Growth evaluation intervention system based on four-start system

Abstract

The invention relates to the technical field of growth evaluation, and discloses a growth evaluation intervention system based on a four-start system, which is based on the four-start system of psychological structure construction, physical storage and cultivation, habit system formation and core ability improvement. Six core modules are arranged: an index library establishment module constructs a dynamic library containing quantification and quality indexes and clearly updates a trigger mechanism; the data collection module summarizes multi-scene four-start data; the association model construction module outputs key data such as a four-dimensional development state through data processing and training verification; the intervention path planning module formulates an intervention plan and allocates implementation subjects; the post-intervention data acquisition module acquires and arranges post-intervention data according to a standard; and the iterative optimization module continuously updates the dynamic index library and the intervention path table based on the related evaluation data. Accurate adaptation and dynamic adjustment of growth evaluation and intervention are achieved, and intervention pertinence and implementation effectiveness are improved.

Description

Technical Field

[0001] This invention relates to the field of growth assessment technology, and more specifically, to a growth assessment intervention system based on a four-stage system. Background Technology

[0002] Growth assessment and intervention are key technologies that support the full-cycle development of individuals. They are widely used in areas such as student ability cultivation, adult behavior improvement, and rehabilitation scenario adaptation. Their core value lies in scientifically assessing and identifying developmental shortcomings, and formulating targeted intervention plans to promote multidimensional growth of individuals.

[0003] However, traditional growth assessment and intervention techniques often rely on statically set assessment indicators, making it difficult to adapt to the differentiated needs of students and adults, as well as different scenarios such as school and rehabilitation. Furthermore, the integration of quantitative and qualitative behavioral indicators is insufficient, resulting in limited accuracy of assessment results. Data collection suffers from fragmentation; a lack of unified collection standards across multiple scenarios and insufficient precision in collecting data related to core weaknesses affect the comprehensiveness of the assessment and analysis. Intervention pathways are often based on fixed templates, failing to fully integrate individual growth trajectories and self-healing potential. The cross-dimensional content allocation lacks scientific basis and a real-time feedback and adjustment mechanism, making it impossible to dynamically optimize intervention duration and frequency based on implementation progress, leading to insufficient targeting and effectiveness. Simultaneously, the assessment, intervention, and optimization processes lack a closed-loop design, making it difficult to iteratively update the indicator system and intervention plan based on actual results, further impacting the sustainability and adaptability of growth support.

[0004] Therefore, it is necessary to design a growth assessment and intervention system based on the four-stage system to solve the problems of static indicator system, fragmented data collection, lack of personalized and dynamic adjustment of intervention path, and lack of closed-loop iteration mechanism in traditional growth assessment and intervention techniques. Summary of the Invention

[0005] In view of this, the present invention proposes a growth assessment and intervention system based on the four-start system, which aims to solve the problems of static indicator system, fragmented data collection, lack of personalized and dynamic adjustment of intervention path, and lack of closed-loop iteration mechanism in traditional growth assessment and intervention technology.

[0006] This invention proposes a growth assessment and intervention system based on a four-stage framework, comprising: The indicator library creation module is configured to establish a dynamic indicator library covering the four-stage system of psychological structure construction, physical fitness reserve cultivation, habit system formation, and core competency enhancement. The dynamic indicator library includes quantitative indicators and qualitative behavioral indicators. It sets update trigger mechanisms for the beginning-of-semester assessment node, intervention implementation completion node, indicator system related policy update node, and self-healing potential verification node. It adds and deletes suitable indicator items and adjusts indicator recording rules according to preset rules, and outputs a standardized four-dimensional indicator system. The data collection module is configured to collect data on target subjects requiring behavioral intervention or ability improvement in multiple scenarios from students and adults based on the four-dimensional indicator system, and summarize the data to form a dataset. The association model building module is configured to receive the dataset, build an association model between the four-dimensional data and the individual's growth trajectory, preprocess the data according to the indicator definitions in the dynamic indicator library, perform feature annotation, association training and verification, and output data on four-dimensional development status, degree of collaborative adaptation, core capability shortcomings and self-healing potential. The intervention path planning module is configured to determine the intervention focus and cross-dimensional content based on the output data of the association model construction module, formulate short-term and long-term intervention plans, allocate implementation entities according to data sources, and form an intervention path table containing intervention projects, scenarios, and time limits. The post-intervention data collection module is configured to collect post-intervention data at fixed intervals, intervention completion nodes, and before the next round of evaluation, in accordance with the requirements of the intervention path table, and maintain the same data collection standards as the data collection module, and match and organize the data according to the dimensions of the indicator library to form a post-intervention dataset. The iterative optimization module is configured to receive the post-intervention dataset, extract indicator suitability assessment data and intervention effectiveness assessment data, and iteratively update the dynamic indicator library and intervention path table of the four-start system.

