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

Abstract

This invention provides an information processing method that allows for the easy output of childcare records and instructional plans for kindergarten children. [Solution] An information processing method according to one embodiment of the present disclosure acquires images of children taken in the nursery school and outputs childcare records and instructional plans by providing the images and the nursery school childcare guidelines to a language model. Furthermore, an information processing method according to one embodiment of the present disclosure outputs childcare records and instructional plans by providing the language model with instructions to refer to the nursery school childcare guidelines according to the age of the child in the images taken in the nursery school.

Description

Technical Field

[0001] This technology relates to an information processing method, a program, an information processing apparatus, and an information processing system.

Background Art

[0002] Conventionally, a system for recording and aggregating children's activities has been disclosed. For example, in the system described in Patent Document 1, a play corner is provided in a nursery room, a tablet and an IC card reader are arranged at each play corner, the IC card reader reads an IC chip when a child starts and ends playing, and the server records, as data, the activities of the child in the play corner of the child based on the child number of the IC chip of the child read by the IC card reader.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, the system described in Patent Document 1 does not consider outputting a simple nursery record and guidance plan for kindergarten children.

[0005] This disclosure has been made in view of such circumstances, and an object thereof is to provide an information processing method or the like that can simply output a nursery record and a guidance plan for kindergarten children. [[ID=4x]]

Means for Solving the Problems

[0006] An information processing method according to an embodiment of the present disclosure acquires an image of a kindergarten child, and outputs a nursery record and a guidance plan by providing the image and nursery guidelines to a language model.

Effects of the Invention

[0007] In one embodiment of the information processing method described herein, it is possible to easily output childcare records and instructional plans for children. [Brief explanation of the drawing]

[0008] [Figure 1] This is an explanatory diagram showing an overview of the information processing system. [Figure 2] This is a block diagram showing an example of the configuration of an information processing device. [Figure 3] This is a block diagram showing an example configuration of a terminal device. [Figure 4] This is an explanatory diagram showing an example of a record layout for an image database. [Figure 5] This is an explanatory diagram showing an example of a record layout for the child database. [Figure 6] This is an explanatory diagram showing an example of the record layout for the overall childcare record database. [Figure 7] This is an explanatory diagram showing an example of a personal childcare record database. [Figure 8] This is an explanatory diagram showing an example of a child extraction module. [Figure 9] This is an explanatory diagram showing an example of input to the language model related to the overall daily childcare record generation process. [Figure 10] This is an explanatory diagram showing an example of the output of the language model in the overall daily childcare record generation process. [Figure 11] This is an explanatory diagram showing an example of input to the language model in the overall summary output processing. [Figure 12] This is an explanatory diagram showing an example of the output of a language model in the overall summary output processing. [Figure 13] This is an explanatory diagram showing an example of input to the language model related to the generation process of individual daily childcare records. [Figure 14] This is an explanatory diagram showing an example of the output of the language model in the individual daily childcare record generation process. [Figure 15] This is an explanatory diagram illustrating an example of input to a language model in the personal summary output processing. [Figure 16] It is an explanatory diagram showing an example of the output to the language model in the personal summary output process. [Figure 17] It is an explanatory diagram showing an example of the overall childcare record correction screen. [Figure 18] It is an explanatory diagram showing an example of the personal childcare record correction screen. [Figure 19] It is a flowchart showing an example of the child extraction process. [Figure 20] It is a flowchart showing an example of the overall daily childcare record generation process. [Figure 21] It is a flowchart showing an example of the overall summary output process. [Figure 22] It is a flowchart showing an example of the personal daily childcare record generation process. [Figure 23] It is a flowchart showing an example of the personal summary output process. [Figure 24] It is a flowchart showing an example of the overall childcare record correction process. [Figure 25] It is a flowchart showing an example of the personal childcare record correction process. [Figure 26] It is an explanatory diagram showing an example of the input to the language model in the personal summary output process according to Embodiment 2. [Figure 27] It is an explanatory diagram showing an example of the output of the language model M in the personal summary output process.

Mode for Carrying Out the Invention

[0009] (Embodiment 1) When the acquired image of the child and the childcare guidelines of the nursery are input, the language model M according to Embodiment 1 outputs childcare records and a guidance plan (activity plan). Hereinafter, Embodiment 1 will be described with reference to the drawings.

[0010] Figure 1 is an explanatory diagram showing an overview of the information processing system S. The information processing system S comprises an information processing device 1, a terminal device 2, a camera 3, and a language model server 4. The camera 3 is a camera installed in the nursery school (for example, a monitoring camera) and transmits images of the children in the nursery school to the information processing device 1. The camera 3 may be a portable camera, a wearable camera, or a camera built into smart glasses, a smartphone, a tablet terminal, or a personal computer. The information processing device 1 transmits the images acquired from the camera 3, along with the nursery school childcare guidelines, to the language model server 4 and inputs them into the language model M of the language model server 4. Hereafter, the transmission of information to the language model server 4 and the input of information to the language model M are considered synonymous. The information processing device 1 acquires the childcare record and lesson plan output by the language model M, which has received the images and nursery school childcare guidelines as input, stores the acquired childcare record and lesson plan, and transmits them to the terminal device 2 upon request from the terminal device 2. The terminal device 2 is, for example, a terminal device used by a childcare worker (user). The recording device may be installed in kindergartens, schools, or after-school childcare facilities, and the language model M may be inputted with the Kindergarten Education Guidelines, Curriculum Guidelines, or After-School Childcare Club Operation Guidelines. In this case, the language model M may also output learning records.

[0011] Figure 2 is a block diagram showing an example configuration of the information processing device 1. The information processing device 1 is, for example, a server computer and comprises a processing unit 11, a storage unit 12, and a communication unit 13. The processing unit 11 is composed of a CPU (Central Processing Unit), MPU (Micro Processing Unit), GPU (Graphical Processing Unit), or quantum processor, and performs various control processing, arithmetic processing, etc., by reading and executing a program P (program product) and a database pre-stored in the storage unit 12. A database server or the like may be provided outside the information processing device 1, and the database may be read from the database server or the like. Furthermore, the information processing device 1 may have its functions realized by multiple server devices or computers. Also, the information processing device 1 may correspond to a node on a blockchain. Furthermore, the processing unit 11 of the information processing device 1 may execute some or all of the processing performed by the terminal device 2.

[0012] The storage unit 12 of the information processing device 1 is, for example, a volatile memory and a non-volatile memory. The storage unit 12 stores a program P, an image DB (Data Base) 121, a child DB 122, a general childcare record DB 123, and an individual childcare record DB 124. The program P may be provided to the information processing device 1 using a storage medium 12a that is readable by a computer. The storage medium 12a is, for example, a portable memory. Examples of portable memory include CD-ROMs, USB (Universal Serial Bus) memory, SD cards, microSD cards, or CompactFlash® memory. If the storage medium 12a is a portable memory, the processing element of the processing unit 11 may read the program P from the storage medium 12a using a reading device (not shown). The read program P is written to the storage unit 12. Furthermore, the program P may be provided to the information processing device 1 by the communication unit 13 communicating with an external device. Details of Image DB121, Child DB122, Overall Childcare Record DB123, and Individual Childcare Record DB124 will be described later.

[0013] The communication unit 13 is a communication module or communication interface for communicating with the terminal device 2, the imaging device 3, and the language model server 4 by wired or wireless connection, and is, for example, a wide-area wireless communication module such as LTE®, 4G, or 5G. The processing unit 11 communicates with the terminal device 2, the imaging device 3, and the language model server 4 via the communication unit 13, for example, through an external network N such as the Internet.

[0014] The language model server 4 stores the language model M. Details of the language model M will be described later. Note that the language model M may also be stored in the information processing device 1 or the terminal device 2.

