a draft text containing text related to the scene. This allows the server to present the user with a draft text corresponding to the specified scene.

[0194] Furthermore, server 20 acquires log information that associates information about the specified child's condition with time information, and generates a prompt that includes instructions to generate a draft text by referring to the log information. This allows the large-scale language model to refer to information other than images, enabling the presentation of a draft text that describes the child's condition in more detail.

[0195] Furthermore, the server 20 accepts the specification of one or more time periods and, by referring to the specified time periods and the time information contained in the log information, generates a prompt that includes instructions to generate a draft document containing text about the child's condition corresponding to the time period. This makes it possible to present a draft document that mentions the child's condition in more detail for the time period specified by the user.

[0196] Furthermore, Server 20 generates a prompt that includes instructions for selecting an attached image for the generated draft document, based on the image identified as relating to a child or similar person. Server 20 retrieves information about the attached image from a large-scale language model and presents the attached image to the user. This reduces the effort required for the user to select an image themselves, thereby streamlining document creation.

[0197] Furthermore, Server 20 stores draft texts obtained from a large-scale language model in a database, associating them with the children or groups designated as targets for evaluation. Server 20 accepts the designation of children or groups to be evaluated from the user. Server 20 extracts draft texts associated with the designated children or groups from the one or more draft texts stored in the database. Server 20 evaluates the designated children or groups to which the children belong by analyzing the text data of the extracted draft texts and presents the evaluation results to the user. Server 20 also performs text mining on the text data of the extracted draft texts and evaluates the designated children or groups by extracting the characteristics of the extracted draft texts. This allows for the evaluation of children based on the draft texts generated by Server 20. Since the draft texts are images of children in child welfare facilities, it is possible to appropriately evaluate the children's condition and behavior in child welfare facilities. Moreover, the draft texts are objectively output by a large-scale language model that reads the images, and are not texts created or modified by childcare workers or others. Therefore, the draft document does not include the subjective perspective of childcare workers or other staff members towards the children. Consequently, by analyzing the text data of the draft document, it is possible to make a more objective evaluation than one made subjectively by childcare workers or other staff members.

[0198] <4. Basic Hardware Configuration of a Computer> Figure 18 is a block diagram showing the basic hardware configuration of computer 90. Computer 90 includes at least a processor 94, main memory 95, auxiliary storage 96, and a communication interface 99. These are electrically connected to each other by a bus.

[0199] The processor 94 is hardware for executing the instruction set written in a program. The processor 94 consists of an arithmetic unit, registers, peripheral circuits, etc.

[0200] Main memory 95 is used to temporarily store programs and data processed by programs, etc. For example, it is volatile memory such as DRAM (Dynamic Random Access Memory).

[0201] Auxiliary storage device 96 refers to a storage device for saving data and programs. Examples include flash memory, HDD (Hard Disc Drive), magneto-optical disk, CD-ROM, DVD-ROM, and semiconductor memory.

[0202] A communication IF99 is an interface for inputting and outputting signals for communication with other computers via a network using wired or wireless communication standards.

[0203] A network consists of various mobile communication systems built using the internet, LANs, wireless base stations, etc. For example, a network includes 3G, 4G, and 5G mobile communication systems, LTE (Long Term Evolution), and wireless networks that can connect to the internet via designated access points (e.g., Wi-Fi®). When connecting wirelessly, communication protocols include, for example, Z-Wave®, ZigBee®, and Bluetooth®. When connecting via a wired connection, the network also includes connections made directly via USB (Universal Serial Bus) cables, etc.

[0204] Furthermore, by distributing all or part of each hardware configuration across multiple computers 90 and connecting them to each other via a network, a computer 90 can be virtually realized. Thus, the concept of computer 90 includes not only a computer 90 housed in a single enclosure or case, but also a virtualized computer system.

[0205] <Basic Functional Configuration of Computer 90> The functional configuration of the computer realized by the basic hardware configuration of computer 90 shown in Figure 18 will be explained. The computer comprises at least one functional unit: a control unit, a memory unit, and a communication unit.

[0206] Furthermore, the functional units of computer 90 can also be realized by distributing all or part of each functional unit across multiple computers 90 interconnected via a network. The concept of computer 90 includes not only a single computer 90 but also a virtualized computer system.

[0207] The control unit is realized when the processor 94 reads various programs stored in the auxiliary storage device 96, loads them into the main memory device 95, and executes processing according to those programs. The control unit can realize various functional units that perform information processing depending on the type of program. In this way, the computer is realized as an information processing device that performs information processing.

