Method and system for multi-dimensional behavior collection of teenagers and time-sharing unlocking of mobile terminal

Abstract

The invention belongs to the technical field of mobile terminal security management and control, and particularly relates to a teenager multi-dimensional behavior acquisition and mobile terminal time-sharing unlocking system which is composed of a teenager mobile terminal, a wearable / IoT device, a parent mobile terminal and a cloud platform, and a bidirectional encrypted data stream closed loop is formed. The invention provides a set of whole-process solution of'behavior acquisition-anti-cheating verification-intelligent decision-time-sharing unlocking 'by fusing IoT (Internet of Things) multi-device coordination, AI (Artificial Intelligence) behavior recognition and reinforcement learning adaptive control technology systems aiming at a teenager mobile terminal use scene. The method is suitable for guide and control scenes used by parents for teenager mobile terminals, further covers a comprehensive quality education scene of learning-exercise-housework collaborative cultivation of teenagers, helps to realize an education goal of'control-guidance ', and has a good market application prospect.

Description

Technical Field

[0001] This invention belongs to the field of mobile terminal security management technology, specifically relating to a method and system for collecting multi-dimensional behavior data of teenagers and unlocking mobile terminals in a time-sharing manner. Background Technology

[0002] Currently, typical products / solutions on the market for guiding and controlling parents' use of mobile devices by teenagers can be divided into three categories, as follows: Pure time management apps, such as iOS's "Screen Time," Android's "Digital Health," and the Xiaotiancai Parent App, operate on the core logic of parents setting a daily entertainment time limit and forcibly locking the screen once the limit is reached. The limitation of these products lies in their focus solely on controlling "duration," failing to link it to core behavioral values ​​for teenagers (such as academic performance or contributions to household chores). Their flaw stems from the lack of a "behavior-permission" linkage mechanism; control merely restricts time without providing positive guidance.

[0003] Task-based incentive products, such as Zuoyebang Parent Edition and Yuanfudao's management module, operate on the core logic of unlocking the device after completing a designated learning task (homework, answering questions). The limitations of this type of product are that it only covers the single dimension of "learning," and the unlocking granularity is "whole device / whole screen," making it unable to accurately control highly addictive applications (such as short videos and games). The root cause of its shortcomings lies in its single dimension and coarse unlocking granularity, limiting its applicability to various scenarios.

[0004] Exercise-based reward systems, such as Keep Kids and Xiaomi Sports Points System, operate on the core logic of earning points through exercise and redeeming those points for physical goods (toys, coupons). The limitations of these products lie in the disconnect between points and device usage permissions, and the lack of anti-cheating mechanisms (such as indoor step counting by shaking the phone or faking exercise tracks). The root cause of these flaws is the absence of a direct link between exercise and device permissions, resulting in a lack of anti-cheating capabilities.

[0005] On the other hand, existing technologies also have the following key drawbacks: 1. The device management permission activation process is cumbersome: it requires authorization of the mobile terminal with the help of a computer or other external devices, and in some scenarios, it also requires downloading special drivers. The operation steps are complicated and the operation threshold for parents is high. 2. The guidance of comprehensive quality is too narrow: focusing only on a single dimension (such as learning or sports) cannot cover the needs of cultivating "learning + sports + housework" in a coordinated manner; for example, after children finish their homework and unlock their mobile phones, they are prone to become addicted to games or short videos, neglecting the development of labor habits and healthy exercise, which directly violates the core requirement of "control is guidance"; 3. Insufficient precision in control: It only supports "unlocking the whole device / locking the whole screen", and cannot implement differentiated control for applications such as "short videos (high addiction), social media (medium addiction), and tools (low addiction)". For example, after unlocking, children can use game applications without restrictions, resulting in a serious disconnect between the control effect and expectations. 4. Lack of anti-cheating mechanisms: The system does not integrate AI recognition and multi-device verification capabilities, and there are vulnerabilities such as "taking photos to upload assignments, shaking the phone indoors to count steps, and repeatedly uploading historical data"; actual test data shows that about 12 out of 100 task submissions were forged data, indicating a high risk of control failure. 5. Poor real-time performance and fragmented user experience: Relying on manual uploads via local scheduled tasks or manual review by parents can easily lead to the problem of "task completion and permission unlocking not being synchronized" (e.g., the child completes the task at 10 am, but the phone is only unlocked at 3 pm); surveys show that 80% of teenagers reported that "after completing the task, they cannot use the terminal in time and are unwilling to participate in management anymore"; 6. Lack of adaptive closed-loop: The system lacks a dynamic task adjustment mechanism, requiring parents to frequently and manually modify goals (e.g., if the child continues to exceed the target, the goal still needs to be manually increased). Actual tests show that parents need to spend an average of 20 minutes per day adjusting the management rules, resulting in high management costs.