[0015] Figure 3 is a block diagram showing an example configuration of the terminal device 2. The terminal device 2 is, for example, a smartphone, a tablet terminal, or a personal computer. In this embodiment, the example of the terminal device 2 being a personal computer will be described. The terminal device 2 comprises a terminal processing unit 21, a storage unit 22, a communication unit 23, an input unit 24, and a display unit 25. The terminal processing unit 21 is composed of a CPU or MPU, and performs various control processing, arithmetic processing, etc. The terminal device 2 may perform all or part of the processing performed by the information processing device 1 or the language model server 4. Furthermore, the terminal device 2 may be configured as an integral part of the imaging device 3.

[0016] The storage unit 22 stores an application program Pa for which childcare records and lesson plans output by the language model M, acquired from the information processing device 1, are displayed, and which accepts corrections to the childcare records and lesson plans. The application program Pa is provided to the terminal device 2, for example, using the storage medium 22a. The terminal processing unit 21 of the terminal device 2 may acquire the application program Pa using the internet and store it in the storage unit 22.

[0017] The input unit 24 receives input from the user (childcare worker) regarding the selection of childcare records and lesson plans to display, as well as corrections to the childcare records and lesson plans. The input unit 24 is composed of, for example, a keyboard or mouse.

[0018] The display unit 25 displays childcare records and instruction plans acquired from the information processing device 1.

[0019] Figure 4 is an explanatory diagram showing an example of the record layout of the image database 121. The image database 121 stores images taken by the camera 3 and information about the children pictured in the images. The management items (fields) of the image database 121 include the image ID field, image data field, shooting date and time field, photographed child field, and photographed class field.

[0020] The image ID field in the image DB 121 stores the ID assigned to each image captured by the camera 3. The image data field stores the captured images, for example, in file format. The date and time of capture field stores the date and time the image was captured. The child captured field stores the child ID of the child in the image, output by the child extraction module (see Figure 8) that received the image as input. The class captured field stores the class ID of the class to which the child in the image belongs. The class ID is identified based on the information associated with the child's child ID. If multiple children in a single image belong to different classes, multiple class IDs are stored in the class captured field. The storage unit 12 may also store multiple image DBs 121 with class IDs as attribute information. In this case, the class ID of the class installed in each of the multiple camera 3 is associated with the image and stored as identification information, and each image is stored in the image DB corresponding to its respective class ID based on the identification information of the camera 3 that captured the image.

[0021] Figure 5 is an explanatory diagram showing an example of the record layout of the child database 122. The child database 122 stores information related to children. The management items (fields) of the child database 122 include the child ID field, name field, class affiliation field, age field, feature field, and target image field.

[0022] The child ID field in the child DB122 stores the ID assigned to the child. The name field stores the child's name. The class field stores the class ID of the class the child belongs to. The age field stores the child's age. The feature field stores the child's features (vector data) extracted, for example, by a learning model from the child's face photograph which has been previously associated with the child ID. The target image field stores the image ID of the image in which the child is pictured. Note that the target image field stores image IDs related to multiple images, and each time the child extraction module (see Figure 8) outputs the child's child ID, the image ID of the image input to the child extraction module is stored in the target image field.

[0023] Figure 6 is an explanatory diagram showing an example of the record layout of the overall childcare record DB123. The overall childcare record DB123 stores childcare records for the entire nursery school, not specific to any particular child. The overall childcare record DB123 includes the overall daily table TAD, overall weekly table TAW, overall monthly table TAM, and overall yearly table TAY. The management items (fields) of each table in the overall childcare record DB123 include the class ID field, record period (day, month / week, month, or year) field, childcare record draft field, childcare record proofreading field, parent contact draft field, parent contact proofreading field, activity content field, activity objective field, child's condition field, observation field, next period (next day, next week, next month, or next year) instruction plan field, health activity evaluation field, human relations activity evaluation field, environmental activity evaluation field, language activity evaluation field, and expressive activity evaluation field.

[0024] The overall daily table TAD stores information output by language model M, which has been inputted with multiple images taken on each day and the childcare guidelines for the nursery school. The class ID field of the overall daily table TAD stores the class ID of the class in which the input image was taken when language model M outputs the childcare record. The record date field stores the date on which the input image was taken when language model M outputs the childcare record. The childcare record draft field stores the childcare record (childcare record draft) for the relevant day output by language model M with the input image. The childcare record correction field stores the childcare record for the relevant day (childcare record correction) that has been corrected by terminal device 2 based on the childcare record draft. The parent contact draft field stores the parent contact (parent contact draft) for the relevant day output by language model M with the input image. The parent contact correction field stores the parent contact (parent contact correction) for the relevant day that has been configured by terminal device 2 based on the parent contact draft.

[0025] The activity content field, activity objective field, child's condition field, observation field, next day's lesson plan field, health activity evaluation field, interpersonal relationship activity evaluation field, environmental activity evaluation field, language activity evaluation field, and expressive activity evaluation field of the overall daily table TAD store the activity content, activity objective, child's condition, observation, next day's lesson plan, health-related activity evaluation, interpersonal relationship activity evaluation, environmental activity evaluation, language activity evaluation, and expressive activity evaluation for the day output by the language model M to which the image has been input.

[0026] The overall weekly table TAW stores information output by language model M, which has been inputted with the contents of records from the overall daily table TAD corresponding to the date of each week. The class ID field of the overall weekly table TAW stores the class ID of the records from the input overall daily table TAD when language model M outputs childcare records. The record week field stores the month and week to which the dates of the multiple records from the input overall daily table TAD correspond. The childcare record draft field stores the childcare record (childcare record draft) for the relevant week output by language model M, which has been input with multiple records from the overall daily table TAD. The childcare record correction field stores the childcare record for the relevant week (childcare record correction) that has been corrected on terminal device 2 based on the childcare record draft. The parent contact draft field stores the parent contact (parent contact draft) for the relevant week output by language model M, which has been input with multiple records from the overall daily table. The Parent Contact Correction field stores the parent contact (parent contact correction) for the relevant week, which is configured in terminal device 2 based on the parent contact draft.

[0027] The activity content field, activity objective field, child's condition field, observation field, next week's lesson plan field, health activity evaluation field, interpersonal relationship activity evaluation field, environmental activity evaluation field, language activity evaluation field, and expressive activity evaluation field of the overall weekly table TAW contain the activity content, activity objective, child's condition, observation, next week's lesson plan, health-related activity evaluation, interpersonal relationship activity evaluation, environmental activity evaluation, language activity evaluation, and expressive activity evaluation for the relevant week, which are output by the language model M that has multiple records from the overall daily table TAD as input.

[0028] The overall monthly table TAM stores information output by language model M, which has been inputted with the contents of the records for the week corresponding to each month from the records in the overall weekly table TAW. The class ID field of the overall monthly table TAM stores the class ID of the records in the input overall weekly table TAW when language model M outputs childcare records. The record month field stores the month to which the weeks of the multiple records in the input overall weekly table TAW belong. The childcare record draft field stores the childcare record (childcare record draft) for the relevant month output by language model M, which has been input with multiple records from the overall weekly table. The childcare record correction field stores the childcare record for the relevant month (childcare record correction) that has been corrected on terminal device 2 based on the childcare record draft. The parent contact draft field stores the parent contact (parent contact draft) for the relevant month output by language model M, which has been input with multiple records from the overall weekly table. The parent contact correction field stores the parent contact (parent contact correction) for the relevant month that has been configured on terminal device 2 based on the parent contact draft.

[0029] The activity content field, activity objective field, child's condition field, observation field, next month's lesson plan field, health activity evaluation field, interpersonal relationship activity evaluation field, environmental activity evaluation field, language activity evaluation field, and expression activity evaluation field of the overall monthly table TAM store the activity content, activity objective, child's condition, observation, next month's lesson plan, health-related activity evaluation, interpersonal relationship activity evaluation, environmental activity evaluation, language activity evaluation, and expression activity evaluation for the relevant month, which are output by the language model M into which multiple records from the overall weekly table are input.

