A method for automatically generating a child AI three-level growth report

Abstract

The application relates to the technical field of children's intelligence education, and discloses a children's AI three-level growth report automatic generation method, which is applied to children's intelligent hardware, is executed by a terminal-side processor in cooperation with a security chip, and comprises the following steps: S1, a three-level architecture pre-fixing solidification stage; S2, a basic data verification level acquisition control stage; S3, a multi-dimension fusion analysis level calculation stage; S4, a report output and file solidification level control stage; and S5, a whole-process compliance verification and closed-loop control stage. Through the exclusive architecture design of three-level serial control, the application realizes the terminal-side local closed-loop operation of the whole process of children's growth report generation, avoids the leakage risk caused by uploading of children's private data to the cloud from the technical bottom layer, and realizes the complete and normal operation of core functions in a network interruption state, and the application is provided with the matching design of kernel-level network isolation, hardware-level permission control and whole-process minimum necessary acquisition.

Description

Technical Field

[0001] This invention relates to the field of children's intelligent education technology, and in particular to a method for automatically generating a three-level AI growth report for children. Background Technology

[0002] With the deep integration of artificial intelligence technology and digital education, children's smart hardware equipped with AI learning assistance and growth companion functions has become a core tool in children's family education and self-learning scenarios. Growth reports, as the core carrier for recording children's growth trajectory and providing feedback on their learning and development status, have become a core and essential function of children's smart education products. At the same time, the state has introduced strict laws, regulations and regulatory requirements for the protection of minors' personal information. The full-process security control of children's privacy data and the compliance of functional implementation have become the core prerequisites for the marketization of children's smart education products. Current growth report technologies in the industry cannot simultaneously meet the multiple requirements of intelligent analysis, protection of children's privacy and safety, and strong compliance with regulations in practical applications, making it difficult to adapt to the actual usage needs and policy regulatory requirements of children's growth scenarios. Summary of the Invention

[0003] The purpose of this invention is to provide a method for automatically generating a three-level AI-powered growth report for children, in order to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: A method for automatically generating a three-level AI-powered growth report for children, the method being applied to children's smart hardware and executed by an edge processor in conjunction with a security chip, includes the following steps: S1. Three-level architecture pre-consolidation stage: Build a dimension-layered three-level report generation architecture, complete the solidification of full-process permission rules and hardware-level root trust configuration, lock the scope of data flow and network access permissions, and establish a three-level process serial control mechanism. S2, Basic Data Verification Level Collection and Control Phase: Complete the offline collection and verification of children's growth-related behavioral sequences on the terminal side, construct continuous time-series behavioral sequences, and execute full lifecycle management of the collected data; S3, Multi-dimensional Fusion Analysis Level Calculation Stage: In the edge-side secure execution environment, the standardization processing of data in each dimension is completed, multi-dimensional dynamic weighted fusion is performed, a comprehensive growth index is generated, and growth trend and benchmarking analysis are completed. S4. In the report output and archive solidification management stage, growth reports are automatically generated in multi-level cycles, the hardware-level tamper-proof verification of the reports is completed, and the encrypted storage and authorized export of the reports and growth archives are performed. S5. In the full-process compliance verification and closed-loop control phase, the compliance verification of the entire report generation process is completed locally, compliance audit records are generated, and hardware-level traceless clearing of temporary data throughout the process is performed to complete the local closed loop of the entire process.

[0005] As a further improvement to this technical solution: S1 includes the following: completing root trust configuration within the device's security chip, solidifying the full-process encryption key, permission verification rules, and compliance rules for minors, building a hardware-level encryption engine and integrity verification engine, establishing a full-process hardware-level permission verification and exception circuit breaker mechanism, and completing the hardware-level direct connection between the security chip and the device processor and local storage; permanently disabling network access permissions for the entire process related to report generation at the device operating system kernel level, writing the permission rules into the operating system kernel configuration, with no cloud modification or remote control entry point; opening an independent secure execution environment at the device operating system kernel level, performing minimum necessary permission control on the report generation-related processes; solidifying the serial execution logic of the basic data verification level, multi-dimensional fusion analysis level, report output, and archive solidification level, setting pre-verification rules for process execution, and only starting the execution of the next level process after the previous level process has completed all verifications and passed, and directly triggering operation interception if any step fails verification, and exception prompts are only completed locally on the device, with no data reporting behavior.

[0006] As a further improvement to this technical solution, S2 includes the following: Before each collection action is initiated, a permission verification request is sent to the security chip. If the verification is successful, the calling permission of the corresponding collection unit is granted; if the verification fails, the collection action is directly intercepted. The device offline collects four types of numerical structured data necessary for growth report analysis, namely, behavioral data related to learning ability, behavioral data related to emotional state, behavioral data related to living habits, and behavioral data related to focus. Unnecessary redundant data such as facial images, complete voice content, and private conversations are not collected. In the local secure execution environment, the discretely collected behavioral data is sorted by timestamp to form a continuous and uninterrupted time-series behavioral sequence, and the marking information of abnormal interruption nodes is supplemented. The original behavioral sequence data is subject to full lifecycle management. The original data caching period is completely bound to the report generation period. After basic verification is completed, hardware-level encryption is performed through the security chip and stored in a local independent encrypted partition. Unnecessary original data is not persistently stored and there is no data transmission channel.