[0006] To address the problems and shortcomings of existing technologies, this invention, targeting the mobile terminal usage scenarios of teenagers, integrates IoT multi-device collaboration, AI behavior recognition, and reinforcement learning adaptive control technology systems to provide a complete solution encompassing "behavior collection - anti-cheating verification - intelligent decision-making - time-based unlocking." This solution is suitable for scenarios where parents guide and manage teenagers' use of mobile terminals, further covering the comprehensive quality education scenario of cultivating teenagers' "learning - sports - housework," and helping to achieve the educational goal of "control as guidance."

[0007] The information disclosed in this background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0008] The purpose of this invention is to provide a method and system for collecting multidimensional behavior data from teenagers and unlocking mobile terminals in a time-sharing manner, so as to solve the above-mentioned problems.

[0009] To achieve the above objectives, this invention provides a system for multi-dimensional behavior data collection and time-sharing unlocking of mobile terminals for teenagers. The system consists of four terminals: a teenager's mobile terminal, a wearable / IoT device, a parent's mobile terminal, and a cloud platform, forming a closed loop of two-way encrypted data flow. The core responsibilities of each terminal are as follows: The mobile terminal for teenagers (collection + execution): includes a data collection module, a local verification module, and an unlocking execution module. It is responsible for collecting real-time behavioral data in three categories: learning, exercise, and housework, and completing the initial local verification of the data. It is also responsible for receiving the duration token JWT (JSON Web Token) issued by the cloud platform and calling the system API to execute the application's time-sharing unlock. Wearable / IoT devices (auxiliary data collection + anti-cheating module): Includes auxiliary data collection modules (Bluetooth jump rope, smart bracelet, dual-frequency GNSS module, etc.) and SE signature modules, which are used to provide auxiliary data collection and verification basis for sports data (such as heart rate, number of jump ropes, outdoor trajectory) and enhance anti-cheating capabilities; Parental mobile terminal (configuration + monitoring terminal): includes a parameter configuration module and a behavior monitoring module, which supports parents to customize "dimensional weight, maximum daily unlock time, and application unlock priority", and view teenager's behavior reports and abnormal alerts (such as data forgery warnings) in real time. Cloud platform (decision and storage): Deploys cross-validation module (CVE), reinforcement learning model, and duration token generation module, responsible for behavioral data verification, dynamic adjustment of tasks for the next day, distribution of duration tokens (JWT), and hardware-level storage of sensitive data.

[0010] Core data flow path: Data collection by youth mobile terminals / IoT devices → Local encryption processing → Upload to the cloud → Cloud cross-validation → Cloud generation of duration tokens / adjustment of next day's tasks → Cloud distribution of JWTs to youth mobile terminals → Youth mobile terminals execute applications for time-sharing unlocking.

[0011] The method for the system of multi-dimensional behavior collection and time-sharing unlocking of mobile terminals for teenagers, applied to the present invention, includes the following steps: Step S0: Without external devices, the mobile terminal automatically activates remote management permissions through accessibility and wireless debugging; details are as follows: S01. Preparation: Obtain page element information through Android Device Monitor or accessibility components, manually open the [Wireless Debugging] interface, and use the Monitor or logs to view and record the IDs and locations of the following components: [Wireless Debugging] Status component ID and location; The ID and location of the [IP address and port] component; The ID and location of the component for "Pairing Devices Using Pairing Code"; Click the "Pair Device with Pairing Code" component. In the pop-up "Pair with Device" window, record the IDs and locations of the "WLAN Pairing Code" and "Pairing IP Address and Port". S02. Self-activate device management permissions, which includes the following steps: Enable accessibility permissions and start the background wireless debugging service; Accessibility features automatically redirect to the debugging mode interface. Use the text search function to find the "Wireless Debugging" related component. If not found, scroll down the screen until it is found, then click the "Wireless Debugging" component to enter its interface. If the feature is not enabled, it will automatically enable "Wireless Debugging". The accessibility feature captures the wireless debugging interface, reads the IP and port number of the connection debugging mode through the component ID and location, and caches this information for subsequent connection debugging. After successful caching, click "Pair Device with Pairing Code" to view the pairing information. Read the pairing code and the temporary pairing IP and port number using the component ID. The debug backend service initiates debug pairing using the pairing code and port until pairing is successful. After successful pairing, initiate a debugging connection using the wireless debugging IP and port until the connection is successful; After a successful connection, execute the device permission activation script to automatically activate device management permissions.