[0030] The overall annual table TAY stores information output by language model M, which has been inputted with the contents of the records for the month corresponding to each year from the overall monthly table TAM. Note that records may be created for each year in the overall annual table TAY. The class ID field of the overall annual table TAY stores the class ID of the records in the input overall monthly table TAM when language model M outputs childcare records. The record year field stores the year to which the months of the multiple records in the input overall monthly table TAM correspond. The childcare record draft field stores the childcare record (childcare record draft) for the relevant year output by language model M, which has been input with multiple records from the overall monthly table. The childcare record correction field stores the childcare record for the relevant year (childcare record correction) that has been corrected on terminal device 2 based on the childcare record draft. The parent contact draft field stores the parent contact (parent contact draft) for the relevant year output by language model M, which has been input with multiple records from the overall monthly table. The Parent Contact Proofreading field stores the parent contact (parent contact proofreading) for the relevant year, which is configured in terminal device 2 based on the parent contact draft.

[0031] The activity content field, activity objective field, child's condition field, awareness field, next year's instruction plan field, health activity evaluation field, interpersonal relationship activity evaluation field, environmental activity evaluation field, language activity evaluation field, and expression activity evaluation field of the overall annual table TAY contain the activity content, activity objective, child's condition, awareness, next year's instruction plan, health-related activity evaluation, interpersonal relationship activity evaluation, environmental activity evaluation, language activity evaluation, and expression activity evaluation for the relevant year, which are output by the language model M into which multiple records from the overall monthly table are input.

[0032] Figure 7 is an explanatory diagram showing an example of the Individual Childcare Record DB124. The Individual Childcare Record DB124 stores childcare records for each child. The Individual Childcare Record DB124 includes the Individual Daily Table TID, Individual Weekly Table TIW, Individual Monthly Table TIM, and Individual Yearly Table TIY. The management items (fields) for each table in the Individual Childcare Record DB124 include the Class ID field, Child ID field, Record Period (Day, Monthly / Week, Month, or Year) field, Childcare Record Draft field, Childcare Record Correction field, Parent Contact Draft field, Parent Contact Correction field, Health Development field, Interpersonal Relationship Development field, Environmental Development field, Language Development field, and Expression Development field.

[0033] The individual daily table TID stores information output by language model M, which has been input with multiple images of specific children taken on each day and information from the corresponding record in the overall daily table TAD (such as childcare records). The class ID field of the individual daily table TID stores the class ID of the class to which the child in the input image belongs when language model M outputs childcare records. The child ID field stores the child ID of the child in the input image when language model M outputs childcare records. The record date field stores the date the input image was taken when language model M outputs childcare records. The childcare record draft field stores the childcare record (childcare record draft) for the specific child for the corresponding day, output by language model M with the input image. The childcare record correction field stores the childcare record for the corresponding day (childcare record correction) that has been corrected by terminal device 2 based on the childcare record draft. The Parent Contact Draft field stores the parent contact (parent contact draft) for a specific child on the relevant day, output by the language model M that received the image as input. The Parent Contact Correction field stores the parent contact (parent contact correction) for the relevant day, which was prepared by terminal device 2 based on the parent contact draft.

[0034] The Health Development field, Relationship Development field, Environment Development field, Language Development field, and Expression Development field in the individual daily table TID store the evaluations of the child's health development, relationship development, environment development, language development, and expression development for that day, which were output by a language model M that received an image of the specific child (specific image) as input.

[0035] The individual weekly table TIW stores information output by the language model M, which has been inputted with the contents of the records in the individual daily table TID corresponding to the date of each week. The class ID field in the individual weekly table TIW stores the class ID of the class to which the specific child belongs, related to the record in the individual daily table TID. The child ID field stores the child ID of the specific child. The record week field stores the month and week to which the dates of the multiple records in the individual daily table TID input into the language model M correspond. The childcare record draft field stores the childcare record (childcare record draft) for the specific child for the week in question, output by the language model M, which has been inputted with the multiple records in the individual daily table TID. The childcare record correction field stores the childcare record for the day in question (childcare record correction) that has been corrected on terminal device 2 based on the childcare record draft. The parent contact draft field stores the parent contact (parent contact draft) for the specific child for the day in question, output by the language model M, which has been inputted with the multiple records in the individual daily table TID. The Parent Contact Proofreading field stores the parent contact (parent contact proofreading) for the relevant day, which is configured in terminal device 2 based on the parent contact draft.

[0036] The Health Development field, Relationship Development field, Environment Development field, Language Development field, and Expression Development field in the individual weekly table TIW store the evaluations of the health development, relationship development, environment development, language development, and expression development of the specific child in question for that week, which are output by the language model M that has multiple records from the individual daily table TID as input.

[0037] The individual monthly table TIM stores information output by language model M, which has been inputted with the contents of the records for the week corresponding to each month from the individual weekly table TIW. The class ID field of the individual monthly table TIM stores the class ID of the class to which a specific child belongs, related to the record in the individual weekly table TIW. The child ID field stores the child ID of a specific child. The recording month field stores the month to which the weeks of the multiple records in the individual weekly table TIW input into language model M correspond. The childcare record draft field stores the childcare record (childcare record draft) for the relevant month, output by language model M, which has been inputted with multiple records from the individual weekly table TIW. The childcare record correction field stores the childcare record for the relevant week (childcare record correction) that has been corrected by terminal device 2 based on the childcare record draft. The parent contact draft field stores the parent contact (parent contact draft) for the relevant week of a specific child, output by language model M, which has been inputted with multiple records from the individual weekly table TIW. The Parent Contact Correction field stores the parent contact (parent contact correction) for the relevant week, which is configured in terminal device 2 based on the parent contact draft.

[0038] The Health Development field, Relationship Development field, Environment Development field, Language Development field, and Expression Development field in the individual monthly table TIM store the evaluations of the health development, relationship development, environment development, language development, and expression development of the specific child in question for that month, which are output by the language model M, into which multiple records from the individual weekly table TIW have been input.

[0039] The individual yearly table TIY stores information output by language model M, which has been inputted with the contents of the records for the month corresponding to each year from the individual monthly table TIM. The class ID field of the individual yearly table TIY stores the class ID of the class to which a specific child belongs, related to the record in the individual monthly table TIM. The child ID field stores the child ID of a specific child. The record year field stores the year to which the months of the multiple records in the individual monthly table TIM input into language model M correspond. The childcare record draft field stores the childcare record (childcare record draft) for a specific child for the year, output by language model M, which has been inputted with multiple records from the individual monthly table TIM. The childcare record correction field stores the childcare record for the month in question (childcare record correction) that has been corrected by terminal device 2 based on the childcare record draft. The parent contact draft field stores the parent contact (parent contact draft) for a specific child for the month in question, output by language model M, which has been inputted with multiple records from the individual monthly table TIM. The Parent Contact Proofreading field stores the parent contact (parent contact proofreading) for the relevant month, which is configured in terminal device 2 based on the parent contact draft.

[0040] The Health Development field, Relationship Development field, Environment Development field, Language Development field, and Expression Development field in the individual yearly table TIY store evaluations of the health development, relationship development, environment development, language development, and expression development of a specific child in that year, which are output by the language model M, into which multiple records from the individual monthly table TIM have been input.