[0007] As a further improvement to this technical solution: S3 includes the following: For the behavioral sequence data in four dimensions—ability, emotion, habit, and focus—extreme value normalization standardization is performed respectively, mapping all dimension data to a unified numerical range of 0 to 1, eliminating the dimensional differences between different dimensions. The calculation formula for the standardization process is: , in the formula, For the first The first dimension The original collected values ​​of each time-series node, For the first The minimum value of each dimension within the current reporting period. For the first The maximum value of each dimension within the current reporting period. For the first The first dimension The output values ​​are standardized from each time-series node; based on the child's age, school stage, and learning scenario, corresponding dimension weight coefficients are matched locally to complete the dynamic weighted fusion of the four dimensions, generating a unified comprehensive growth index. The weight coefficients can only be adjusted after local hardware-level authorization, with no cloud-based modification access. The calculation formula for the weighted fusion is as follows: , in the formula, The final output is the comprehensive growth index. The standardized capability dimension feature sequence, This is the standardized sequence of emotion dimension features. The standardized habitual dimension feature sequence, The standardized attention dimension feature sequence, , , , These are the dynamic weight coefficients corresponding to the four dimensions, and the weight coefficients satisfy... Based on single-dimensional standardized data and a comprehensive growth index, time-series trend analysis, benchmark comparison analysis of the same age group, identification of growth weaknesses and marking of strengths are completed locally. All calculation processes are completed on the device side without any data transmission.

[0008] As a further improvement to this technical solution: S4 includes the following: Report generation is automatically triggered locally on a multi-level cycle (daily, weekly, monthly, semester). Based on multi-dimensional fusion analysis results, a complete report is generated, including data statistics, trend analysis, a summary of strengths, suggestions for improving weaknesses, and personalized growth plans. The report template is pre-stored locally on the device, and the entire generation process is executed offline without cloud resource loading. After the report is generated, a unique hash verification code is generated for each report via a security chip. The verification code is strongly bound to the report content, generation time, and device hardware serial number. The verification code is encrypted and stored together with the report. Any modification to the report content will result in a mismatch of the verification code. The calculation formula for generating the verification code is: , in the formula, To report a unique hash checksum, Released by the State Cryptography Administration Cryptographic hash algorithms The binary sequence of the report content. The standard timestamp generated for the report, This is the device's unique hardware serial number. The root key is embedded in the security chip; after hardware-level encryption of the report and corresponding verification code, it is stored in an independent encrypted partition on the device, without network access. Only after local hardware-level identity verification can the report be authorized for export. The exported report embeds the verification code, which can be used to complete the authenticity verification locally. There is no cloud transmission or storage throughout the process; according to the semester or academic year cycle, all multi-level reports, verification codes, and growth comprehensive index time series within the corresponding cycle are integrated locally into a standardized growth digital file. The file supports local authorized migration and local authorized destruction.

[0009] As a further improvement to this technical solution, S5 includes the following: After the report is generated and stored, a full-process compliance verification is performed locally to confirm that there is no excessive data collection, no data transmission, and no compliance anomalies. A compliance audit record is generated, encrypted and stored in a security chip, and not uploaded to the cloud. Temporary cached data during the report generation process is cleared, and hardware-level traceless clearing is performed, leaving no data residue and completing the full-process local closed loop.

[0010] As a further improvement to this technical solution: all steps of the method are stored in a computer-readable storage medium in the form of computer-executable instructions. When the computer-executable instructions are loaded and executed by the processor of the children's smart hardware, the corresponding step of automatically generating a three-level AI growth report for children is realized.

[0011] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention, through a dedicated architecture design with three-level serial control, achieves closed-loop operation of the entire process of generating children's growth reports locally on the client side. It avoids the risk of leakage caused by uploading children's privacy data to the cloud from the underlying technology. The entire process has no cloud dependency and no data transmission. The core functions can run normally even when the network is offline. At the same time, through the supporting design of kernel-level network isolation, hardware-level permission control, and minimum necessary data collection throughout the process, it fully meets the requirements of relevant laws and regulations on the protection of minors' personal information, and fundamentally solves the privacy leakage and compliance risk problems existing in the prior art.

[0012] 2. This invention employs a three-tiered execution logic with layered dimensions, designing an independent technical link specifically for children's growth report scenarios. Completely decoupled from other functional modules of the device, it ensures the entire growth report generation process is unbypassable and tamper-proof. Simultaneously, through a hardware-level tamper-proof verification mechanism, it constructs a continuous, traceable, and reliable standardized digital archive of children's growth, adapting to the recording and analysis needs throughout the entire growth cycle. While strictly protecting children's privacy and security, it meets the core needs of parents and children for recording growth trajectories and analyzing growth status, significantly enhancing the compliance and market competitiveness of children's smart education products. Furthermore, it forms a unique technical barrier, increasing the patent protection value and feasibility of rights protection.

[0013] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it according to the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. Specific embodiments of the present invention are given in detail below with reference to the accompanying drawings. Attached Figure Description

[0014] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1 This is a schematic diagram of the method structure for automatically generating a three-level AI-powered growth report for children. Detailed Implementation

[0015] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are for illustrative purposes only and are not intended to limit the scope of the invention. The invention is described more specifically in the following paragraphs by way of example with reference to the accompanying drawings. It should be noted that the drawings are in a very simplified form and use non-precise proportions, and are only used to facilitate and clarify the illustration of the embodiments of the present invention.