[0012] Step S10: IoT device binding and certificate verification; Process: The youth mobile terminal scans nearby authorized IoT devices (such as Bluetooth jump ropes and smart bracelets) via Bluetooth 5.2 to obtain the device's unique SN code; the SN code is uploaded to the cloud platform, and the cloud platform verifies whether the device is an authenticated device (matching the preset device whitelist).

[0013] Error handling: If verification fails (unauthorized device), a pop-up window will appear on the terminal prompting "Please bind an authenticated IoT device" and a notification will be pushed to the parent's terminal simultaneously.

[0014] Step S20: Multidimensional behavioral data collection, collecting behavioral data on three categories of adolescents: learning, sports, and housework, and calculating scores; details are as follows: S21. Learning behavior collection, including: Paper-based assignments: The paper-based assignments are photographed using a mobile device's camera, and the question information is extracted using OCR (Optical Character Recognition). An AI big data model automatically grades the objective questions, and the subjective questions are graded by connecting to the teacher's API. Online learning: The mobile device for teenagers has built-in lightweight learning functions (such as online quizzes and knowledge point exercises). After completing a complete task, entertainment time is accumulated and unlocked (parents can dynamically configure the "task-reward time" rules through the APP). Score Calculation: Learning Score L = Accuracy Rate > 0.75? Base Score Corresponding to Reward Time: 0 (Base score range 0-120 points, preset by parents); S22. Motion behavior data collection, including: Mobile device: Dual-frequency GNSS (GPS L1 / L5 + Beidou B1I / B2a, positioning accuracy ≤3 meters) collects outdoor movement trajectory, pedometer filters invalid steps such as "shaking the phone indoors", and triggers facial recognition to verify the user's identity at irregular intervals; Wearable devices: Smart bracelets synchronize exercise heart rate (60-120 beats / minute is considered acceptable) and exercise duration; Bluetooth jump ropes synchronize "number of consecutive jumps in a single session (≥10 times is valid) and total number of jumps"; Scoring calculation: Exercise score M = GNSS trajectory integrity × 0.3 + heart rate target rate × 0.2 + exercise frequency target rate × 0.5 (maximum score 100 points); S23. Collection of household chores behavior, including: Process: Shoot a "10-second video of the housework process + a photo of the result", use the YOLOv8 AI model to identify the type of housework (such as sweeping, washing dishes, with an accuracy of ≥92%) and the standardization of the actions; compare the result quality with the "standard cleaning effect library" (such as ≤5% stain residue after sweeping); Score calculation: Household chores score H = type matching degree × 0.3 + action standard score × 0.4 + result quality score × 0.3 (maximum score 100 points).

[0015] Step S30: Initial verification of the local rule engine; details are as follows: S31. Verification Dimensions: Data Completeness (e.g., learning data must include "homework photos + grading results", exercise data must include "track + heart rate"), Data Reasonableness (e.g., single rope skipping ≤ 300 times, housework time ≥ 5 minutes); S32. Error Handling: If the data is incomplete or unreasonable, a pop-up window will appear on the terminal to indicate the specific problem (e.g., "Please supplement the running track for 10 consecutive minutes"), and the data will not be uploaded to the cloud platform for the time being.

[0016] Step S40: Encrypt and upload data to the cloud platform; details are as follows: S41. Encryption Strategy: Adopts dual protection of "HTTPS transmission encryption + end-to-end SM4 algorithm encryption". Uploaded data includes "learning / exercise / housework score (L / M / H), device SN code, timestamp, and terminal hardware fingerprint". S42. Anti-tampering design: The hardware fingerprint is bound to the uploaded data. The cloud platform can verify whether the data comes from the target terminal through the fingerprint, preventing data forgery across devices.

[0017] Step S50: The cloud platform uses the Cross-Validation Module (CVE) for anomaly detection and outputs the confidence score; details are as follows: S51. Verification Dimensions: Positioning Consistency (error between mobile phone GNSS positioning and wristband positioning ≤ 10 meters), Behavioral Continuity (deviation between daily learning data and historical data ≤ 20%), Forgery Detection (AI detection of whether photos / videos have been tampered with or are historical data). S52. Confidence level determination: The verification will pass automatically if the confidence level is ≥0.75; if the confidence level is <0.75, a manual review by parents will be triggered (if the review is not completed within 12 hours, the verification will be considered as failing).

[0018] Step S60: Adaptively adjust the task objective through a reinforcement learning model to form an adaptive closed loop; Key parameters include: Status s: Adolescent's overall completion rate (score) sequence over the past 7 days; Action a: Increase the target for the next day (-10% / 0 / +10%). Reward r: r = 0.7 × daily score + 0.3 × (1 - |Δavg|) (Δavg is the average of the score differences over the past 3 days); Adjustment logic: When r≥0.8, the task target for the next day is increased by 10%; when 0.6≤r<0.8, the task target for the next day remains unchanged; when r<0.6, the task target for the next day is decreased by 10%, and the learning / exercise / household chores target (L0 / M0 / H0) for the next day is updated.