[0041] Figure 8 is an explanatory diagram showing an example of a child extraction module 125. The child extraction module 125 is composed of, for example, an autoencoder. When an image is input to the child extraction module 125, the child extraction module 125 refers to the child ID and child feature vector pair in the child DB 122 and outputs the child ID corresponding to the feature vector that is most similar to the feature vector of the child in the image. The child extraction module 125 may also be composed of an object detection machine learning model including a neural network such as CNN (Convolutional Neural Network), R-CNN (Regions-CNN), Fast R-CNN, Faster R-CNN, SSD (Single Shot Multibook Detector), or YOLO (You Only Look Once). In this case, the child extraction module 125 is pre-trained to output the child ID of the child in the image when an image is input, using training data that associates child IDs with child feature vectors recorded in the child DB 122. Furthermore, the child selection module 125 may use methods other than neural networks, such as decision trees, k-nearest neighbors, or SVM.

[0042] The child ID output by the child extraction module 125 is stored in the child field of the record related to the image input to the child extraction module 125 in the image database 121. In addition, the target image field in the child database 122 related to the child ID output by the child extraction module 125 stores the image ID of the image in which the child with that child ID is pictured, and the image ID of the image input to the child extraction module 125 when that child ID was output.

[0043] Figure 9 is an explanatory diagram showing an example of input to the language model M related to the overall daily childcare record generation process. The language model M is composed of, for example, a GPT (Generative Pretrained Transformer). The language model M includes an aspect as an LLM (Large Language Model) that produces output according to the instructions contained in the prompt based on the input prompt and data, and an aspect as a VLM (Vision Language Model) that produces output according to the instructions contained in the prompt based on the input prompt and image. The language model M may also be an MLLM (Multimodal Large Language Model).

[0044] In the overall daily childcare record generation process, the language model M receives images, childcare guidelines, and prompts as input. The images input to the language model M include images from the image database that correspond to a specific class ID and date. In the example shown in Figure 9, multiple images corresponding to a specific class ID and date are combined and input to the language model M as a single image, but each of the multiple images may also be input to the language model M.

[0045] The Childcare Guidelines for Nursery Schools, which are entered into Language Model M, are those issued by the Ministry of Health, Labour and Welfare that define the content of childcare in nursery schools and related operations. The Childcare Guidelines for Nursery Schools include descriptions of the content of childcare in nursery schools, such as the feelings, motivation, and attitudes that children are expected to acquire, and the tasks that childcare workers must perform, as well as guidelines for the content of childcare records and instructional plans. The Childcare Guidelines for Nursery Schools also include descriptions that apply to children of all ages and descriptions that apply to children of specific ages. In addition to or in lieu of the Childcare Guidelines for Nursery Schools, Language Model M may also be entered with childcare-related materials, guidelines, manuals, or daily records such as childcare notes created by childcare workers for each child, or accumulated and memorized childcare records.

[0046] In the overall daily childcare record generation process, the prompts input to the language model M consist of instructions, roles, reference information, the child's age, and rules. The prompt instructions shown in Figure 9 include instructions that, based on the input image, refer to the childcare guidelines according to the age of the child in the image, and output activity content, activity objectives, child's behavior, observations, childcare record, parent contact (comments for parents), lesson plan, and activity evaluation.

[0047] In the overall daily childcare record generation process, the prompts input to the language model M include the role (query) given to the language model M. In this example, the sentence, "You are a childcare and early childhood education professional. In the past, as a childcare worker and kindergarten teacher, you have supported the growth of many children while being conscious of 'praising and nurturing children'," is given to the language model M as its role.

[0048] In the overall daily childcare record generation process, the reference information of the prompts input to the language model M includes instructions to output by referring to the childcare guidelines provided to the language model M, and instructions to determine the section to refer to according to the age of the child. As mentioned above, the childcare guidelines include descriptions targeting children of specific ages. When the prompt is input, the language model M refers to the descriptions in the childcare guidelines that target the age of the child, as described later, and outputs the childcare record and instruction plan.

[0049] In the overall daily childcare record generation process, the child's age in the prompt input to language model M is the age of the child belonging to the class in which the image input to language model M was taken. The child's age is shown. Note that the child's age may also be determined based on the child shown in the image by referring to child DB122.

[0050] In the overall daily childcare record generation process, the prompt rules input to the language model M include detailed content for each of the following: activity content, activity objectives, child's behavior, observations, childcare record, parent communication, instruction plan, and activity evaluation, as well as the specification of the output format. In the example shown in Figure 9, the content of the activity content is "What kind of activities are the children doing?", and the format specification is "Output concisely in bullet points." The content of the activity objectives is "What were the intentions of the childcare worker in planning this activity?", and the format specification is "Output concisely in bullet points, broken down into elements of the childcare guidelines." The content of the child's behavior is "How are the children feeling about this activity?", and the format specification is "Output concisely in bullet points." The content of the observations is "Are there any points that the childcare worker should carefully observe as observations regarding the child's growth?", and the format specification is "Output concisely in bullet points, broken down into elements of the childcare guidelines."

[0051] The content of the childcare record is "a record of the day the images were taken, based on the activity content, the objectives of the activity, the children's behavior, and observations." The specified format is "the format of a childcare diary." The content of the parent communication (comments for parents) is "a description of the day's activities, focusing on the children's emotions," and the specified format is "to be written in a soft, informal style, excluding greetings." The content of the lesson plan is "considering the children's growth in the next step based on the intention of the day's activities, and outlining the intention and content of the activities at an appropriate level for the children's age," and the specified format is "to be written concisely in bullet points, broken down into elements of the childcare guidelines."

[0052] The activity evaluation will consist of "an evaluation of the five aspects of childcare in nurseries—health, human relations, environment, language, and expression—based on the activity content and objectives," and the specified format is "a 5-point scale from 1 to 5, where 1 is the least relevant to the aspect and 5 is the most relevant." The prompt may include commands to be output in structured text such as JSON notation, array notation, or Markdown notation.

[0053] Figure 10 is an explanatory diagram showing an example of the output of the language model M in the overall daily childcare record generation process. The language model M outputs activity content, activity objectives, child's condition, observations, childcare record, parent contact, next day's lesson plan, and activity evaluation (health-related activity evaluation, interpersonal relationship-related activity evaluation, environmental activity evaluation, language-related activity evaluation, and expression-related activity evaluation) in the specified format, according to the instructions and rules included in the input prompt. The content output by the language model M is stored in the overall daily table TAD of the overall childcare record DB123.

[0054] Figure 11 is an explanatory diagram showing an example of input to the language model M in the overall summary output processing. When growth information for a predetermined period is input from the growth information of each field, including childcare records and instructional plans stored in the overall childcare record DB123, the language model M outputs a summary of the predetermined period. For example, when growth information for one week (7 days) from the growth information recorded in the records of the overall daily table TAD is input to the language model M, the language model M outputs a summary of each item of the growth information for one week (childcare record, parent contact, activity content, activity objectives, child's condition, observations, instructional plan for next week, health activity evaluation, human relations activity evaluation, environment activity evaluation, language activity evaluation, expression activity evaluation), and the processing unit 11 of the information processing device 1 stores the output summary as growth information for that month and week in the overall weekly table TAW. When growth information for one month (4 or 5 weeks) from the records in the overall weekly table TAW is input to the language model M, the language model M outputs a summary of each item of the one month's worth of growth information, and the processing unit 11 of the information processing device 1 stores the output summary as the growth information for that month in the overall monthly table TAM. When growth information for one year (12 months) from the records in the overall monthly table TAM is input to the language model M, the language model M outputs a summary of each item of the one year's worth of growth information, and the processing unit 11 of the information processing device 1 stores the output summary as the growth information for that year in the overall annual table TAY.

[0055] In the overall summary output process, the language model M receives growth information for a predetermined period, childcare guidelines, and prompts recorded in the overall childcare record DB123. The following describes the instructions for inputting daily growth information and outputting weekly growth information as a summary. The language model M is given the contents of the records for 7 days in the overall daily table TAD as growth information. The childcare guidelines input in the overall summary output process are the same as those described above.