[0007] Furthermore, after the iterative optimization module iteratively updates the dynamic indicator library and intervention path table of the four-start system, it also includes: Based on the updated dynamic indicator library of the four-start system, the weight of the data collection dimensions of the data collection module was optimized, and differentiated data collection priorities were set for different target groups such as students and adults, as well as different scenarios such as campus and rehabilitation. Based on the intervention effectiveness assessment data, a dynamic adjustment threshold is configured for the short-term intervention plan. The real-time intervention data collected by the post-intervention data collection module is compared with the dynamic adjustment threshold, and the intervention path planning module is triggered to adjust the intervention project duration and implementation frequency. The iterative correlation model collaborative mapping parameters are synchronized to the data collection module to guide it in improving the accuracy of data collection related to core capability shortcomings. At the same time, the indicator adaptability assessment results are fed back to the implementation entity of the target object's scenario, providing specific guidance on indicator understanding and data recording.

[0008] Furthermore, when configuring dynamic adjustment thresholds for short-term intervention programs and triggering intervention pathway adjustments, the following are included: Based on the target group type and the core competency shortcomings, the basic adjustment thresholds for the short-term intervention plan are determined. These basic adjustment thresholds are set according to four dimensions: psychological structure construction, physical fitness reserve cultivation, habit system formation, and core competency enhancement. By combining the intervention response efficiency of similar target objects in historical intervention data, the basic adjustment threshold is calibrated to obtain the dynamic adjustment threshold range for the appropriate individual. The dynamic adjustment threshold range is divided into a lower threshold and an upper threshold. The intervention participation rate, behavioral change magnitude, and training execution achievement rate data collected in real time by the post-intervention data acquisition module are compared with the dynamically adjusted threshold range: When real-time data is below the lower threshold, the intervention path planning module is triggered to shorten the duration of a single intervention project and increase the frequency of implementation. When real-time data exceeds the upper limit threshold, the intervention path planning module is triggered to extend the interval period of intervention projects and optimize the cross-dimensional content allocation. When real-time data is between the lower and upper thresholds, keep the original intervention plan parameters unchanged and record the data fluctuation trend as the basis for the next round of threshold calibration.

[0009] Furthermore, the calibration of the basic adjustment threshold by combining the intervention response efficiency of similar target objects in historical intervention data includes: The target objects of the same type are classified in two ways: group type and core competency weakness dimension, and an intervention response database of the same type of objects is constructed. The intervention response database includes historical basic threshold, intervention cycle achievement rate, weakness improvement rate and cross-dimensional collaborative improvement data. Define intervention response efficiency evaluation indicators, including short-term intervention target achievement rate, data fluctuation stability, and dimensional adaptation fit, and calculate the comprehensive response efficiency value according to preset weights. Preset optimal response thresholds and acceptable response thresholds, and classify calibration levels based on overall response efficiency values: When the overall response efficiency value is greater than or equal to the optimal response threshold, the calibration coefficient is 1.2; When the overall response efficiency value is greater than or equal to the qualified response threshold and less than the optimal response threshold, the calibration coefficient is 1.0; When the overall response efficiency value is less than the qualified response threshold, the calibration coefficient is 0.8; The calibrated dynamic thresholds are calculated separately for each dimension of the four-start system: the lower threshold is equal to the basic lower threshold multiplied by the calibration coefficient, and the upper threshold is equal to the basic upper threshold multiplied by the calibration coefficient. If the sample size of historical data for the same target object is greater than or equal to the preset sample threshold, the weighted average method is used to correct the calibration coefficient. The weight factors are allocated according to the evaluation priority of short-term intervention target achievement rate, data fluctuation stability, and dimensional fit. If the sample size is less than the preset sample threshold, the system's preset dimension adaptation calibration benchmark value is called to supplement and correct it, and finally the dynamic adjustment threshold range of the adapted individual is determined.