[0016] Please see Figure 1 In this embodiment of the invention, a method for automatically generating a three-level AI-powered growth report for children includes the following steps: S1. Three-level architecture pre-consolidation stage: Build a dimension-layered three-level report generation architecture, complete the solidification of full-process permission rules and hardware-level root trust configuration, lock the scope of data flow and network access permissions, and establish a three-level process serial control mechanism. S2, Basic Data Verification Level Collection and Control Phase: Complete the offline collection and verification of children's growth-related behavioral sequences on the terminal side, construct continuous time-series behavioral sequences, and execute full lifecycle management of the collected data; S3, Multi-dimensional Fusion Analysis Level Calculation Stage: In the edge-side secure execution environment, the standardization processing of data in each dimension is completed, multi-dimensional dynamic weighted fusion is performed, a comprehensive growth index is generated, and growth trend and benchmarking analysis are completed. S4. In the report output and archive solidification management stage, growth reports are automatically generated in multi-level cycles, the hardware-level tamper-proof verification of the reports is completed, and the encrypted storage and authorized export of the reports and growth archives are performed. S5, the full-process compliance verification and closed-loop control stage, completes the full-process compliance verification of report generation locally, generates compliance audit records, performs hardware-level traceless clearing of temporary data throughout the process, and completes the full-process local closed loop; Specific application scenarios and implementing entities: This method is applied to children's smart hardware, including children's learning machines, educational tablets, AI companion devices and other terminals. It is executed by the device's on-side processor in conjunction with the security chip, without the intervention of the cloud server, which meets the core technical requirements of local closed loop on the device side. The core content of each step and the function of the corresponding device: S1 Three-Tier Architecture Pre-Consolidation Phase: This step forms the underlying foundation of the entire method. Its core function is to build a layered three-tier report generation architecture, solidify permission rules and hardware root trust, and lock data flow boundaries. The core devices involved in this step include a security chip root trust configuration device, a kernel-level network isolation device, an independent secure execution environment construction device, and a three-tier process serial control device. This device solidifies the serial execution logic of the basic data verification level, the multi-dimensional fusion analysis level, and the report output and file solidification level. It sets a pre-condition rule that the next level process can only be started after the previous level process has passed the verification, ensuring that the three-tier process cannot be bypassed or tampered with. It is designed independently for the growth report scenario and is completely decoupled from other functional modules of the device, achieving strong control over the entire report generation process. The security chip root trust configuration device is used to build hardware-level trust roots and encryption verification capabilities. The kernel-level network isolation device is used to close network access permissions from the system bottom layer to prevent data from being transmitted outside. The independent secure execution environment construction device is used to create an isolated running space for the entire report generation process to avoid interference from external processes and data reading. S2 Basic Data Verification and Acquisition Control Phase: This step provides a trusted data foundation for report generation. Its core function is to complete the offline collection, verification, and time-series processing of children's growth behavior data, achieving full lifecycle control of the collected data. The core devices involved in this step include a pre-collection hardware-level permission verification device, a minimum necessary behavior sequence collection device, a behavior sequence continuous processing device, and a data lifecycle control device. This device is specifically designed with collection rules for growth report scenarios, collecting only four types of numerical structured data, strictly adhering to the minimum necessary collection principle, and completely distinguishing itself from the collection logic of other scenarios such as anti-interference and status recognition. This avoids the risk of leakage of children's privacy data from the source of collection, ensuring that the collected data is only used for growth report analysis. The pre-collection hardware-level permission verification device is used for pre-collection permission control of collection actions, the behavior sequence continuous processing device is used to construct continuous time-series behavior sequences, and the data lifecycle control device is used to achieve full-process control of collected data from caching to encrypted storage. S3 Multi-Dimensional Fusion Analysis Calculation Stage: This step is the core analysis stage for report generation. Its core function is to complete the standardization, weighted fusion, and trend analysis of multi-dimensional growth data, generating a comprehensive growth index and providing core analysis results for the report content. The core devices involved in this step include a single-dimensional data standardization processing device, a multi-dimensional dynamic weighted fusion device, and a growth trend and benchmarking analysis device. The two devices are equipped with dedicated standardized calculation formulas and dynamic weighted fusion calculation formulas, respectively, designed specifically for multi-dimensional data analysis in growth reports. This achieves unified quantification and fusion of data from different dimensions, clearly distinguishing it from feature fusion algorithms in other scenarios. The growth trend and benchmarking analysis device is used to complete the trend judgment, benchmarking comparison, and strength and weakness identification of growth data. S4 Report Output and Archive Solidification Management Phase: This step is the core implementation stage of the method. Its core function is to automatically generate, verify, tamper-proof, store, and authorize the export of multi-level periodic growth reports, thus constructing standardized digital growth archives. The core devices involved in this step include a multi-level periodic report automatic generation device, a hardware-level tamper-proof verification device, a local encrypted storage and authorized export device, and a standardized digital growth archive solidification device. This device is equipped with a hash verification code generation formula based on the national cryptographic SM3 algorithm, realizing full-content tamper-proof verification of growth reports and archives, ensuring the credibility and traceability of growth records from the technical level, and is specifically designed for children's digital growth archive scenarios. The multi-level periodic report automatic generation device is used to automatically generate reports according to preset cycles, the local encrypted storage and authorized export device is used for encrypted storage and compliant authorized access of reports, and the standardized digital growth archive solidification device is used to complete the integration and solidification of long-term growth archives. S5 Full-Process Compliance Verification and Closed-Loop Control Phase: This step is the compliance final stage of the method. Its core function is to complete full-process compliance verification, audit record retention, and temporary data clearing, achieving a local closed loop for the entire process. The core devices involved in this step include a full-process compliance verification device and a temporary data hardware-level clearing device. The full-process compliance verification device is used to complete the compliance verification of the entire process operation and generate audit records. The temporary data hardware-level clearing device is used to complete the traceless clearing of temporary cached data, ensuring that no data remains.