[0208] The memory unit is implemented by a main memory 95 and an auxiliary memory 96. The memory unit stores data, various programs, and various databases. The processor 94 can also reserve memory areas corresponding to the memory unit in the main memory 95 or the auxiliary memory 96 according to the program. The control unit can also cause the processor 94 to perform operations such as adding, updating, and deleting data stored in the memory unit according to the various programs.

[0209] A database, specifically a relational database, is used to manage and link data sets called tables, which are structurally defined by rows and columns. In a database, tables are called tables, the columns of a table are called columns, and the rows of a table are called records. In a relational database, relationships can be established and linked between tables.

[0210] Typically, each table has a key column to uniquely identify records, but setting a key on a column is not mandatory. The control unit can instruct the processor 94 to add, delete, or update records in specific tables stored in the memory unit according to various programs.

[0211] The communication unit is implemented by the communication IF99. The communication unit provides the functionality to communicate with other computers 90 via the network. The communication unit can receive information transmitted from other computers 90 and input it to the control unit. The control unit can cause the processor 94 to perform information processing on the received information according to various programs. The communication unit can also transmit information output from the control unit to other computers 90.

[0212] The functions realized by the components described herein may be implemented in a circuitry or processing circuitry, including general-purpose processors, application-specific processors, integrated circuits, ASICs (Application Specific Integrated Circuits), CPUs (a Central Processing Unit), conventional circuits, and / or combinations thereof, programmed to realize the functions described herein. A processor includes transistors and other circuits and is considered a circuitry or processing circuitry. A processor may be a programmed processor that executes a program stored in memory. In this specification, circuitry, unit, and means are hardware programmed to perform or execute the functions described herein. Such hardware may be any hardware disclosed herein, or any hardware known to be programmed to perform or execute the functions described herein. If the hardware is a processor that is considered to be a type of circuitry, then the circuitry, means, or unit is a combination of hardware and software used to constitute the hardware and / or processor.

[0213] While several embodiments of this disclosure have been described above, these embodiments can be implemented in a variety of other forms, and various omissions, substitutions, and modifications are permitted without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents.