[0010] Furthermore, the association model construction module preprocesses and annotates the data, including: The dataset is divided into four sub-datasets according to the four-start system dimensions: psychological structure construction, physical fitness reserve cultivation, habit system formation, and core competency enhancement. The preprocessing stage involves data filtering: a preset threshold for the proportion of missing values ​​is set, and sample data with a proportion of missing values ​​greater than the threshold are removed. Interpolation is used to correct outlier data. Perform data standardization: convert quantitative indicators into a uniform scale using the Z-score method, and convert qualitative behavioral indicators into computable data according to preset coding rules; In the feature annotation stage, according to the definition of the dynamic indicator library, dimension labels, scene labels, subject labels and priority labels are added to the preprocessed data: dimension labels correspond to four-dimensional classification, scene labels distinguish between campus, family, practice and rehabilitation scenarios, subject labels indicate the data source, and priority labels are divided into first-level labels, second-level labels and third-level labels according to the importance of the indicators. The data sources include behavioral records, training logs, self-evaluation reports, peer evaluation results, teacher evaluation forms, and professional assessment reports.

[0011] Furthermore, the priority label division rules include: A preset indicator importance assessment system is established, which includes three dimensions: dimensional influence, data sensitivity, and intervention relevance. Each dimension is assigned a preset calculation weight, and the weight allocation rule is that dimensional influence is greater than data sensitivity, which is greater than intervention relevance, and the total weight is 1. Set a primary threshold and a secondary threshold, where the primary threshold is greater than the secondary threshold. Priority labels are assigned based on the overall score. When the overall score of the importance of an indicator is greater than or equal to the first-level threshold, it is marked as a first-level label, corresponding to the core intervention indicator; When the overall score of the indicator's importance is greater than or equal to the secondary threshold but less than the primary threshold, it is labeled as a secondary label, corresponding to an important monitoring indicator; When the overall score of the indicator's importance is less than the second-level threshold, it is labeled as a third-level label, corresponding to the regular recorded indicator; Among them, cognitive calibration in the psychological structure construction dimension and sustainable growth performance indicators in the core competency enhancement dimension are marked as first-level labels by default, while daily routine and diet norms in the physical fitness reserve cultivation dimension and daily routine execution indicators in the habit system formation dimension are marked as second-level labels by default.

[0012] Furthermore, the association relationship training and verification of the association model construction module includes: A four-dimensional sub-model is constructed using the gradient boosting tree algorithm, which learns the association rules of indicators in the dimensions of psychological structure construction, physical fitness reserve cultivation, habit system formation, and core competency improvement. An inter-dimensional collaborative mapping layer is constructed using an attention mechanism to calculate the contribution coefficient of each dimension to the improvement of core capabilities and to establish an interconnection network. During training, a five-fold cross-validation method is used, with a preset error threshold. The dataset is divided into a training set and a validation set according to a preset ratio, and the model parameters are iteratively optimized until the prediction error of the validation set is less than the error threshold. The verification process involves calling upon historical intervention data sets, setting a pre-defined fit threshold, and comparing the fit between the model's output data on core weaknesses and self-healing potential with the actual intervention effects. When the fit is greater than or equal to the fit threshold, the model is deemed to have passed verification. When the fit is less than the fit threshold, return to the data preprocessing stage to re-optimize the feature annotation rules.

[0013] Furthermore, the intervention path planning module allocates implementation entities according to data sources, including: When the data comes from classroom status records and group activity participation data in a campus setting, the implementing entity is set as the teacher, and the collaborating entity is configured as the school psychological counselor. The teacher is responsible for the implementation of the intervention project and daily records, while the psychological counselor provides psychological technical support. When the data comes from the daily routine records and self-management behavior data of the family scene, the implementing entity is set as the parent and the collaborating entity is set as the class teacher. The parent takes the lead in the implementation of intervention in the family scene, and the class teacher regularly updates the school performance and provides guidance. When the data comes from social participation feedback and skills application performance data in practical scenarios, the implementing entity is set as the practical activity mentor, and the collaborating entity is configured as professionals in relevant fields. The mentor is responsible for on-site intervention and guidance, and provides skills adaptation suggestions to professionals in relevant fields. When the data comes from behavioral improvement records and symptom relief data in rehabilitation scenarios, the implementing entity is set as professional intervention personnel, and the collaborating entity is configured as rehabilitation institution technicians. Professional intervention personnel develop rehabilitation intervention plans, and technicians are responsible for assisting in the implementation of training. All implementing entities and collaborating entities intervene in the progress in real time through the system data interface, forming a closed-loop implementation mechanism of primary responsibility execution and collaborative support.