[0056] The prompts entered during the overall summary output process consist of instructions, roles, reference information, the child's age, and rules. The instructions included in the prompts entered during the overall summary output process include instructions to summarize and output each item of the input growth information (contents of multiple records in the overall daily table TAD) (e.g., "Summarize the growth information from the input overall daily table TAD and output a weekly summary of each item"). The roles, reference information, the child's age, and rules entered into the language model M during the overall summary output process are the same as those entered during the overall daily childcare record generation process.

[0057] Figure 12 is an explanatory diagram showing an example of the output of the language model M in the overall summary output process. The language model M outputs a weekly summary in the specified format for the activity content, activity objectives, children's behavior, observations, childcare records, parent communication, instructional plan, and activity evaluation (health-related activity evaluation, interpersonal relationship-related activity evaluation, environmental activity evaluation, language-related activity evaluation, and expression-related activity evaluation). The content output by the language model M is stored in the overall weekly table TAW of the overall childcare record DB123.

[0058] The output of the contents stored in the overall monthly table TAM and the overall annual table TAY by the language model M is performed by the same process as described in Figures 11 and 12. The explanations in Figures 11 and 12 describe the overall summary output process for outputting the overall monthly table TAM by replacing "day" and "daily" with "week" and "weekly," and "weekly" with "monthly." Similarly, the explanations in Figures 11 and 12 describe the overall summary output process for outputting the overall annual table TAY by replacing "day" and "daily" with "monthly" and "monthly," and "weekly" with "year" and "yearly." Note that when the language model M outputs the contents stored in the overall monthly table TAM or the overall annual table TAY, the contents stored in the overall daily table TAD (growth information) may also be input to the language model M.

[0059] Figure 13 is an explanatory diagram showing an example of input to the language model M related to the individual daily childcare record generation process. In the individual daily childcare record generation process, the language model M is input with an image taken of a specific child (specific image), the childcare guidelines for the nursery school, the overall childcare record for the day the specific image was taken (first childcare record), the childcare record for the specific child for the month prior to the day the specific image was taken (first growth information), and a prompt.

[0060] The specific image input to the language model M is an image taken of a specific child from among the images stored in the image database on a specific date, that is, an image from a record in the image database 121 where the child ID of the specific child is stored in the child-taking field. In the example shown in Figure 13, multiple images taken of a specific child on a specific date are arranged and combined and input to the language model M as a single image, but each of the multiple images may also be input to the language model M. The childcare guidelines for the nursery school that are input to the language model M in the individual daily childcare record generation process are the same as those described above.

[0061] The first childcare record entered into language model M includes the childcare record (draft childcare record) of the record in the overall daily table TAD relating to the day the specific image was taken. The first childcare record may also include the childcare record correction of the record in the overall daily table TAD relating to the day the specific image was taken.

[0062] The first growth information input to the language model M includes, for example, the childcare record (draft childcare record) for a specific child in the individual monthly table TIM for the month preceding the day the specific image was taken. The first growth information may also include proofread childcare records. Furthermore, the first growth information may include childcare records for months prior to the month preceding the day the specific image was taken, the week preceding the day the specific image was taken, the week prior to the week preceding the week preceding the week preceding the day the specific image was taken, the year preceding the day the specific image was taken, or the year prior to the year preceding the year preceding the day the specific image was taken. In addition, the first growth information may include information regarding parental communication or the child's development.

[0063] In the individual daily childcare record generation process, the prompts input to the language model M consist of instructions, roles, reference information, child information, and rules. The prompt instructions shown in Figure 9 include instructions to output the childcare record (second childcare record), parent contact, and child development for the day the specific image was taken, for the child who appears in all of the input specific images, based on the content of the input specific image, the first childcare record, and the first growth information.

[0064] In the process of generating individual daily childcare records, the prompts input to the language model M include the role (query) given to the language model M. In this example, the sentence, "You are a childcare and early childhood education professional. In the past, as a childcare worker and kindergarten teacher, you have supported the growth of many children while being mindful of 'praising and nurturing children'," is given to the language model M as its role.

[0065] In the individual daily childcare record generation process, the reference information of the prompt input to the language model M includes instructions to output by referring to the childcare guidelines provided to the language model M, and instructions to determine the section to refer to according to the child's age. As mentioned above, the childcare guidelines include descriptions targeting children of specific ages. When the prompt is input, the language model M refers to the descriptions in the childcare guidelines that target the child's age, as described later, and outputs the childcare record and instruction plan. The reference information also includes the childcare record (draft childcare record), activity content, and activity objectives of the record for the day the specific image was taken in the overall daily table TAD, and the childcare record (draft childcare record) of the record for the month preceding the day the specific image was taken for the specific child in the individual monthly table TIM.

[0066] In the individual daily childcare record generation process, the child information entered into the language model M includes the child ID, child name, and child age of the specific child captured in the image entered into the language model M. The child information includes information read from the child DB122.

[0067] The prompt rules input to the language model M during the individual daily childcare record generation process include detailed content for childcare records, parent / guardian communications, and the child's development, as well as the specification of the output format. Furthermore, the prompt rules input to the language model M during the individual daily childcare record generation process include a common rule to faithfully represent the child's appearance in all input images, while being mindful of changes from the first growth information (previous month's childcare record).

[0068] In the example shown in Figure 13, the content of the childcare record (second childcare record) is "characteristic observations unique to the child, excluding information about the activities," and the format is specified as "the format of a childcare diary." The content of the parent communication is "communication focusing on the child's unique emotions and observations, excluding greetings and activity details," and the format is specified as "output in a soft, informal style." The content of the child's development is "a record of the child's development based on the activities, activity objectives, and the first childcare record, covering five aspects: health, interpersonal relationships, environment, language, and expression," and the format is specified as "output for each of the five aspects." The prompt may also include commands to output in structured text such as JSON notation, array notation, or Markdown notation.

[0069] Figure 14 is an explanatory diagram showing an example of the output of the language model M in the individual daily childcare record generation process. The language model M outputs a report of second growth information, including childcare records (second childcare records), parent contact, and records of the child's development (health development, interpersonal development, environmental development, language development, and expression development), in the specified format, according to the instructions and rules contained in the input prompt. The report output by the language model M is stored in the individual daily table TID of the individual childcare record DB124. In other words, the language model M outputs a report containing second growth information for each child, and the report is stored for each child in the storage unit 12 of the information processing device 1. Note that for aspects of the child's development that cannot be evaluated from the image provided to the language model M, the output may be "None in particular". In the example shown in Figure 14, the "health" aspect is not evaluated, and "None in particular" is output.

[0070] Figure 15 is an explanatory diagram showing an example of input to the language model M in the personal summary output processing. When personal growth information for a predetermined period is input from personal growth information, including childcare records, stored in the personal childcare record DB124, the language model M outputs a summary of that predetermined period. For example, if personal growth information for one week (7 days) of a specific child recorded in the personal daily table TID is input to the language model M, the language model M outputs a summary of each item of the personal growth information for that week (childcare record, parent contact, health-related development, interpersonal relationship development, environmental development, language development, and expression development), and the processing unit 11 of the information processing device 1 stores the output summary as personal growth information for that month and week in the personal weekly table TIW. When personal growth information for a specific child recorded in the records of the individual weekly table TIW for one month (4 or 5 weeks) is input into the language model M, the language model M outputs a summary of each item of the individual growth information for that month for the specific child, and the processing unit 11 of the information processing device 1 stores the output summary as the individual growth information for that month for the specific child in the individual monthly table TIM. When personal growth information for a specific child recorded in the records of the individual monthly table TIM for one year (12 months) is input into the language model M, the language model M outputs a summary of each item of the individual growth information for that year, and the processing unit 11 of the information processing device 1 stores the output summary as the individual growth information for that year for the specific child in the individual annual table TIY.