[0214] <Note> The details described in each of the above embodiments are noted below. (Note 1) A program for operating a computer comprising a processor and memory, the program causing the processor to perform the following steps: receiving a designation from a first user of a child or a group to which a child belongs as the target for generating draft text; identifying one or more images from among images recorded in a database that depict child welfare activities as images related to the designated child or group as the target for generating text; generating a prompt that includes reading the context of the identified images and generating draft text about the designated child or group based on the context read; obtaining draft text from a large-scale language model by inputting a prompt containing information about the images into the large-scale language model; and presenting the obtained draft text to the first user. (Note 2) The draft document is a draft of the text to be included in the document for recording activities related to child welfare, as described in Appendix 1. (Note 3) The program described in Appendix 1 or Appendix 2, which, in the step of receiving the specification of the target to be generated, accepts the specification of the type of document, and in the step of generating a prompt, generates a prompt containing different instructions depending on the specified type. (Note 4) The program described in Appendix 3 generates a prompt that includes information about the output format of the draft text, which varies depending on the specified type, in the step of generating the prompt. (Note 5) The program described in any of Appendix 1 to Appendix 4, wherein in the step of receiving a specification of the target to be generated, the program receives a specification of one or more time periods, the program causes the processor to perform a step of determining an image corresponding to the time period based on the specified time period and time information associated with the identified image, and in the step of generating a prompt, the program generates a prompt that includes an instruction to generate a draft text containing text about the time period by reading the status of the image corresponding to the time period. (Note 6) The program described in Appendix 5, wherein in the step of identification, multiple images are identified; in the step of receiving the specification of the target to be generated, multiple time zones are specified; in the step of determining the image corresponding to the time zone, multiple images corresponding to each of the specified time zones are determined; and in the step of generating a prompt, the status of the multiple images corresponding to the multiple time zones is read, thereby generating a prompt that includes instructions to generate sentences separated by multiple time zones. (Note 7) A program according to any one of the appendices 1 to 6, wherein in the step of receiving the specification of the target to be generated, the program receives the specification of one or more scenes, and in the step of generating a prompt, the program generates a prompt that includes determining an image corresponding to the scene from the specified images, reading the situation of the image corresponding to the scene, and generating a draft text that includes text about the scene. (Note 8) The program is one of the programs described in any of Appendix 1 to Appendix 7, which causes a processor to perform the step of obtaining log information in which information about the state of children belonging to a specified child or group is recorded in association with time information, and in the step of generating a prompt, generates a prompt that includes an instruction to generate a draft text by referring to the log information. (Note 9) The program described in Appendix 8, which, in the step of receiving the specification of the target to be generated, accepts the specification of one or more time periods, and in the step of generating a prompt, refers to the specified time period and the time information contained in the log information to generate a prompt that includes an instruction to generate a draft text containing text about the state of a child corresponding to the time period. (Note 10) A program according to any one of the appendices 1 to 9, wherein in the step of generating a prompt, it generates a prompt that includes an instruction to select an attached image for the generated draft text from an identified image; in the step of obtaining the draft text, it obtains information about the attached image; and in the step of presenting the draft text, it presents the obtained attached image. (Note 11) The program is one of the programs described in any of Appendix 1 to Appendix 10, which causes the processor to perform the following steps: storing draft sentences obtained from a large-scale language model in a database in association with a child or group designated as the target of generation; receiving a designation from a second user for a child or group to which a child belongs as the target of evaluation; extracting draft sentences from one or more draft sentences stored in the database that are associated with the child or group to which a child belongs designated as the target of evaluation; evaluating the child or group to which a child belongs designated as the target of evaluation by analyzing the text data of the extracted draft sentences; and presenting the evaluation results to the second user. (Note 12) The program described in Appendix 11 evaluates the designated child or the group to which the child belongs by performing text mining on the extracted text data of the draft texts during the evaluation step and extracting the characteristics of the extracted draft texts. (Note 13) A method performed by a computer equipped with a processor, the method comprising: the step of receiving from a first user the designation of a child or a group to which a child belongs as the target for generating draft text; the step of identifying one or more images from among images recorded in a database that depict child welfare activities as images related to the designated child or group as the target for generating text; the step of generating a prompt that includes reading the context of the identified images and generating draft text about the designated child or group based on the context read; the step of inputting the prompt containing information about the images into a large-scale language model to obtain draft text from a large-scale language model; and the step of presenting the obtained draft text to the first user. (Note 14) An information processing device comprising a control unit, wherein the control unit performs the following steps: receiving a designation from a first user of a child or a group to which a child belongs as the target for generating a draft document; identifying one or more images from among images recorded in a database that depict child welfare activities as images related to the designated child or group as the target for generation; generating a prompt that includes reading the context of the identified images and generating a draft document about the designated child or group based on the context read; obtaining a draft document from a large-scale language model by inputting a prompt containing information about the images into a large-scale language model; and presenting the obtained draft document to the first user. (Note 15) A system comprising: means for receiving a designation from a first user of a child or a group to which a child belongs as the target for generating draft text; means for identifying one or more images from a database of images recording child welfare activities as images related to the designated child or group as the target for generation; means for generating a prompt that includes reading the context of the identified images and generating a draft text about the designated child or group based on the read context; means for obtaining a draft text from a large-scale language model by inputting a prompt containing information about the images into the large-scale language model; and means for presenting the obtained draft text to the first user. [Explanation of symbols]

[0215] 1...System 10...Terminal device 12...Communication interface 120...Communication unit 13...Input device 14...Output device 141...Display 15...Memory 16...Storage 17...Audio processing unit 180...Storage unit 19...Processor 190...Control unit 20...Server

Claims

1. A program for operating a computer comprising a processor and memory, wherein the program is configured to operate the processor, The first step involves receiving a designation from the first user as the target for generating draft texts, or a designation of a child or a group to which a child belongs. The steps include: identifying one or more images from among the images recorded in the database that depict child welfare activities as images related to the child or group designated as the target for generation; A step of generating a prompt that includes reading the situation of the identified image and generating a draft text relating to the child or group designated as the target of the generation, based on the situation read. The steps include: inputting the prompt containing information about the image into the large-scale language model to obtain the draft text from the large-scale language model; The steps include presenting the acquired draft document to the first user, A program that executes something.

2. The program according to claim 1, wherein the aforementioned draft text is a draft text to be included in a document for recording activities related to child welfare.

3. In the step of receiving the designation of the target to be generated, the designation of the type of document is received, In the step of generating the prompt, a prompt is generated that includes different instructions depending on the specified type. The program according to claim 1.

4. In the step of generating the prompt, a prompt is generated that includes information regarding the output format of the draft document, which varies depending on the specified type. The program according to claim 3.