[0014] Furthermore, the qualitative behavioral indicators are converted into computable data according to preset coding rules, including: Qualitative behavioral indicators are divided into four categories of coding objects according to the four-start system dimensions: psychological tendency, physical performance, habit execution, and ability application. Each category of coding object corresponds to a preset behavioral description dictionary. A three-level coding hierarchy is established: first-level dimension code, second-level representation code, and third-level strength code; A fixed weight relationship is assigned to each level: the weight of the second-level performance code is greater than the weight of the third-level strength code, which is greater than the weight of the first-level dimension code. The formula for calculating the encoding value remains unchanged. The formula for calculating the encoding value is: the encoding value is equal to the sum of the first-level dimension code multiplied by the weight of the first-level dimension code, the second-level performance code multiplied by the weight of the second-level performance code, and the third-level strength code multiplied by the weight of the third-level strength code, multiplied by 10, and the result is rounded to the nearest integer. Establish a coding mapping table: psychological tendency class corresponds to first-level dimension code 1, physical performance class corresponds to first-level dimension code 2, habit execution class corresponds to first-level dimension code 3, and ability application class corresponds to first-level dimension code 4. Coding validity verification: A consistency threshold is preset, 10% of the qualitative data samples are randomly selected, and two professional evaluators independently code the data. The coding consistency coefficient is calculated, and the coding rule takes effect when the consistency coefficient is greater than or equal to the consistency threshold. If the consistency coefficient is less than the consistency threshold, the fuzzy description of the behavior description dictionary is revised, typical behavior examples are added, and the code is recoded until the consistency requirements are met. The revised coding rules are then stored in the dynamic indicator library as the basis for updates.

[0015] Furthermore, the evaluation method for the self-healing potential data output by the correlation model construction module includes: The self-healing potential assessment is determined in three dimensions: willingness to adjust autonomously, ability to repair through dimensional synergy, and environmental adaptability, which correspond to the core driving factors of the four-start system. Collect basic data on self-healing potential: including natural changes in indicators of target subjects during periods without intervention, records of proactive improvement behaviors, cross-scenario behavior transfer performance, and evaluation data on external environmental support; Calculate the self-healing potential score for each dimension, set the self-healing potential level threshold, which is divided into high potential threshold, medium potential threshold and low potential threshold, and the correlation model outputs targeted self-healing potential data according to the level; If there is no historical data for the target object during the intervention period, the baseline value of self-healing potential in the same dimension of the same group is used. The deviation rate between the target object's initial indicator data and the group mean is used for correction. The deviation rate is equal to the difference between the target object's initial value and the group mean divided by the group mean. The correction formula is: the corrected self-healing potential value is equal to the sum of the baseline value multiplied by 1 and the deviation rate multiplied by 0.5. Finally, the corrected self-healing potential level and data are output.

[0016] Furthermore, it also includes a data security and privacy protection module, which is configured to conduct full lifecycle security management of all data involving the personal information of the target object obtained by the data collection module and the post-intervention data acquisition module, and to encrypt the data during transmission using an asymmetric encryption algorithm.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. By using the multi-node update triggering mechanism of the dynamic indicator library and the functions of adding and deleting indicator items and adjusting recording rules, the quantitative indicators and qualitative behavioral indicators are integrated and adapted to accurately match the differentiated needs of different groups of students / adults and different scenarios of campus / rehabilitation, thus solving the problem of insufficient adaptability of the static indicator system; 2. Establish a unified data collection standard based on the standardized four-dimensional indicator system, combine the correlation model with the collaborative mapping parameters to enhance the accuracy of data collection for core capabilities that are lacking, and integrate data from multiple scenarios to form a complete dataset to solve the problems of lack of unified data collection standards and insufficient accuracy of core data. 3. Based on individual growth trajectory, four-dimensional development status and self-healing potential data, develop personalized intervention plans. Through a dynamic threshold adjustment mechanism, compare intervention data in real time and flexibly adjust intervention time limits, implementation frequency and cross-dimensional content ratio to solve the problems of templated intervention plans and delayed adjustments. 4. By extracting post-intervention evaluation data through the iterative optimization module, continuously updating the dynamic indicator library and intervention path table, and simultaneously optimizing data collection weights, model parameters, and implementation subject guidelines, a complete closed loop of "assessment-intervention-feedback-optimization" is formed, solving the problem of disconnect between assessment and intervention and the inability to continuously adapt to individual development. Attached Figure Description