[0019] Step S70: Calculate the overall completion score based on the behavioral data that has passed the confidence test, and generate a duration token; Overall completion score calculation: score = wL × (L / L0) + wM × (M / M0) + wH × (H / H0); Among them, wL / wM / wH are dimension weights, with a value range of 0-1; the sum of wL / wM / wH is 1, with a default of 0.4:0.4:0.2, which parents can customize; the value range of L / M / H is 0-120 points; the value range of L0 / M0 / H0 is 50-100 points; Time token calculation: T=round(k×score×Tmax) (k is the age coefficient: 1.0 for 6-12 years old, 1.2 for 13-18 years old; Tmax is the maximum unlock time per day, ranging from 0 to 120 minutes, with a default of 60 minutes, which can be set by parents); if score<0.6, then T=0; if T>Tmax, then T=Tmax.

[0020] Step S80: The cloud platform issues a time token (JWT) to the youth's mobile terminal; The JWT structure includes: header (specifying the RS256 encryption algorithm), payload (containing duration token T, unlock period, device SN code, and daily validity identifier), and signature (encrypted using a cloud private key to prevent tampering). Security design: JWT is bound to the device's serial number and is valid only for the day of use to prevent reuse across devices or expiration.

[0021] Step S90: The youth mobile terminal calls the system API to perform time-sharing unlocking; Process: The mobile terminal for teenagers receives the JWT → verifies the validity of the signature → calls the Android ScreenTime / iOS Family Controls system API → writes the unlocking rules (such as "short videos are available from 18:00 to 19:00") to the hardware secure storage area (such as Android RPMB, iOS Keychain). Example: If T=60 minutes, you can configure "Short videos 18:00-19:00 (30 minutes), games 20:00-21:00 (30 minutes)", and automatically lock highly addictive apps during other time periods.

[0022] The hardware coordination relationships of the system of this invention are shown in Table 1 below: Table 1 Hardware Coordination Relationship Table

[0023] Compared with existing technologies, and based on actual test data (100 units × 30 days), the present invention has the following beneficial effects: (1) In view of the defects of existing technologies where local files can be tampered with by ROOT, the system of the present invention adopts TrustZone+RPMB hardware storage. There was not a single case of bypassing the control on 100 ROOT terminals, achieving 0 cracking records. Sensitive data is 100% tamper-proof. Parents do not need to worry about "children cracking the control", which can effectively improve their trust in the system.

[0024] (2) In view of the defects of existing technology that motion data can be backfilled and forged, the system of the present invention adopts SE hardware signature (IoT device). Only 3 out of 1000 cheating attempts are successful (due to temporary SE failure), and the cheating success rate is reduced from 12% to 0.3%. The behavioral data is real and reliable, providing an accurate basis for intelligent decision-making.

[0025] (3) In view of the defects of existing technologies, such as indoor shaking of mobile phone to scan steps and distortion of motion data, the present invention adopts dual-frequency GNSS + barometer outdoor verification. Only 45 misjudgments were made out of 1000 motion data, the misjudgment rate was reduced by 85%, and the real motion recognition rate was ≥95.5%, which is conducive to guiding children to participate in real outdoor sports and cultivating healthy living habits.

[0026] (4) In view of the shortcomings of the existing technology in terms of the completion of photo / video forgery tasks, the system of the present invention adopts TOF liveness detection + AI behavior recognition. 200 forgery attempts all failed, and only 40 errors occurred in 500 housework recognition attempts. The forgery success rate was 0%, and the behavior recognition rate was ≥92%. This is conducive to ensuring that "the person completes the task" and cultivating the self-management ability of teenagers.

[0027] (5) In view of the shortcomings of the existing technology that parents need to spend 20 minutes a day to adjust the task goals, the present invention adopts reinforcement learning adaptive task, which only requires 6 manual corrections in 30 days (originally 60 times), the goal adjustment accuracy rate is ≥90%, parental intervention is reduced by 80%, which can reduce the management cost for parents and improve the ease of use of the system.