[0017] S1 includes the following: Root trust configuration is completed within the device's security chip; full-process encryption keys, permission verification rules, and compliance rules for minors are solidified; a hardware-level encryption engine and integrity verification engine are built; a full-process hardware-level permission verification and exception circuit breaker mechanism is established; and a direct hardware-level connection is completed between the security chip and the device processor and local storage. Network access permissions for all report generation processes are permanently disabled at the device operating system kernel level; permission rules are written into the operating system kernel configuration, eliminating cloud modification and remote control access points. An independent secure execution environment is created at the device operating system kernel level to implement minimum necessary permission control over report generation processes. The serial execution logic of basic data verification, multi-dimensional fusion analysis, report output, and file solidification levels is solidified; pre-process verification rules are set; the next level of process can only be started after the previous level has completed all verifications and passed; failure at any stage directly triggers operation interception; exception prompts are only completed locally on the device, with no data reporting behavior. Specifically, the security chip's root trust configuration: This section clarifies the configuration method of the security chip, embedding encryption keys, permission verification rules, and compliance rules for minors within the security chip. It establishes a hardware-level encryption engine and integrity verification engine, creating a full-process hardware-level permission verification and anomaly circuit breaker mechanism, and completing the direct hardware-level connection between the security chip and the device processor and local storage. The hardware-level encryption engine is used for national-level encryption and decryption operations on all data throughout the process; the integrity verification engine verifies the integrity of rule files and program files to prevent malicious tampering; and the anomaly circuit breaker mechanism terminates the corresponding process at the hardware level when verification fails, providing the ultimate hardware guarantee for a forced local closed loop. The direct connection design between the security chip and hardware units ensures that there is no possibility of bypassing verification and encryption operations. Kernel-level network isolation: This section clarifies the implementation method of network isolation. At the device operating system kernel level, network access permissions for all processes related to the entire report generation process are permanently disabled. The permission rules are written into the operating system kernel configuration, with no cloud modification or remote control entry point. The data transmission channel is completely closed from the system bottom layer, and cannot be modified or closed by the application layer, users, or the cloud. There are no operable switches, which fundamentally eliminates the possibility of data being uploaded to the cloud. Independent Secure Execution Environment: This section clarifies how to build a secure execution environment. An independent secure execution environment is created at the device operating system kernel level, and minimum necessary permission control is performed on the report generation-related processes. This environment is completely isolated from other application processes on the device, and only the relevant processes of this method can access it, ensuring that the report generation execution process is not interfered with or read by other applications, and eliminating the risk of unauthorized access to data on the local machine. Three-level serial process control: This section details the implementation logic of the three-level serial process control device, solidifying the serial execution logic of the basic data verification level, multi-dimensional fusion analysis level, and report output and archive solidification level. It sets pre-verification rules for process execution, ensuring that the next level of process can only be started after the previous level has completed all verifications and passed. Failure at any stage directly triggers an operation interception. Anomaly alerts are only processed locally on the device, with no data reporting. This ensures that report generation must follow the logical sequence of data collection and verification, multi-dimensional analysis, and report output solidification, without skipping or bypassing steps. This guarantees the standardization and data reliability of the entire report generation process. The fact that anomaly alerts are only processed locally with no data reporting further strengthens privacy compliance.

[0018] S2 includes the following: Before each data collection action is initiated, a permission verification request is sent to the security chip. If the verification is successful, the corresponding data collection unit's access permission is granted; if the verification fails, the data collection action is directly intercepted. The device collects four types of numerical structured data necessary for growth report analysis offline: learning ability-related behavioral data, emotional state-related behavioral data, lifestyle habit-related behavioral data, and focus-related behavioral data. Unnecessary redundant data such as facial images, complete voice content, and private conversations are not collected. In the local secure execution environment, the discretely collected behavioral data is sorted by timestamp to form a continuous, uninterrupted temporal behavioral sequence, and the marking information of abnormal interruption nodes is completed. The original behavioral sequence data is subject to full lifecycle management. The original data caching period is completely bound to the report generation period. After basic verification is completed, hardware-level encryption is performed through the security chip and the data is stored in a local independent encrypted partition. Unnecessary original data is not persistently stored on the ground, and there is no data transmission channel. Specifically, pre-collection permission verification: This section clarifies the pre-collection control rules. Before each collection action is initiated, a permission verification request is sent to the security chip. If the verification is successful, the corresponding collection unit's access permission is granted; if the verification fails, the collection action is directly intercepted. This sets up a hardware-level pre-gate for data collection, locking in the legality of the collection behavior from the source, avoiding unauthorized collection behavior, and eliminating the possibility of collecting beyond the scope. Minimal Necessary Data Collection: This section details the implementation rules for the minimum necessary behavioral sequence collection device, clarifying that four types of numerical structured data necessary for growth report analysis must be collected offline on the device: learning ability-related behavioral data, emotional state-related behavioral data, lifestyle habit-related behavioral data, and attention-related behavioral data. Unnecessary redundant data such as facial images, complete audio content, and private conversations are not collected. Strictly adhering to the minimum necessary collection principle in the "Regulations on the Protection of Children's Personal Information," only numerical data necessary for growth report analysis is collected, avoiding the risk of privacy leaks during the collection process. Furthermore, the entire collection process is executed offline without any network requests, clearly distinguishing it from the collection logic of other scenarios. Behavioral Sequence Continuous Processing: This section clarifies the construction method of time-series behavioral sequences. In a local secure execution environment, discretely collected behavioral data is sorted by timestamps to form a continuous, uninterrupted time-series behavioral sequence, supplementing the marking information of abnormal interruption nodes. This ensures the temporal continuity of growth behavioral data, providing a complete and continuous data source for subsequent growth trend analysis and avoiding distortion of analysis results due to data breakpoints. Data Lifecycle Management: This section clarifies the full-process management rules for collected data. It implements full lifecycle management for raw behavioral sequence data, with the raw data caching cycle fully bound to the report generation cycle. After basic verification, hardware-level encryption is performed via a security chip, and the data is stored in a local independent encrypted partition. Unnecessary raw data is not persistently stored, and there is no data transmission channel. This achieves full lifecycle management of collected data from generation to deletion. Raw data is only temporarily cached in secure memory, and unnecessary data is not stored after the task is completed, eliminating the risk of local leakage of raw collected data and fully complying with the requirements for protecting children's privacy data.