5. In the step of receiving the designation of the target to be generated, the designation of one or more time slots is accepted. The program causes the processor to perform the step of determining an image corresponding to a specified time period based on the time information associated with the identified image. In the step of generating the prompt, the system reads the image status corresponding to the time period and generates a prompt that includes an instruction to generate a draft text containing text related to the time period. The program according to claim 1.

6. In the aforementioned identification step, multiple images are identified, In the step of receiving the specification of the target to be generated, the specification of multiple time periods is accepted, In the step of determining the image corresponding to the aforementioned time period, a plurality of images corresponding to each of the specified plurality of time periods are determined, In the step of generating the prompt, the system reads the status of each of the multiple images corresponding to the multiple time periods, and generates a prompt that includes an instruction to generate sentences separated for each of the multiple time periods. The program according to claim 5.

7. In the step of receiving the specification of the target to be generated, the specification of one or more scenes is received, In the step of generating the prompt, the prompt is generated which includes determining an image corresponding to the scene from the identified images, reading the situation in the image corresponding to the scene, and generating a draft text that includes text relating to the scene. The program according to claim 1.

8. The program is provided to the processor: The system performs the step of obtaining log information in which information regarding the status of a designated child or a child belonging to the group is associated with time information. In the step of generating the prompt, a prompt is generated that includes an instruction to generate the draft text by referring to the log information. The program according to claim 1.

9. In the step of receiving the designation of the target to be generated, the designation of one or more time slots is accepted. In the step of generating the prompt, the system generates a prompt that includes an instruction to generate a draft text containing text relating to the child's condition corresponding to the specified time period, by referring to the specified time period and the time information contained in the log information. The program according to claim 8.

10. In the step of generating the prompt, a prompt is generated that includes an instruction to select an attached image for the generated draft text from the identified image. In the step of obtaining the aforementioned draft document, information regarding the attached image is obtained, In the step of presenting the draft document, the acquired attached image is presented. The program according to claim 1.

11. The program is provided to the processor: The steps include: storing the draft text obtained from the large-scale language model in a database in association with the child or group designated as the target of generation; The second user then specifies the child or the group to which the child belongs as the target of evaluation. The steps include: extracting from one or more draft documents stored in the database a draft document associated with the child designated as the subject of evaluation or the group to which the child belongs; The steps include: analyzing the text data of the extracted draft text to evaluate the child or group to which the child belongs, and The steps include presenting the evaluation results to the second user, The program according to claim 1, which causes to execute.

12. In the evaluation step, text mining is performed on the extracted text data of the draft text, and the characteristics of the extracted draft text are extracted to evaluate the child or the group to which the child belongs, which is designated as the subject of evaluation. The program according to claim 11.

13. A method performed by a computer having a processor, wherein the processor The first step involves receiving a designation from the first user as the target for generating draft texts, or a designation of a child or a group to which a child belongs. The steps include: identifying one or more images from among the images recorded in the database that depict child welfare activities as images related to the child or group designated as the target for generation; A step of generating a prompt that includes reading the situation of the identified image and generating a draft text relating to the child or group designated as the target of the generation, based on the situation read. The steps include: inputting the prompt containing information about the image into the large-scale language model to obtain the draft text from the large-scale language model; The steps include presenting the acquired draft document to the first user, How to do it.

14. An information processing apparatus comprising a control unit, wherein the control unit is The first step involves receiving a designation from the first user as the target for generating draft texts, or a designation of a child or a group to which a child belongs. The steps include: identifying one or more images from among the images recorded in the database that depict child welfare activities as images related to the child or group designated as the target for generation; A step of generating a prompt that includes reading the situation of the identified image and generating a draft text relating to the child or group designated as the target of the generation, based on the situation read. The steps include: inputting the prompt containing information about the image into the large-scale language model to obtain the draft text from the large-scale language model; The steps include presenting the acquired draft document to the first user, An information processing device that performs the following actions.

15. A means for receiving a designation from the first user as the target for generating draft text, a child or a group to which a child belongs, A means for identifying one or more images from among the images of child welfare activities stored in the database as images related to the child or group designated as the target of generation, A means for generating a prompt, which includes reading the situation of the identified image and generating a draft text relating to the child or group designated as the target of the generation, based on the situation read. A means for obtaining the draft text from the large-scale language model by inputting the prompt containing information about the image into the large-scale language model, A means for presenting the acquired draft document to the first user, A system that is equipped with [the following].

Images