[0071] In the individual summary output process, the language model M is inputted with the childcare records of a specific child for a predetermined period, as well as personal growth information including the child's development, childcare guidelines, and prompts, all recorded in the individual childcare record DB124. The following describes the instructions for inputting daily personal growth information and outputting weekly personal growth information as a summary. The language model M is given the contents stored in the records for 7 days of the individual daily table TID as personal growth information. The childcare guidelines input in the individual summary output process are the same as those described above.

[0072] The prompts entered during the individual summary output process consist of instructions, roles, reference information, child information, and rules. The instructions included in the prompts entered during the individual summary output process include instructions to summarize and output each item of the input individual growth information (contents of multiple records in the individual daily table TID) (e.g., "Summarize the individual growth information pertaining to a specific child in the input individual daily table TID and output a weekly summary of each item"). The roles, reference information, child information, and rules entered into the language model M during the individual summary output process are the same as those entered during the individual daily childcare record generation process.

[0073] Figure 16 is an explanatory diagram showing an example of the output of the language model M in the individual summary output processing. The language model M outputs a weekly summary in the specified format of childcare records, parent communication, and the child's development (health development, interpersonal development, environmental development, language development, and expression development) according to the instructions and rules included in the input prompt. The content output by the language model M is stored in the individual weekly table TIW in the individual childcare record DB124.

[0074] The output of the contents stored in the individual monthly table TIM and the individual annual table TIY by the language model M is performed by the same process as described in Figures 15 and 16. The explanations in Figures 15 and 16 describe the personal summary output process, which outputs the information stored in the individual monthly table TIM by replacing "day" and "daily" with "week" and "weekly," and "weekly" with "monthly." Similarly, the explanations in Figures 15 and 16 describe the personal summary output process, which outputs the contents stored in the individual annual table TIY by replacing "day" and "daily" with "monthly" and "monthly," and "weekly" with "year" and "yearly." When the language model M outputs the contents stored in the individual monthly table TIM or the individual annual table TIY, the contents stored in the individual daily table TID (personal growth information) may also be input to the language model M.

[0075] Figure 17 is an explanatory diagram showing an example of the overall childcare record correction screen. The terminal processing unit 21 of the terminal device 2 retrieves the childcare record draft and parent contact draft stored in the overall childcare record DB 123 from the information processing device 1 and displays them on the display unit 25. The terminal processing unit 21 also receives input for correction of the displayed childcare record or parent contact from the input unit 24 and transmits the input (corrected) childcare record (childcare record correction) or parent contact (parent contact correction) to the information processing device 1.

[0076] The overall childcare record proofreading screen displays the childcare record period, a childcare record draft display field, a childcare record proofreading field, a parent contact draft display field, a parent contact proofreading field, and a proofreading complete button. The childcare record draft display field displays the childcare record draft read from the overall childcare record DB123. The childcare record proofreading field displays the childcare record draft read from the overall childcare record DB123 in an editable format. The terminal processing unit 21 accepts input for proofreading the childcare record in the childcare record proofreading field. The parent contact draft display field displays the parent contact draft read from the overall childcare record DB123. The parent contact proofreading field displays the parent contact draft read from the overall childcare record DB123 in an editable format. The terminal processing unit 21 accepts input for proofreading the parent contact in the parent contact proofreading field. In Figures 17 and 18, the corrected parts are highlighted. When the proofreading complete button is pressed, the terminal processing unit 21 sends the corrected childcare record (childcare record proofreading) and the corrected parent contact (parent contact proofreading) to the information processing device 1. Furthermore, the terminal processing unit 21 may transmit the childcare record correction or parent contact correction to a terminal owned by the child's guardian, or it may transmit a request to the information processing unit 1 to transmit the childcare record correction or parent contact correction to a terminal owned by the child's guardian. In addition, the terminal processing unit 21 may transmit the childcare record correction or parent contact correction to a server device, application management device, or system management device, etc., different from the information processing unit 1, or it may transmit a request to the information processing unit 1 to transmit the childcare record correction or parent contact correction to a server device, application management device, or system management device, etc., different from the information processing unit 1.

[0077] In the example shown in Figure 17, the daily childcare records and parent contacts relating to the entire system, specifically the childcare record drafts stored in the childcare record draft field and the parent contact drafts stored in the parent contact draft field of the overall daily table TAD, are displayed. The date of the childcare record is also displayed. At this time, the childcare record corrections sent to the information processing device 1 are stored in the childcare record correction field of the overall daily table TAD, and the parent contact corrections are stored in the parent contact correction field of the overall daily table TAD. Furthermore, if the weekly childcare record drafts and parent contact drafts relating to the entire system are displayed, the childcare record corrections and parent contact corrections are stored in the overall weekly table TAW; if the monthly childcare record drafts and parent contact drafts relating to the entire system are displayed, the childcare record corrections and parent contact corrections are stored in the overall monthly table TAM; and if the annual childcare record drafts and parent contact drafts relating to the entire system are displayed, the childcare record corrections and parent contact corrections are stored in the overall annual table TAY.

[0078] Figure 18 is an explanatory diagram showing an example of the individual childcare record correction screen. The terminal processing unit 21 of the terminal device 2 retrieves the childcare record draft and parent contact draft stored in the information processing device 1 or the individual childcare record DB 124 and displays them on the display unit 25. The terminal processing unit 21 also receives input for correction of the displayed childcare record or parent contact from the input unit 24 and transmits the input (corrected) childcare record (childcare record correction) or parent contact (parent contact correction) to the information processing device 1.

[0079] The individual childcare record proofreading screen displays the childcare record period, child information (child ID and child's name), a childcare record draft display field, a childcare record proofreading field, a parent contact draft display field, a parent contact proofreading field, and a proofreading complete button. The childcare record draft display field displays the childcare record draft for a specific child, read from the individual childcare record DB124. The childcare record proofreading field displays the childcare record draft for a specific child, read from the individual childcare record DB124, in an editable format. The terminal processing unit 21 accepts input for childcare record proofreading in the childcare record proofreading field. The parent contact draft display field displays the parent contact draft for a specific child, read from the individual childcare record DB124. The parent contact proofreading field displays the parent contact draft for a specific child, read from the individual childcare record DB124, in an editable format. The terminal processing unit 21 accepts input for parent contact proofreading in the parent contact proofreading field. When the calibration complete button is pressed, the terminal processing unit 21 transmits the calibrated childcare record (calibrated childcare record) and the calibrated parent contact (calibrated parent contact) to the information processing unit 1. The terminal processing unit 21 may also transmit the calibrated childcare record or the calibrated parent contact to a terminal owned by the parent of the specific child, or it may transmit a request to the information processing unit 1 to transmit the calibrated childcare record or the calibrated parent contact to a terminal owned by the parent of the specific child.

[0080] In the example shown in Figure 18, the daily childcare record and parent contact information for a specific child, specifically the childcare record draft stored in the childcare record draft field of the individual daily table TID, and the parent contact draft stored in the parent contact draft field, are displayed. The date of the childcare record and the child ID and name of the specific child related to the displayed childcare record are also displayed. At this time, the childcare record corrections transmitted to the information processing device 1 are stored in the childcare record correction field of the individual daily table TID for the relevant date and the record related to the specific child, and the parent contact corrections are stored in the parent contact correction field of the individual daily table TID for the relevant date and the record related to the specific child. If the weekly childcare record draft and parent contact draft for the entire group are displayed, the childcare record corrections and parent contact corrections are stored in the individual weekly table TIW; if the monthly childcare record draft and parent contact draft for the entire group are displayed, the childcare record corrections and parent contact corrections are stored in the individual monthly table TIM; and if the annual childcare record draft and parent contact draft for the entire group are displayed, the childcare record corrections and parent contact corrections are stored in the individual annual table TIY. In other words, the terminal processing unit 21 of terminal device 2 receives a request for correction of a specific child's childcare record (second growth information), and the processing unit 11 of information processing device 1 stores the received second growth information for each child in the individual childcare record DB 124.