[0028] (6) In view of the fact that the whole device unlocking of the existing technology can easily lead to the excessive use of highly addictive applications, the system of the present invention adopts application-level time-sharing unlocking, reducing the average daily game time from 90 minutes to 54 minutes, and reducing the game / short video usage time by 40%, thereby achieving precise control of high-risk applications and balancing entertainment needs with positive guidance. Attached Figure Description

[0029] Figure 1 This is the overall system architecture diagram of the present invention, showing the connection relationship of the four terminals (youth mobile terminal / IoT device / parent terminal / cloud platform). The encrypted data flow is marked with red arrows and the control command flow is marked with blue arrows, clearly indicating the core modules of each terminal (such as cloud CVE engine and terminal acquisition module). Figure 2 This is a flowchart of the self-activation device permission process, showing the process of a mobile terminal achieving self-activation. Figure 3 It is a flowchart for multi-dimensional behavior collection and anti-cheating, with steps S10-S90 as the core, and the "input data", "output data" and "abnormal branches" of each step are marked (such as jumping to parent review when the confidence level of S50 is <0.75). Figure 4 It is a hot-load / time-sharing unlock sequence diagram, showing the sequence of "cloud-based JWT distribution → terminal verification → writing to the secure area → API call → application unlocking", and marking the time consumed in each step (e.g., JWT verification ≤ 100ms). Figure 5 This is a schematic diagram of the duration token generation and adaptive algorithm. The left side shows the logic of score calculation (weight × dimension completion rate) → T generation; the right side shows the reinforcement learning closed loop (s→a→r→update target), with the parameter value range marked. Detailed Implementation

[0030] The technical solution of this invention patent will be clearly and completely described below. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this invention.

[0031] In the description of this invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship are based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention.

[0032] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0033] Example 1 Operating environment: Android system; Mobile device for teenagers: Xiaomi 13 (Android 14, Snapdragon 8 Gen2 chip, 4K+OIS camera, supports ARM TrustZone). IoT devices: Bluetooth jump rope (model SK-01, equipped with nRF52832 chip + PN7462 SE security element), smart bracelet (model BW-03, equipped with nRF52832 chip, supports heart rate / movement trajectory collection). Parent's mobile device: iPhone 15 (iOS 17, with the "Digital Education Helper for Parents" APP installed); Cloud platform: Alibaba Cloud ECS (4 cores, 8GB RAM, Ubuntu 22.04 operating system, MySQL 8.0.32 database, PyTorch 2.0.1 deep learning framework).

[0034] The input targets are shown in Table 2 below: Table 2 Input Target for Example 1

[0035] The key parameters are as follows: Dimension weights: wL=0.4, wM=0.4, wH=0.2; The age coefficient k = 1.2 (15 years old, which falls within the 13-18 age range). The maximum daily unlock time, Tmax, is 60 minutes. CVE cross-validation confidence threshold = 0.75.

[0036] The execution process is as follows: S10 (IoT Pairing): The Xiaomi 13 scans and pairs the SK-01 jump rope and BW-03 smart band via Bluetooth 5.2, and uploads the device serial number to the cloud platform for verification; the cloud platform confirms that the device is an authorized device and establishes an AES-128 encrypted communication link (takes approximately 500ms). S20 (Data Acquisition): Learning: Answered 9 math olympiad questions correctly (all objective questions were correct, and 36 / 40 points were scored on subjective questions), and answered 18 English questions correctly (90% accuracy rate). The calculated score was L=92 points (approximately 120 minutes). Exercise: 1.1km outdoor run (5m altitude), 550 rope skips (≥10 skips per session), combined with heart rate data from the wristband (average 85 beats / min, meeting the standard), the calculated M=91 points (time spent approximately 45 minutes). Household chores: Sweeping 35㎡ (cleanliness 95%), recording the process video and photos of the results, AI recognition judged it as valid household chores, calculated H=94.5 points (time taken about 20 minutes). S30 (Local Verification): The local rule engine verifies the integrity (including "homework photos + grading results", "exercise trajectory + heart rate", and "housework videos + photos") and reasonableness (e.g., if the number of jump ropes is 550 or less, it is considered reasonable if the preset reasonable threshold is 800). Verification passes (takes approximately 100ms). S40 (encrypted upload): Data is uploaded using HTTPS+SM4 encryption (including L / M / H scores, device SN code, timestamp, and hardware fingerprint), taking approximately 3 seconds in a 5G network environment; S50 (Cloud Verification): The CVE engine verifies the consistency of positioning (positioning error between mobile phone and wristband is 8 meters ≤ 10 meters) and the continuity of behavior (deviation between daily learning data and historical data is 15% ≤ 20%). The confidence level is calculated to be 0.92 ≥ 0.75, and the verification passes (takes about 2 seconds). S60 (Adaptive Task Adjustment): Retrieve the score sequence of the past 7 days [0.85, 0.88, 0.90, 0.89, 0.91, 0.93, 0.99], calculate the reward r = 0.7 × 0.99 + 0.3 × (1 - |0.02|) = 0.987 ≥ 0.8, and increase the task target for the next day by 10% (takes approximately 1 second). S70 (Time Token Generation): Calculate score≈0.99, T=round(1.2×0.99×60)=71→60 (because T≤Tmax) (takes about 50ms). S80 (JWT issuance): A JWT (including the unlocking time period 18:00-19:00 / 20:00-21:00, device SN code, and valid identifier for the day) is generated in the cloud and issued to Xiaomi 13 (takes about 1 second). S90 (Time-sharing unlock): Xiaomi 13 verifies the validity of the JWT signature, writes the unlock rules into the TrustZone secure zone, and calls the ScreenTime API to complete the unlock (total time is approximately 280ms).