[0019] S3 includes the following: For behavioral sequence data across four dimensions—ability, emotion, habits, and focus—extreme value normalization is performed to standardize the data, mapping all dimensions to a uniform numerical range of 0 to 1, eliminating dimensional differences between dimensions. The standardization calculation formula is as follows: , in the formula, For the first The first dimension The original collected values ​​of each time-series node, For the first The minimum value of each dimension within the current reporting period. For the first The maximum value of each dimension within the current reporting period. For the first The first dimension The output values ​​are standardized from each time-series node; based on the child's age, school stage, and learning scenario, corresponding dimension weight coefficients are matched locally to complete the dynamic weighted fusion of the four dimensions, generating a unified comprehensive growth index. The weight coefficients can only be adjusted after local hardware-level authorization, with no cloud-based modification access. The calculation formula for the weighted fusion is as follows: , in the formula, The final output is the comprehensive growth index. The standardized capability dimension feature sequence, This is the standardized sequence of emotion dimension features. The standardized habitual dimension feature sequence, The standardized attention dimension feature sequence, , , , These are the dynamic weight coefficients corresponding to the four dimensions, and the weight coefficients satisfy... Based on single-dimensional standardized data and a comprehensive growth index, time-series trend analysis, benchmark comparison analysis of the same age group, identification of growth weaknesses and marking of strengths are completed locally. All calculation processes are completed on the device side without any data transmission. Specifically, single-dimensional data standardization processing: This section elaborates on the implementation logic and calculation formula of the single-dimensional data standardization processing device, and clarifies that extreme value normalization standardization processing is performed on behavioral sequence data of four dimensions: ability, emotion, habit, and focus, respectively, to map all dimension data to a unified numerical range of 0 to 1 and eliminate the dimensional differences of data of different dimensions. Detailed formula annotation: The calculation formula for standardization is as follows , in the formula, For the first The first dimension The original collected values ​​of each time-series node, For the first The minimum value of each dimension within the current reporting period. For the first The maximum value of each dimension within the current reporting period. For the first The first dimension The standardized output values ​​of each time-series node; Formula explanation: This formula is specifically designed for multi-dimensional data processing in growth reports, clearly distinguishing it from standardized formulas used in other scenarios. It performs full-cycle extreme value normalization on a dimension-by-dimensional basis, fully preserving the time-series change trend of a single dimension within the reporting period. At the same time, it eliminates the dimensional differences between different dimensions of data, allowing different types of growth data to have a unified calculation benchmark. This provides standardized input data for subsequent multi-dimensional weighted fusion, ensuring the accuracy and rationality of the fusion calculation results. Multi-dimensional Dynamic Weighted Fusion: This section details the implementation logic and calculation formulas of the multi-dimensional dynamic weighted fusion device. It clarifies that based on the child's age, school stage, and learning scenario, corresponding dimension weight coefficients are matched locally to complete the dynamic weighted fusion of the four dimensions, generating a unified comprehensive growth index. The weight coefficients can only be adjusted after local hardware-level authorization; there is no cloud-based modification option. Detailed formula annotation: The calculation formula for weighted fusion is as follows: , in the formula, The final output is the comprehensive growth index. The standardized capability dimension feature sequence, This is the standardized sequence of emotion dimension features. The standardized habitual dimension feature sequence, The standardized attention dimension feature sequence, , , , These are the dynamic weight coefficients corresponding to the four dimensions, and the weight coefficients satisfy... Formula Explanation: This formula is completely different from the fixed-weight feature fusion formulas used in other scenarios. It adopts a dynamic weight matching mechanism, which can adjust the weight ratio of different dimensions based on the child's age, school stage, and learning scenario. For example, in the subject learning scenario, the dimensions of ability and concentration have higher weights, while in the habit formation scenario, the dimensions of habit and emotion have higher weights. This achieves personalized growth analysis that fits different scenarios and different children. It is specifically designed for the growth report scenario and has no technical overlap. Through this formula, standardized data from four dimensions can be integrated into a unified comprehensive growth index, which intuitively reflects the child's overall growth status and provides core quantitative indicators for trend analysis and benchmarking in growth reports. Growth Trend and Benchmarking Analysis: This section clarifies the specific implementation method of growth analysis. Based on single-dimensional standardized data and a comprehensive growth index, time-series trend analysis, benchmark comparison analysis with peers of the same age group, identification of growth weaknesses, and marking of strengths are completed locally. All calculation processes are completed on the device side without any data transmission. All analysis and calculations required for the growth report are completed on the device side without cloud computing support. It can be executed normally even without an internet connection, while ensuring that no children's growth data is transmitted outside the device during the analysis process, fully complying with privacy compliance requirements.