[0018] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are configured only to illustrate preferred embodiments and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A functional block diagram of a growth assessment and intervention system based on a four-stage system provided in an embodiment of the present invention. Detailed Implementation

[0019] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] Reference Figure 1 In some embodiments of this application, a growth assessment and intervention system based on a four-stage system includes: an indicator library establishment module, a data collection module, an association model construction module, an intervention path planning module, a post-intervention data collection module, and an iterative optimization module.

[0021] The indicator library creation module is configured to establish a dynamic indicator library covering the four-stage system of psychological structure construction, physical fitness reserve cultivation, habit system formation, and core competency enhancement. The dynamic indicator library includes quantitative indicators and qualitative behavioral indicators. It sets update trigger mechanisms for the beginning-of-semester assessment node, intervention implementation completion node, indicator system related policy update node, and self-healing potential verification node. It adds and deletes suitable indicator items and adjusts indicator recording rules according to preset rules, and outputs a standardized four-dimensional indicator system. The data collection module is configured to collect data on students and adults who require behavioral intervention or ability improvement in multiple scenarios based on a four-dimensional indicator system, and then aggregate the data to form a dataset. The association model building module is configured to receive datasets, build an association model between four-dimensional development data and individual growth trajectory, preprocess the data according to the indicator definitions of the dynamic indicator library, perform feature annotation, association training and verification, and output data on four-dimensional development status, degree of collaborative adaptation, core capability shortcomings and self-healing potential. The intervention path planning module is configured to determine the intervention focus and cross-dimensional content based on the output data of the association model building module, formulate short-term and long-term intervention plans, allocate implementation entities according to data sources, and form an intervention path table containing intervention projects, scenarios, and time limits. The post-intervention data collection module is configured to collect post-intervention data at fixed intervals, intervention completion nodes, and before the next round of evaluation, in accordance with the requirements of the intervention path table, and maintain the same data collection standards as the data collection module. The data is then matched and organized according to the dimensions of the indicator library to form the post-intervention dataset. The iterative optimization module is configured to receive the post-intervention dataset, extract indicator suitability assessment data and intervention effectiveness assessment data, and iteratively update the dynamic indicator library and intervention path table of the four-start system.

[0022] Specifically, when the system detects that the current time has reached the preset start time of the semester-end assessment (e.g., the first week of each semester), or when the start instruction for the semester-end assessment is manually entered, the update mechanism for the semester-end assessment node is triggered; when the post-intervention data collection module completes the data collection of the intervention completion node and feeds back the intervention implementation end signal, and this signal matches the preset intervention cycle completion condition in the intervention path table, the update mechanism for the intervention implementation completion node is triggered; when the system obtains updated education and rehabilitation policy documents related to the four-start system indicators through an external policy interface, or when a verified policy adjustment notice is manually uploaded, the update mechanism for the indicator system associated policy update node is triggered; when the associated model construction module completes the self-healing potential verification calculation and outputs the "self-healing potential assessment completed" result, and this result meets the preset verification data integrity requirements, the update mechanism for the self-healing potential verification node is triggered.

[0023] Specifically, the addition and deletion of adaptive indicator entries and the adjustment of indicator recording rules are based on preset rules. The core principle is to formulate clear technical rules based on the core objectives of indicator adaptability, practicality, data quality, and accurate system assessment and effective intervention. When adding an indicator entry, if the update mechanism is triggered and it is found that the existing indicator cannot cover the core assessment dimensions, the correlation between the unincluded behavioral characteristics and the improvement of core capabilities reaches a preset threshold, or it needs to adapt to policy / new scenario requirements, then a corresponding quantitative / qualitative indicator is added and labeled accordingly. When deleting an indicator entry, if the usage rate or intervention correlation of an indicator is lower than the preset threshold for two consecutive periods, or if it becomes invalid due to policy adjustments or scenario changes, then it is deleted, with priority given to retaining core and important monitoring indicators. When adjusting recording rules, quantitative indicators need to optimize the collection frequency, granularity, and verification standards based on data fluctuation stability and collaborative adaptability. For qualitative behavioral indicators, the behavioral description dictionary is revised to address issues of coding consistency and record integrity, the responsible recording entity is clarified, or the process is simplified. All rules are deeply integrated with modules such as data collection and correlation model construction to ensure that the indicator library fits the needs of the group and the scenario, and that the collected data is standardized and usable.