[0037] The output is as follows: Unlock effect: Short video apps are available from 18:00 to 19:00 (30 minutes), game apps are available from 20:00 to 21:00 (30 minutes), and utility apps (such as calculators) are available all day; Performance metrics: Total unlocking time was approximately 280ms, with no service interruption (0ms interruption). Anti-cheating effectiveness: No data forgery incidents within 30 days, and CVE did not trigger manual parental review; Compliance: Parental authorization was obtained before data collection, which complies with the relevant requirements of the "Regulations on the Protection of Minors Online".

[0038] Example 2 The difference from Example 1 is that the operating environment is the iOS system; Teen device: iPhone 14 (iOS 17, A15 chip, relies on Keychain secure area to store sensitive data). IoT Pairing: Complete the pairing of IoT devices (such as smart bracelets that support HomeKit) through Apple HomeKit certification to ensure device compatibility; Unlock API: Call the "ApplicationActivity" interface under the iOS Family Controls framework (supported by iOS 17.0 and above) to achieve application-level time-sharing unlocking.

[0039] The actual test results are as follows: Unlocking time: ≤320ms (due to differences in the Keychain security zone verification process compared to Android RPMB, the time is slightly longer than that of Android). Anti-cheating effectiveness: Consistent with the Android system, the cheating rate for motion data is 0.3%, and the pass rate for task forgery is 0%. Adaptation conclusion: The core algorithms (such as reinforcement learning and score calculation) can be fully reused. Only the "secure storage module + system API call logic" needs to be adjusted, resulting in low adaptation costs (the development cycle is shortened by 40%).

[0040] In addition, the present invention also provides a hardware expansion solution, including: Smartwatch compatibility: Supports Apple Watch Series 8 and Huawei Kids Watch 5X, with the addition of a "Sleep Achievement" dimension (incorporating sleep achievement data into the calculation of household chores score H), expanding to a four-dimensional management system of "learning - exercise - household chores - sleep"; the sleep achievement standard is "daily sleep time ≥ 8 hours, deep sleep percentage ≥ 25%", and sleep data is collected through the watch's sensors.

[0041] Smart home collaboration: Connects with Ecovacs X2 robot vacuum cleaner and Siemens SJ636 dishwasher to automatically acquire household chores data such as "assisting with sweeping and setting the table" for teenagers, reducing manual photo taking; actual test data shows that the smart home data synchronization success rate is ≥98%.

[0042] Campus scenario adaptation: Integrate with DingTalk Campus Edition to obtain teenagers' "classroom focus (such as whether they frequently lose focus) and homework submission status", and incorporate it into their learning score L; realize "home-school collaborative management", and parents can view campus behavior data in real time.

[0043] This invention also provides an unlocking granularity expansion scheme, as shown in Table 3 below: Table 3 Unlock Granularity Expansion Scheme

[0044] The foregoing description of specific exemplary embodiments of the invention is for illustrative and explanatory purposes. These descriptions are not intended to limit the invention to the precise forms disclosed, and it will be apparent that many changes and variations can be made in accordance with the foregoing teachings. The exemplary embodiments were chosen and described in order to explain the specific principles of the invention and its practical application, thereby enabling those skilled in the art to implement and utilize various different exemplary embodiments of the invention, as well as various different choices and variations. The scope of the invention is intended to be defined by the claims and their equivalents.

Claims

1. A system for multi-dimensional behavior data collection and time-sharing unlocking of mobile terminals for teenagers, characterized in that, include: The youth mobile terminal includes a data collection module, a local verification module, and an unlocking execution module. It is responsible for collecting real-time behavioral data in three categories: learning, exercise, and housework, and completing the initial local verification of the data. The youth mobile terminal is also responsible for receiving the time token (JWT) issued by the cloud platform and calling the system API to execute application time-sharing unlocking. Wearable / IoT devices, including an auxiliary data acquisition module and an SE signature module, are used to provide auxiliary data acquisition and verification basis for sports data, and enhance anti-cheating capabilities; The parent mobile terminal includes a parameter configuration module and a behavior monitoring module, allowing parents to customize "dimensional weights, maximum daily unlock time, and application unlock priority", and view real-time reports on adolescent behavior and abnormal alerts. The cloud platform includes a cross-validation module (CVE), a reinforcement learning model, a duration token generation module, and a secure storage module. It is responsible for behavioral data verification, dynamic adjustment of tasks for the next day, distribution of duration tokens (JWT), and hardware-level storage of sensitive data.