[0081] Figure 19 is a flowchart illustrating an example of the child extraction process. The processing unit 11 of the information processing device 1 acquires the captured image from the camera 3 (S1). The processing unit 11 inputs the acquired image into the child extraction module 125 (S2) and outputs the child ID of the child in the image (S3). Based on the output child ID, the processing unit 11 refers to the child DB to identify the class ID of the class in which the image was taken (S4). The processing unit 11 stores the image, child ID, and class ID in the image DB 121 (S5) and terminates the process.

[0082] Figure 20 is a flowchart showing an example of the overall daily childcare record generation process. The processing unit 11 of the information processing device 1 executes the overall daily childcare record generation process at a predetermined time each day, for example. Alternatively, the processing unit 11 may execute the overall daily childcare record generation process when, for example, a command to create an overall childcare record is input from the terminal device 2.

[0083] The processing unit 11 of the information processing device 1 reads images of the class and date for which childcare records are created from the image DB 121 (S11). The processing unit 11 inputs the read images, the childcare guidelines, and prompts to the language model M (S12). The childcare guidelines input to the language model M may be stored in the storage unit 12 or read from a web page, etc. Also, the prompts input to the language model M in the overall daily childcare record generation process may be stored in the storage unit 12 in advance or may be input, for example, from the terminal device 2. The processing unit 11 acquires the activity content, activity objectives, child's condition, observations, childcare record, parent contact, instruction plan, and activity evaluation output by the language model M according to the input prompts (S13). The processing unit 11 stores the acquired activity content, activity objectives, child's condition, observations, childcare record, parent contact, instruction plan, and activity evaluation in the overall daily table TAD of the overall childcare record DB 123 (S14) and terminates the process.

[0084] Figure 21 is a flowchart showing an example of the overall summary output process. In the following, shorter periods are referred to as lower-order periods, and longer periods as higher-order periods. That is, weekly, monthly, and yearly periods are higher-order than daily, weekly, and monthly periods. For example, when the processing unit 11 of the information processing device 1 has completed registering records corresponding to one higher-order period in the lower-order childcare record table of the overall childcare record DB 123, it starts the overall summary processing and outputs the contents to be stored in the higher-order childcare record table to the language model M.

[0085] The processing unit 11 of the information processing device 1 reads childcare records and instructional plans (growth information) for a predetermined period from the lower-level childcare record table contained in the overall childcare record DB 123 (S21). The processing unit 11 inputs the read childcare records and instructional plans for the predetermined period, the childcare center guidelines, and prompts into the language model M (S22). The childcare center guidelines input into the language model M may be stored in the memory unit 12 or read from a web page, etc. Also, the prompts input into the language model M in the overall summary output process may be stored in the memory unit 12 in advance or may be input, for example, from the terminal device 2. The processing unit 11 obtains summaries of the activity content, activity objectives, children's behavior, observations, childcare records, parent contact, instructional plans, and activity evaluations for higher-level periods output by the language model M according to the input prompts (S23). The processing unit 11 stores the activity content, activity objectives, children's behavior, observations, childcare records, parent communication, instruction plan, and activity evaluation output by the language model M into a higher-level childcare record table included in the overall childcare record DB 123 (S24), and then terminates processing.

[0086] Figure 22 is a flowchart showing an example of the individual daily childcare record generation process. The processing unit 11 of the information processing device 1 executes the individual daily childcare record generation process at a predetermined time each day, for example. Alternatively, the processing unit 11 may execute the individual daily childcare record generation process when, for example, a command to create an individual childcare record is input from the terminal device 2.

[0087] The processing unit 11 of the information processing device 1 reads a specific image of a specific child from the image DB 121 on the date on which the individual childcare record is created (S31). The processing unit 11 reads the entire childcare record (first childcare record) for the day on which the individual childcare record (second childcare record) is created from the overall daily table TAD (S32). The processing unit 11 reads the childcare record (first growth information) for the record of the month preceding the day on which the specific image of the specific child was taken from the individual monthly table TIM (S33). The processing unit 11 inputs the read specific image, childcare guidelines, first childcare record, first growth information, and prompts into the language model M (S34). The childcare guidelines input into the language model M may be stored in the storage unit 12 or read from a web page or the like. The prompts input into the language model M during the individual daily childcare record generation process may be stored in the storage unit 12 beforehand or may be input, for example, from the terminal device 2. The processing unit 11 acquires the childcare record (second childcare record), parent contact, and child development record output by the language model M according to the input prompt (S35). The processing unit 11 stores the acquired childcare record (second childcare record), parent contact, and child development record in the individual daily table TID of the individual childcare record DB124 (S36), and then terminates processing.

[0088] Figure 23 is a flowchart illustrating an example of individual summary output processing. Hereafter, in terms of units representing periods, shorter periods are referred to as lower-order periods, and longer periods as higher-order periods. That is, weekly, monthly, and yearly periods are higher-order than daily, weekly, and monthly periods. For example, when the processing unit 11 of the information processing device 1 has completed registering records corresponding to one higher-order period in the lower-order childcare record table of the individual childcare record DB 124, it starts overall record processing and outputs the contents to be stored in the higher-order childcare record table to the language model M.

[0089] The processing unit 11 of the information processing device 1 reads childcare records, parent contacts, and records of the child's development for a predetermined period from the lower-level childcare record table contained in the individual childcare record DB 124 (S41). The processing unit 11 inputs the read childcare records, parent contacts, records of the child's development, the childcare center guidelines, and prompts to the language model M (S42). The childcare center guidelines input to the language model M may be stored in the storage unit 12 or read from a web page, etc. Also, the prompts input to the language model M in the individual summary output processing may be stored in the storage unit 12 in advance or may be input, for example, from the terminal device 2. The processing unit 11 obtains the summaries of the higher-level childcare records, parent contacts, and records of the child's development output by the language model M according to the input prompts (S43). The processing unit 11 stores the childcare records, parent contacts, and child development output by the language model M in a higher-level childcare record table included in the overall childcare record DB 123 (S44), and then terminates processing.

[0090] According to the configuration and processing of Embodiment 1, the language model M outputs childcare records and instructional plans based on images of the children taken and the childcare guidelines for the nursery school. This makes it possible to easily obtain the children's childcare records and instructional plans.

[0091] Figure 24 is a flowchart showing an example of the overall childcare record calibration process. The terminal processing unit 21 of terminal device 2 receives input specifying the period (day, month / week, month, or year) of the overall childcare record to be calibrated (S51). The terminal processing unit 21 of terminal device 2 transmits the specified period to the information processing device 1 (S52).

[0092] The processing unit 11 of the information processing device 1 reads the draft childcare record and draft parent contact for the specified period, received from the terminal device 2, from the table in the overall childcare record DB 123 that corresponds to the specified period (S53). The processing unit 11 transmits the read draft childcare record and draft parent contact to the terminal device 2 (S54).

[0093] The terminal processing unit 21 of terminal device 2 displays the overall childcare record correction screen, which shows the childcare record draft and parent contact draft received from the information processing device 1, on the display unit 25 (S55). The terminal processing unit 21 accepts input for the correction of the childcare record and parent contact (S56). The terminal processing unit 21 transmits the corrected childcare record and parent contact (childcare record correction and parent contact correction) to the information processing device 1 (S57) and terminates the process.

[0094] The processing unit 11 of the information processing device 1 stores the childcare record correction and parent contact correction received from the terminal device 2 in the childcare record draft field and parent contact draft field of the record from which the childcare record draft and parent contact draft were read from among the records in any of the tables of the overall childcare record DB 123 (S58), and then terminates processing.