[0024] The above embodiments, through the multi-node update triggering mechanism of the dynamic indicator library and the functions of adding and deleting indicator entries and adjusting recording rules, achieve the integration and adaptation of quantitative indicators and qualitative behavioral indicators, accurately matching the differentiated needs of different groups of students / adults and different scenarios of campus / rehabilitation, and solving the problem of insufficient adaptability of static indicator systems; establish a unified collection standard based on the standardized four-dimensional indicator system, and enhance the data collection accuracy of core capability shortcomings by combining the collaborative mapping parameters of the correlation model, and integrate four-dimensional data from multiple scenarios to form a complete dataset, solving the problems of lack of unified standards for data collection and insufficient accuracy of core data; formulate personalized intervention plans based on individual growth trajectory, four-dimensional development status and self-healing potential data, and compare intervention data in real time through a dynamic threshold adjustment mechanism to flexibly adjust intervention time limits, implementation frequency and cross-dimensional content ratio, solving the problems of templated intervention plans and delayed adjustments; extract post-intervention evaluation data through an iterative optimization module, continuously update the dynamic indicator library and intervention path table, and simultaneously optimize data collection weights, model parameters and implementation subject guidance, forming a complete closed loop of "assessment-intervention-feedback-optimization", solving the problems of disconnect between assessment and intervention and inability to continuously adapt to individual development.

[0025] Specifically, after the iterative optimization module updates the dynamic indicator library and intervention path table of the four-start system, it also includes: Based on the updated dynamic indicator library of the four-start system, the weight of the data collection dimensions of the data collection module was optimized, and differentiated data collection priorities were set for different target groups such as students and adults, as well as different scenarios such as campus and rehabilitation. By combining intervention effectiveness assessment data, a dynamic adjustment threshold is configured for short-term intervention plans. The real-time intervention data collected by the post-intervention data collection module is compared with the dynamic adjustment threshold, and the intervention path planning module is triggered to adjust the intervention project time limit and implementation frequency. The iterative correlation model collaborative mapping parameters are synchronized to the data collection module to guide it in improving the accuracy of data collection related to core capability shortcomings. At the same time, the indicator adaptability assessment results are fed back to the implementation entity of the target object's scenario, providing specific guidance on indicator understanding and data recording.

[0026] Specifically, when configuring dynamic adjustment thresholds for short-term intervention programs and triggering intervention pathway adjustments, the following includes: Based on the target group type and the core competency shortcomings, the basic adjustment thresholds for the short-term intervention plan are determined. The basic adjustment thresholds are set separately for four dimensions: psychological structure construction, physical fitness reserve cultivation, habit system formation, and core competency enhancement. By combining the intervention response efficiency of similar target subjects in historical intervention data, the basic adjustment threshold is calibrated to obtain the dynamic adjustment threshold range for the appropriate individual. The dynamic adjustment threshold range is divided into a lower threshold and an upper threshold. The intervention participation rate, behavioral change magnitude, and training execution achievement rate data collected in real time by the post-intervention data acquisition module are compared with the dynamically adjusted threshold range: When real-time data is below the lower threshold, the intervention path planning module is triggered to shorten the duration of a single intervention project and increase the frequency of implementation. When real-time data exceeds the upper limit threshold, the intervention path planning module is triggered to extend the interval period of intervention projects and optimize the cross-dimensional content allocation. When real-time data is between the lower and upper thresholds, keep the original intervention plan parameters unchanged and record the data fluctuation trend as the basis for the next round of threshold calibration.