[0020] S4 includes the following: Automatic local report generation on a daily, weekly, monthly, and semester-based schedule; generation of complete reports based on multi-dimensional fusion analysis results, including data statistics, trend analysis, strengths summaries, suggestions for improvement, and personalized growth plans. Report templates are pre-stored locally on the device, and the entire generation process is performed offline without cloud resource loading. After report generation, a unique hash checksum is generated for each report via a security chip. This checksum is strongly bound to the report content, generation time, and device hardware serial number. The checksum is encrypted and stored along with the report; any modification to the report content will result in a checksum mismatch. The formula for calculating the checksum is: , in the formula, To report a unique hash checksum, Released by the State Cryptography Administration Cryptographic hash algorithms The binary sequence of the report content. The standard timestamp generated for the report, This is the device's unique hardware serial number. The root key is embedded in the security chip; after hardware-level encryption of the report and corresponding verification code, it is stored in an independent encrypted partition on the device, without network access. It can only be authorized to export the report after passing the device's local hardware-level identity verification. The exported report embeds the verification code, which can be used to complete the authenticity verification locally. There is no cloud transmission or storage throughout the process; according to the semester or academic year cycle, all multi-level reports, verification codes, and growth comprehensive index time series within the corresponding cycle are integrated into a standardized growth digital file locally. The file supports local authorized migration and local authorized destruction. Specifically, the system features automatic generation of multi-level periodic reports: This section clarifies the implementation rules for report generation. Reports are automatically triggered locally on a daily, weekly, monthly, and semester-based basis. Based on multi-dimensional fusion analysis results, a complete report is generated, including data statistics, trend analysis, a summary of strengths, suggestions for improving weaknesses, and personalized growth plans. The report template is pre-stored locally on the device, and the entire generation process is executed offline without loading cloud resources. This achieves automated and multi-period coverage of growth reports, meeting the recording and analysis needs of children at different time dimensions during their growth. Furthermore, the entire report generation process is offline, without cloud resource dependence, and can operate normally even without internet access. Hardware-level tamper-proof verification: This section details the implementation logic and calculation formulas of the hardware-level tamper-proof verification device. It clarifies that after report generation, a unique hash checksum is generated for each report via a security chip. This checksum is strongly bound to the report content, generation time, and device hardware serial number. The checksum is encrypted and stored along with the report; any modification to the report content will result in a checksum mismatch. Detailed formula annotation: The calculation formula for generating the checksum is as follows: , in the formula, To report a unique hash checksum, Released by the State Cryptography Administration Cryptographic hash algorithms The binary sequence of the report content. The standard timestamp generated for the report, This is the device's unique hardware serial number. The root key is embedded in the security chip. Formula explanation: This formula uses the national commercial cryptography algorithm SM3, specifically designed for the credibility requirements of growth reports and digital archives. It is completely different from classification reasoning and strategy evaluation formulas used in other scenarios, with no technological overlap. The formula performs a joint hash operation on the report content, generation time, device serial number, and security chip root key. The generated checksum is unique, unforgeable, and tamper-proof. Any modification to the report content will result in a hash checksum mismatch, ensuring the immutability and traceability of growth reports and records from a fundamental technical perspective, providing hardware-level credibility protection for children's digital growth archives. Local Encrypted Storage and Authorized Export: This section clarifies the implementation rules for report storage and export. The report and its corresponding verification code are hardware-encrypted and stored in a separate encrypted partition on the device, with no network access. Authorization to export the report is only granted after successful local hardware-level identity verification. The exported report embeds a verification code, allowing for local authenticity verification. There is no cloud transmission or storage throughout the entire process. This approach secures the report from both storage and access perspectives. Hardware-level encryption ensures that report data cannot be illegally read, and local hardware-level identity verification ensures that only authorized users can access the report. The export process involves no cloud intermediaries, eliminating the risk of privacy leaks during report transmission. Standardized Digital Growth Profiles: This section clarifies the rules for constructing digital growth profiles. Based on semester and academic year cycles, all multi-level reports, check codes, and time-series data of comprehensive growth indices within the corresponding cycle are integrated locally into standardized digital growth profiles. These profiles support local authorized migration and destruction. The discrete periodic reports are integrated into complete, continuous, and tamper-proof digital growth profiles, conforming to common data standards in educational settings. Furthermore, the support for local authorized migration and destruction fully meets the compliance requirements for the transferability, forgetting, and destruction of children's privacy data.