2. A method applied to the system for multi-dimensional behavior collection and time-sharing unlocking of mobile terminals for adolescents as described in claim 1, characterized in that, Includes the following steps: S0. Activate remote management permissions for mobile terminals; S10, IoT device binding and certificate verification; S20. Multi-dimensional behavioral data collection: Collecting behavioral data on three categories of adolescents: learning, sports, and housework. S30. Initial verification is completed through the local rules engine; S40, encrypt and upload data to the cloud platform; The S50 cloud platform uses the cross-validation module CVE for anomaly detection. S60, reinforcement learning models adaptively adjust task objectives; S70: Calculate the overall completion score based on the behavioral data passed with confidence, and generate a duration token; S80 and cloud platform issue time tokens (JWT) to youth mobile terminals; S90, youth mobile terminals call system API to perform time-sharing unlocking.

3. The method for a system for multi-dimensional behavior collection and time-sharing unlocking of mobile terminals for adolescents, as described in claim 2, is characterized in that... Activating remote management permissions for mobile terminals in S0 includes the following steps: S01. Obtain page element information through Android Device Monitor or accessibility components, manually open the [Wireless Debugging] interface, and use the Monitor or logs to view and record the IDs and locations of the following components: [Wireless Debugging] Status component ID and location; The ID and location of the [IP address and port] component; The ID and location of the component for "Pairing Devices Using Pairing Code"; Click the "Pair Device with Pairing Code" component. In the pop-up "Pair with Device" window, record the IDs and locations of the "WLAN Pairing Code" and "Pairing IP Address and Port". S02. Self-activate device management permissions, which includes the following steps: Enable accessibility permissions and start the background wireless debugging service; Accessibility features automatically redirect to the debugging mode interface. Use the text search function to find the "Wireless Debugging" related component. If not found, scroll down the screen until it is found, then click the "Wireless Debugging" component to enter its interface. If the feature is not enabled, it will automatically enable "Wireless Debugging". The accessibility feature captures the wireless debugging interface, reads the IP and port number of the connection debugging mode through the component ID and location, and caches this information for subsequent connection debugging. After successful caching, click "Pair Device with Pairing Code" to view the pairing information. Read the pairing code and the temporary pairing IP and port number using the component ID. The debug backend service initiates debug pairing using the pairing code and port until pairing is successful. After successful pairing, initiate a debugging connection using the wireless debugging IP and port until the connection is successful; After a successful connection, execute the device permission activation script to automatically activate device management permissions.

4. The method for a system for multi-dimensional behavior collection and time-sharing unlocking of mobile terminals for adolescents, as described in claim 2, is characterized in that... The S10 process involves IoT device binding and certificate verification, specifically including the following steps: S11. The youth mobile terminal scans nearby authorized IoT devices via Bluetooth to obtain the device's unique SN code; S12. Upload the SN code to the cloud platform. The cloud platform verifies whether the device is an authenticated device. S13. If the verification fails, a pop-up window will appear on the terminal prompting "Please bind the authenticated IoT device" and a notification will be pushed to the parent's terminal simultaneously.

5. The method for a system for multi-dimensional behavior collection and time-sharing unlocking of mobile terminals for adolescents, as described in claim 2, is characterized in that... In S20, multidimensional behavioral data is collected and scores are calculated, specifically including the following steps: S21. Learning behavior collection, including: The paper assignments are photographed using a mobile device's camera, the question information is extracted using OCR, the objective questions are automatically graded by an AI model, and the subjective questions are graded by connecting to the teacher's API. The mobile device for teenagers has a built-in lightweight learning function, and entertainment unlock time is accumulated after completing online learning tasks. The formula for calculating the learning behavior score is: Learning Score L = Accuracy Rate > 0.75? The base score corresponding to the reward time is 0; the base score range is 0-120 points, preset by the parents. S22. Motion behavior data collection, including: The mobile app collects outdoor exercise tracks, the pedometer filters out invalid steps, and triggers facial recognition check-in to verify the user's identity at irregular intervals. Wearable devices simultaneously collect exercise heart rate, exercise duration, and the rate at which exercise goals are met; The formula for calculating the exercise behavior score is: Exercise score M = GNSS trajectory integrity × 0.3 + heart rate target achievement rate × 0.2 + exercise frequency target achievement rate × 0.5; where the maximum exercise score is 100 points. S23. Collection of household chores behavior, including: Shoot a "10-second video of the housework process + a photo of the result", and use the YOLOv8 AI model to identify the type of housework and the standardization of the actions; Compare the results with the "standard cleaning effect library" to evaluate the quality of the results; The formula for calculating the housework behavior score is: Housework score H = type matching degree × 0.3 + action standard score × 0.4 + result quality score × 0.3; where the maximum score for housework is 100 points.