[0027] Specifically, the baseline adjustment threshold is calibrated by combining the intervention response efficiency of similar target groups in historical intervention data, including: The target objects of the same type are classified in two ways: group type and core competency weakness dimension. An intervention response database for the same type of object is constructed. The intervention response database includes historical basic threshold, intervention cycle achievement rate, weakness improvement rate and cross-dimensional collaborative improvement data. Define intervention response efficiency evaluation indicators, including short-term intervention goal achievement rate, data fluctuation stability, and dimensional fit, and calculate the comprehensive response efficiency value according to preset weights. Preset optimal response thresholds and acceptable response thresholds, and classify calibration levels based on overall response efficiency values: When the overall response efficiency value is greater than or equal to the optimal response threshold, the calibration coefficient is 1.2; When the overall response efficiency value is greater than or equal to the qualified response threshold and less than the optimal response threshold, the calibration coefficient is 1.0; When the overall response efficiency value is less than the qualified response threshold, the calibration coefficient is 0.8; The calibrated dynamic thresholds are calculated separately for each dimension of the four-start system: the lower threshold is equal to the basic lower threshold multiplied by the calibration coefficient, and the upper threshold is equal to the basic upper threshold multiplied by the calibration coefficient. If the sample size of historical data for the same target object is greater than or equal to the preset sample threshold, the weighted average method is used to correct the calibration coefficient. The weight factors are allocated according to the evaluation priority of short-term intervention target achievement rate, data fluctuation stability, and dimensional fit. If the sample size is less than the preset sample threshold, the system's preset dimension adaptation calibration benchmark value is called to supplement and correct it, and finally the dynamic adjustment threshold range of the adapted individual is determined.

[0028] Specifically, based on the updated dynamic indicator library of the Four-Start System, the data collection module's collection dimension weights are optimized—with a preset base weight of 0.5 for primary labels, 0.3 for secondary labels, and 0.2 for tertiary labels. Considering different target groups such as primary school students, middle school students, working professionals, and rehabilitation populations, and the characteristics of different school and rehabilitation scenarios, a priority coefficient of 1.2 is set for the psychological structure construction and habit system formation dimensions of student groups in school scenarios, and a priority coefficient of 1.3 is set for the physical fitness reserve cultivation and core competency improvement dimensions of adult groups in rehabilitation scenarios. The comprehensive weight is calculated according to "priority coefficient × label base weight," and the collection order is determined. Simultaneously, intervention effectiveness evaluation data is used to tailor short-term intervention plans. The system sets thresholds for intervention participation and training execution achievement rate in percentage form, and thresholds for behavioral change magnitude in 100-point scale. It compares the real-time intervention data from the post-intervention data collection module with these thresholds and triggers the intervention path planning module to adjust the intervention project time limit and implementation frequency. It also synchronizes the collaborative mapping parameters of the correlation model with a correlation coefficient ≥ 0.7 for the core weakness dimension after iteration to the data collection module, guiding it to improve the accuracy of quantitative indicator collection related to core capability weaknesses to ±5% and the Kappa coefficient of qualitative indicator coding consistency ≥ 0.85. At the same time, it provides feedback on indicator suitability assessment results to the implementing entity through the system interface and provides special guidance including indicator definition, recording format, and abnormal data processing procedures. Specifically, when configuring dynamic adjustment thresholds and triggering intervention path adjustments for short-term intervention plans, basic adjustment thresholds are first set according to the four dimensions of the four-stage intervention system, based on the target group type and core competency gap dimensions (psychological structure construction dimension: intervention participation 40%-75%, behavioral change range 8-18 points, training execution achievement rate 45%-80%; physical fitness reserve cultivation dimension: intervention participation 35%-70%, behavioral change range 10-20 points, training execution achievement rate 50%-85%; habit system formation dimension: intervention participation 45%-80%, behavioral change range 7-16 points, training execution achievement rate 40%-75%; core competency enhancement dimension: intervention participation 50%-85%, behavioral change range 7-16 points, training execution achievement rate 40%-75%; physical fitness reserve cultivation dimension: intervention participation 50%-85%, behavioral change range 7-16 points, training execution achievement rate 40%-75%; physical fitness reserve ...). The change range is 9-19 points, and the training execution achievement rate is 55%-90%. Then, combined with the intervention response efficiency of similar target objects in historical intervention data, a dynamic adjustment threshold range containing a lower threshold and an upper threshold is obtained. Subsequently, the real-time collected data on intervention participation, behavioral change range, and training execution achievement rate are compared with this range. When the real-time data is less than the lower threshold, the duration of a single intervention project is shortened by 20%-30%, and the implementation frequency is increased to 1.5-2 times the original frequency. When it is greater than the upper threshold, the interval of the intervention project is extended by 50%, and the cross-dimensional content ratio is optimized by 50% each for the core weakness dimension and the collaboration dimension. When it is within the range, the original parameters remain unchanged, and the post-intervention data collection module records the data fluctuation trend once a day. Specifically, the process of calibrating the basic adjustment threshold by combining the intervention response efficiency of similar target groups in historical intervention data includes: classifying similar target groups by both group type and core competency weakness dimension, constructing an intervention response database containing historical basic thresholds, intervention cycle achievement rate, weakness improvement rate, and cross-dimensional collaborative improvement data; defining evaluation indicators such as short-term intervention goal achievement rate (weight 0.4), data fluctuation stability (weight 0.3), and dimensional fit (weight 0.3); and calculating a comprehensive response efficiency value (out of 100 points) by weighted summation; and setting an optimal response threshold of 85 points and 6... The acceptable response threshold is 0 points. The calibration coefficient is 1.2 when the comprehensive response efficiency value is ≥85 points, 1.0 when the comprehensive response efficiency value is ≤60 points and <85 points, and 0.8 when <60 points. The dynamic thresholds for each dimension are calculated according to the formula "lower threshold = basic lower threshold × calibration coefficient, upper threshold = basic upper threshold × calibration coefficient". When the historical data sample size of the same target object is ≥50 groups, the calibration coefficient is corrected by the weighted average method consistent with the weight of the evaluation index. When the sample size is <50 groups, the system's preset dimension adaptation calibration benchmark value is called to supplement and correct the coefficient. Finally, the dynamic adjustment threshold range of the adapted individual is determined.