[0021] S5 includes the following: After the report is generated and stored, the entire process of compliance verification is completed locally to confirm that there is no collection beyond the scope, no data transmission, and no compliance anomalies. A compliance audit record is generated and encrypted and stored in the security chip without being uploaded to the cloud. Temporary cached data during the report generation process is cleared, and hardware-level traceless clearing is performed, leaving no data residue and completing the entire process in a local closed loop. Specifically, the full-process compliance verification: This section clarifies the implementation rules for compliance verification. After the report is generated and stored, the full-process compliance verification is completed locally to confirm that there is no unauthorized data collection, no data leakage, and no compliance anomalies. A compliance audit record is generated and encrypted and stored in a secure chip, without being uploaded to the cloud. This ensures that the entire report generation process complies with the relevant requirements of the "Regulations on the Protection of Children's Personal Information" and the "Personal Information Protection Law." The compliance audit record is only encrypted and stored locally and is not uploaded to the cloud, which not only achieves traceability of compliance operations but also avoids additional privacy risks caused by uploading audit information. Temporary data hardware-level erasure: This section clarifies the implementation rules for temporary data erasure, clearing temporary cached data during the report generation process. It performs hardware-level, traceless erasure, leaving no data residue and completing a local closed loop for the entire process. Hardware-level data erasure is performed through a security chip to ensure that no temporary cached data remains, completely eliminating the risk of temporary data leakage and completing a local closed loop for the entire process from data collection to report archiving.

[0022] All steps of the method are stored in a computer-readable storage medium in the form of computer-executable instructions. When the computer-executable instructions are loaded and executed by the processor of the children's smart hardware, the corresponding steps of automatically generating the children's AI Level 3 growth report are realized. Specifically, the method clarifies that all steps are stored in a computer-readable storage medium in the form of computer-executable instructions. When the computer-executable instructions are loaded and executed by the processor of the children's smart hardware, the corresponding steps of automatically generating the children's AI three-level growth report are realized. The computer-readable storage medium is located in the local encrypted storage partition of the children's smart hardware, which can only be read and executed by the device's local processor. There is no cloud access point for reading or modification, which is consistent with the core requirement of the method's local closed loop.

[0023] The method of use and working principle of this invention are as follows: Usage: After the device is powered on, it automatically loads and starts a three-level hierarchical report generation architecture. This architecture locks the data flow boundaries and network access permissions of the entire report generation process from the technical level. During the child's use of the device, the device completes the collection, verification, and time-series processing of growth-related behavioral data in a local offline environment. It strictly follows the preset three-level serial execution logic, completing the entire process of basic data verification, multi-dimensional fusion analysis, report generation, and file solidification in sequence. There is no cloud intervention or data transmission throughout the process. The device automatically generates and encrypts the growth report according to the preset cycle. Parents can only view and export the corresponding growth report and digital file through the device's local hardware-level identity verification via a direct local connection. The entire process does not rely on a cloud server, and all core functions can be executed normally even when the network is disconnected.

[0024] Working Principle: The core execution framework is a three-tiered serial control architecture consisting of basic data verification, multi-dimensional fusion analysis, and report output / archive solidification. The underlying technology is a forced local closed loop on the device side. Through security chip and system kernel-level configuration, the entire data flow is locked at both hardware and system levels, permanently closing network access permissions for related processes to prevent the possibility of children's growth data being leaked. The device completes the minimum necessary collection and full lifecycle management of growth behavior data in a local offline environment. Through dedicated standardized processing and dynamic weighted fusion algorithms, it performs multi-dimensional growth data analysis and calculation in a local secure execution environment. Based on the analysis results, it automatically generates multi-level periodic growth reports. A unique and tamper-proof verification code is generated for each report using a national cryptographic hash algorithm. Hardware-level encrypted storage and authorization control of reports and growth archives are simultaneously completed. The entire process of collection, calculation, analysis, generation, and storage is completed independently on the device side without cloud server intervention. Furthermore, through full-process compliance verification and temporary data erasure mechanisms, a complete local compliance closed loop from data collection to archive archiving is achieved.

[0025] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any way. Those skilled in the art can readily implement the present invention based on the description and drawings above. However, any modifications, alterations, and variations made by those skilled in the art without departing from the scope of the present invention using the disclosed technical content are equivalent embodiments of the present invention. Furthermore, any modifications, alterations, and variations made to the above embodiments based on the essential technology of the present invention are still within the protection scope of the present invention.

Claims

1. A method for automatically generating a three-level AI-powered growth report for children, characterized in that, The method is applied to children's smart hardware and is executed by an edge processor in conjunction with a security chip, including the following steps: S1. Three-level architecture pre-consolidation stage: Build a dimension-layered three-level report generation architecture, complete the solidification of full-process permission rules and hardware-level root trust configuration, lock the scope of data flow and network access permissions, and establish a three-level process serial control mechanism. S2, Basic Data Verification Level Collection and Control Phase: Complete the offline collection and verification of children's growth-related behavioral sequences on the terminal side, construct continuous time-series behavioral sequences, and execute full lifecycle management of the collected data; S3, Multi-dimensional Fusion Analysis Level Calculation Stage: In the edge-side secure execution environment, the standardization processing of data in each dimension is completed, multi-dimensional dynamic weighted fusion is performed, a comprehensive growth index is generated, and growth trend and benchmarking analysis are completed. S4. In the report output and archive solidification management stage, growth reports are automatically generated in multi-level cycles, the hardware-level tamper-proof verification of the reports is completed, and the encrypted storage and authorized export of the reports and growth archives are performed. S5. In the full-process compliance verification and closed-loop control phase, the compliance verification of the entire report generation process is completed locally, compliance audit records are generated, and hardware-level traceless clearing of temporary data throughout the process is performed to complete the local closed loop of the entire process.