[0095] Figure 25 is a flowchart showing an example of the individual childcare record calibration process. The terminal processing unit 21 of terminal device 2 accepts input specifying the period (day, month / week, month, or year) of the entire childcare record to be calibrated and the child's ID (S61). The terminal processing unit 21 of terminal device 2 transmits the specified period and the child's ID to the information processing device 1 (S62).

[0096] The processing unit 11 of the information processing device 1 reads the draft childcare record and the draft parent contact document related to the specified period and the child ID of the specified child, received from the terminal device 2, from the table in the individual childcare record DB 124 that corresponds to the specified period unit (S63). The processing unit 11 transmits the read draft childcare record and the draft parent contact document to the terminal device 2 (S64).

[0097] The terminal processing unit 21 of terminal device 2 displays the individual childcare record proofreading screen, which shows the childcare record draft and parent contact draft received from the information processing device 1, on the display unit 25 (S65). The terminal processing unit 21 accepts input for proofreading of the childcare record (second childcare record) and parent contact (S66). The terminal processing unit 21 transmits the proofread childcare record and parent contact (proofread childcare record and proofread parent contact) to the information processing device 1 (S67) and terminates the process.

[0098] The processing unit 11 of the information processing device 1 stores the childcare record correction and parent contact correction received from the terminal device 2 in the childcare record draft field and parent contact draft field of the record from which the childcare record draft and parent contact draft were read from one of the records in the individual childcare record DB 124 (S68), and then terminates processing.

[0099] (Embodiment 2) In the individual summary output processing, the items of growth information entered and the items of the output summary may differ. The language model M according to Embodiment 1 outputs childcare records and instructional plans (activity plans) when childcare records, etc., for a specific child over a predetermined period are entered. Embodiment 2 will be described below with reference to the drawings.

[0100] Figure 26 is an explanatory diagram showing an example of input to the language model M in the personal summary output processing according to Embodiment 2. Figure 24 describes an example in which childcare records of a specific child stored in the personal monthly table TIM are input and a summary is output. In the personal summary processing according to Embodiment 2, 12 months' worth of records stored in the overall monthly table TAM are input.

[0101] The instructions included in the prompts entered during the individual summary processing according to Embodiment 2 refer to the nursery school childcare guidelines according to the age of the specific child, and include instructions to output the focus for the specific child's final year, individual focus, and the development of childcare and the child's growth, based on the monthly activity content and activity objectives, as well as the childcare record and child development information for the specific child each month.

[0102] The rules included in the prompts entered during the individual summary processing according to Embodiment 2 include the focus for the final year, the focus for the individual, and the specification of the content and format of the development of childcare and the child's growth. The content of the focus for the final year is "What goals did the child have for the year, based on the content and objectives of the monthly activities?" and the format specification is "Organize and output the key points." The content of the focus for the individual is "Looking back on the year based on the monthly childcare records of a specific child and information on the child's growth, what was the most important point in the child's development?" and the format specification is "Output in a sentence of XX characters or less." The content of the development of childcare and the child's growth is "Based on the monthly childcare records of a specific child and information on the child's growth, what was the most remarkable point in the child's growth, and records of the child's growth that can be used as a reference for childcare workers in the following year?" and the format specification is "Divide into multiple episodes and output solutions for matters that may become problems in the following year."

[0103] Figure 27 is an explanatory diagram showing an example of the output of the language model M in the individual summary output processing. The language model M outputs a yearly summary in the specified format, including the final year's priorities, individual priorities, and the development of childcare and the children's growth, according to the instructions and rules contained in the input prompt. The content output by the language model M may be stored in the individual yearly table TIY of the individual childcare record DB124.

[0104] The embodiments disclosed herein should be considered in all respects as illustrative and not restrictive. The technical features described in each embodiment can be combined with each other, and the scope of the present invention is intended to include all modifications within the claims and scope equivalent to the claims. Furthermore, the independent and dependent claims described in the claims can be combined with each other in any combination, regardless of the form of reference. In addition, the claims use a multi-claim format in which claims refer to two or more other claims (multi-claim format), but are not limited to this. They may also be described using a multi-claim format in which at least one multi-claim refers to another multi-claim (multi-multi-claim format). [Explanation of Symbols]

[0105] 1: Information Processing Device 2: Terminal device 3: Imaging device 4: Language Model Server 11: Processing Section 12: Storage section 12a:Storage medium 13: Communications Department 21: Terminal Processing Unit 22: Storage section 22a:Storage medium 23: Communications Department 24: Input section 25:Display section 125: Preschooler Extraction Module DB121: Image DB122: Preschool children DB123: Overall childcare record DB124: Individual childcare record M: Language model N: Network P: Program Pa: App Program S: Information Processing System TAD: Overall Daily Table TAM: Overall Monthly Table TAW: Overall Weekly Table TAY: Overall Annual Table TID: Personal Daily Table TIM: Individual Monthly Table TIW: Individual Weekly Table TIY: Individual Annual Table

Claims

1. We retrieve images taken of the children. By providing the aforementioned images and the childcare guidelines to the language model, childcare records and instructional plans are output. Information processing methods.

2. By giving the language model instructions to refer to the childcare guidelines for nursery schools according to the age of the child captured in the aforementioned image, childcare records and instructional plans are output. The information processing method according to claim 1.

3. By providing the aforementioned image and the nursery school childcare guidelines to the language model, comments for the parents of the children will be output. The information processing method according to claim 1 or 2.

4. The system reads out childcare records and guidance plans, including information on the children's growth, generated over a specified period. By providing the retrieved growth information for a predetermined period and the nursery school childcare guidelines to the language model, a summary of the child's growth information for that period is output. The information processing method according to claim 1 or 2.

5. We obtained multiple images taken of multiple children, By providing the aforementioned multiple images and the nursery school childcare guidelines to the language model, the first childcare record and instruction plan are output. The information processing method according to claim 1 or 2.

6. First growth information is acquired, which includes a specific image taken of a specific child and childcare records output for a predetermined period of time for the said specific child. By providing the aforementioned specific image, the first growth information, and the first childcare record to the language model, the second growth information, including the second childcare record of the specific child, is output. The information processing method according to claim 5.

7. A report containing the aforementioned second growth information is output for each child. The aforementioned report is memorized for each child. The information processing method according to claim 6.

8. We accept the correction of the aforementioned second growth information. The calibrated growth information from the previous 2nd generation is stored for each child. The information processing method according to claim 6.

9. Based on the content shown in the aforementioned image, the language model is given prompts that include instructions to output childcare records and instructional plans that include items in accordance with the Childcare Guidelines, thereby outputting childcare records and instructional plans. The information processing method according to claim 1 or 2.

10. Based on the content of multiple specific images taken of a specific child, and growth information including the child's past childcare records that are common to all of the multiple specific images, the language model is given prompts that include instructions to output childcare records for the specific child that include items in accordance with the childcare guidelines for the nursery school, thereby outputting the childcare records for the specific child. The information processing method according to claim 1 or 2.

11. We retrieve images taken of the children. By providing the aforementioned images and the childcare guidelines to the language model, childcare records and instructional plans are output. A program that instructs a computer to perform a process.

12. We retrieve images taken of the children. By providing the aforementioned images and the childcare guidelines to the language model, childcare records and instructional plans are output. Processing section An information processing device equipped with the following features.

13. A camera and Information processing device and Equipped with, The aforementioned photographic device takes images of the children, The processing unit of the aforementioned information processing device is The image of the child taken by the aforementioned camera is acquired from the aforementioned camera, By providing the aforementioned images and the childcare guidelines to the language model, childcare records and instructional plans are output. Information processing system.

Images