6. The method for a system for multi-dimensional behavior collection and time-sharing unlocking of mobile terminals for adolescents, as described in claim 2, is characterized in that... In S30, the initial verification is completed through the local rules engine, specifically including the following steps: S31. Verify the integrity and reasonableness of the data; S32. If the data is incomplete or unreasonable, a pop-up window will appear on the terminal to indicate the specific problem, and the data will not be uploaded to the cloud platform for the time being.

7. The method for a system for multi-dimensional behavior collection and time-sharing unlocking of mobile terminals for adolescents, as described in claim 2, is characterized in that... The encrypted upload of data to the cloud platform in S40 includes the following steps: S41. It adopts a dual-protection encryption strategy of "HTTPS transmission encryption + end-to-end SM4 algorithm encryption". The uploaded data includes: learning / exercise / housework scores, device SN code, timestamp, and terminal hardware fingerprint. S42. Bind the hardware fingerprint to the uploaded data. The cloud platform verifies whether the data comes from the target terminal through the fingerprint, preventing cross-device data forgery.

8. The method for a system for multi-dimensional behavior collection and time-sharing unlocking of mobile terminals for adolescents, as described in claim 2, is characterized in that... The S50 cloud platform uses the Cross-Validation Module (CVE) for anomaly detection, specifically including the following steps: S51. Verify the consistency of location and the continuity of behavior, and perform forgery identification; S52. Determine the confidence level: Automatically pass when the confidence level is ≥0.75; trigger manual review by parents when the confidence level is <0.

75.

9. The method for a system for multi-dimensional behavior collection and time-sharing unlocking of mobile terminals for adolescents, as described in claim 2, is characterized in that... In S60, the task objective is adaptively adjusted through a reinforcement learning model, and the reward formula is: r = 0.7 × daily score + 0.3 × (1 - |Δavg|); Wherein, score represents the overall completion rate of teenagers in the past 7 days; Δavg is the average of the score differences in the past 3 days; The adjustment logic is as follows: When r ≥ 0.8, the target for the next day is increased by 10%. When 0.6 ≤ r < 0.8, the task objective remains unchanged the next day; When r < 0.6, the task objective for the next day is reduced by 10%, and the learning / exercise / household chores objective for the next day is updated.

10. The method for a system for multi-dimensional behavior collection and time-sharing unlocking of mobile terminals for adolescents, as described in claim 2, is characterized in that... The process of generating a duration token in S70 includes the following steps: S71. Calculate the overall completion rate using the following formula: score=wL×(L / L0)+wM×(M / M0)+wH×(H / H0); Among them, wL / wM / wH are dimension weights, with a default of 0.4:0.4:0.2, which parents can customize; S72. Calculate the duration token using the following formula: T = round(k × score × Tmax); Where k is the age coefficient: 1.0 for 6-12 years old and 1.2 for 13-18 years old; Tmax is the maximum daily unlock time, which is 60 minutes by default; If score < 0.6, then T = 0; if T > Tmax, then T = Tmax.

11. The method for a system for multi-dimensional behavior collection and time-sharing unlocking of mobile terminals for adolescents, as described in claim 2, is characterized in that... In S80, a duration token (JWT) is distributed to the youth mobile terminal via a cloud platform. The duration token (JWT) is bound to the device's serial number (SN) and is valid only for the day to prevent reuse across devices or expiration. The duration token structure includes a header, a payload, and a signature. The header specifies the RS256 encryption algorithm. The payload includes the duration token (T), the unlocking period, the device's serial number (SN), and a valid-for-the-day identifier. The signature is encrypted using a cloud-based private key to prevent tampering.

12. The method for a system for multi-dimensional behavior collection and time-sharing unlocking of mobile terminals for adolescents, as described in claim 2, is characterized in that... In the S90, youth mobile terminals call the system API to perform time-sharing unlocking. The specific process is as follows: The youth mobile terminal receives the JWT → verifies the signature validity → calls the Android ScreenTime / iOS FamilyControls system API → writes the unlock rules to the hardware secure storage area.

Images