2. The method for automatically generating a three-level AI-powered growth report for children according to claim 1, characterized in that, The S1 includes the following: completing root trust configuration within the device security chip, solidifying the full-process encryption key, permission verification rules and compliance rules for minors, building a hardware-level encryption engine and integrity verification engine, establishing a full-process hardware-level permission verification and abnormal circuit breaker mechanism, and completing the hardware-level direct connection between the security chip and the device processor and local storage. At the device operating system kernel level, network access permissions for all processes related to report generation are permanently disabled, and permission rules are written into the operating system kernel configuration, eliminating cloud modification and remote control access. An independent secure execution environment is created at the device operating system kernel level to implement minimum necessary permission control over the report generation-related processes. The serial execution logic of basic data verification, multi-dimensional fusion analysis, report output, and file persistence is solidified, and pre-verification rules for process execution are set. The next level of process can only be started after the previous level process has completed all verifications and passed. If any verification fails, operation interception is triggered directly. Anomaly prompts are only completed locally on the device, and there is no data reporting behavior.

3. The method for automatically generating a three-level AI-powered growth report for children according to claim 1, characterized in that, The S2 includes the following: before each acquisition action is initiated, a permission verification request is sent to the security chip. If the verification is successful, the access permission of the corresponding acquisition unit is granted. If the verification fails, the acquisition action is directly intercepted. The system collects four types of numerical structured data necessary for growth report analysis offline: learning ability-related behavioral data, emotional state-related behavioral data, lifestyle habit-related behavioral data, and focus-related behavioral data. Unnecessary redundant data such as facial images, complete audio content, and private conversations are not collected. In a local secure execution environment, the discretely collected behavioral data is sorted by timestamp to form a continuous, uninterrupted temporal behavioral sequence, and the marking information for abnormal interruption nodes is supplemented. Full lifecycle management is implemented for the original behavioral sequence data, with the original data caching period completely bound to the report generation period. After basic verification, hardware-level encryption is performed via a security chip, and the data is stored in a local independent encrypted partition. Unnecessary original data is not persistently stored, and there is no data transmission channel.

4. The method for automatically generating a three-level AI-powered growth report for children according to claim 1, characterized in that, S3 includes the following: For behavioral sequence data across four dimensions—ability, emotion, habits, and focus—extreme value normalization was performed to map all data to a uniform numerical range of 0 to 1, eliminating dimensional differences between the different dimensions. The standardization calculation formula is as follows: , in the formula, For the first The first dimension The original collected values ​​of each time-series node, For the first The minimum value of each dimension within the current reporting period. For the first The maximum value of each dimension within the current reporting period. For the first The first dimension The output values ​​are standardized from each time-series node; based on the child's age, school stage, and learning scenario, corresponding dimension weight coefficients are matched locally to complete the dynamic weighted fusion of the four dimensions, generating a unified comprehensive growth index. The weight coefficients can only be adjusted after local hardware-level authorization, with no cloud-based modification access. The calculation formula for the weighted fusion is as follows: , in the formula, The final output is the comprehensive growth index. The standardized capability dimension feature sequence, This is the standardized sequence of emotion dimension features. The standardized habitual dimension feature sequence, The standardized attention dimension feature sequence, , , , These are the dynamic weight coefficients corresponding to the four dimensions, and the weight coefficients satisfy... Based on single-dimensional standardized data and a comprehensive growth index, time-series trend analysis, benchmark comparison analysis of the same age group, identification of growth weaknesses and marking of strengths are completed locally. All calculation processes are completed on the device side without any data transmission.

5. The method for automatically generating a three-level AI-powered growth report for children according to claim 1, characterized in that, S4 includes the following: Reports are automatically generated locally on a daily, weekly, monthly, and semester-based schedule. Based on multi-dimensional fusion analysis results, a complete report is generated, including data statistics, trend analysis, a summary of strengths, suggestions for improving weaknesses, and personalized growth plans. Report templates are pre-stored locally on the device, and the entire generation process is performed offline without loading cloud resources. After report generation, a unique hash checksum is generated for each report via a security chip. This checksum is strongly bound to the report content, generation time, and device hardware serial number. The checksum is encrypted and stored along with the report; any modification to the report content will result in a checksum mismatch. The formula for calculating the checksum is as follows: , in the formula, To report a unique hash checksum, Released by the State Cryptography Administration Cryptographic hash algorithms The binary sequence of the report content. The standard timestamp generated for the report, This is the device's unique hardware serial number. The root key is embedded in the security chip; after hardware-level encryption of the report and corresponding verification code, it is stored in an independent encrypted partition on the device, without network access. Only after local hardware-level identity verification can the report be authorized for export. The exported report embeds the verification code, which can be used to complete the authenticity verification locally. There is no cloud transmission or storage throughout the process; according to the semester or academic year cycle, all multi-level reports, verification codes, and growth comprehensive index time series within the corresponding cycle are integrated locally into a standardized growth digital file. The file supports local authorized migration and local authorized destruction.

6. The method for automatically generating a three-level AI-powered growth report for children according to claim 1, characterized in that, The S5 includes the following: After the report is generated and stored, complete the full-process compliance verification locally to confirm that there is no excessive collection, no data transmission, and no compliance anomalies. Generate a compliance audit record, encrypt and store it in a secure chip, and do not upload it to the cloud. Clear the temporary cached data during the report generation process, perform hardware-level traceless clearing, leave no data residue, and complete the full-process local closed loop.

7. The method for automatically generating a three-level AI-powered growth report for children according to claim 1, characterized in that, All steps of the method are stored in a computer-readable storage medium in the form of computer-executable instructions. When the computer-executable instructions are loaded and executed by the processor of the children's smart hardware, the corresponding step of automatically generating a three-level AI growth report for children is realized.

Images