[0029] Understandably, by optimizing the weights and priorities of data collection dimensions in the data collection module, configuring and calibrating dynamic adjustment thresholds for short-term interventions tailored to individuals, and synchronously linking model parameters to enhance the accuracy of core weakness data collection, combined with real-time intervention data to dynamically adjust intervention duration and frequency, and optimizing cross-dimensional content allocation, while simultaneously constructing an intervention response database for similar objects through dual classification, calculating the comprehensive response efficiency value by weighting according to evaluation indicators and correcting the calibration coefficient based on sample size, accurate threshold adaptation is achieved. Furthermore, feedback on indicator suitability assessment results and provision of specific guidance are provided to the implementing entity, ultimately improving the targeting and accuracy of data collection, ensuring the dynamic adaptability and individual suitability of intervention plans, strengthening the standardization of intervention execution, and comprehensively enhancing the overall effectiveness of growth assessment and intervention.

[0030] Specifically, the association model building module preprocesses and annotates the data, including: The dataset is divided into four sub-datasets according to the four-start system dimensions: psychological structure construction, physical fitness reserve cultivation, habit system formation, and core competency enhancement. The preprocessing stage involves data filtering: a preset threshold for the proportion of missing values ​​is set, and sample data with a proportion of missing values ​​greater than the threshold are removed. Interpolation is used to correct outlier data. Perform data standardization: convert quantitative indicators into a uniform scale using the Z-score method, and convert qualitative behavioral indicators into computable data according to preset coding rules; In the feature annotation stage, according to the definition of the dynamic indicator library, dimension labels, scene labels, subject labels and priority labels are added to the preprocessed data: dimension labels correspond to four-dimensional classification, scene labels distinguish between campus, family, practice and rehabilitation scenarios, subject labels indicate the data source, and priority labels are divided into first-level labels, second-level labels and third-level labels according to the importance of the indicators. The data sources include behavioral records, training logs, self-evaluation reports, peer evaluation results, teacher evaluation forms, and professional assessment reports.

[0031] Specifically, the rules for dividing priority labels include: A pre-defined indicator importance assessment system is established, which includes three dimensions: dimension impact, data sensitivity, and intervention relevance. Each dimension is assigned a pre-defined weight, with the weight allocation rule being that dimension impact is greater than data sensitivity, which is greater than intervention relevance, and the total weight is 1. Set a primary threshold and a secondary threshold, where the primary threshold is greater than the secondary threshold. Priority labels are assigned based on the overall score. When the overall score of the importance of an indicator is greater than or equal to the first-level threshold, it is marked as a first-level label, corresponding to the core intervention indicator; When the overall score of the indicator's importance is greater than or equal to the secondary threshold but less than the primary threshold, it is labeled as a secondary label, corresponding to an important monitoring indicator; When the overall score of the indicator's importance is less than the second-level threshold, it is labeled as a third-level label, corresponding to the regular recorded indicator; Among them, cognitive calibration in the psychological structure construction dimension and sustainable growth performance indicators in the core competency enhancement dimension are marked as first-level labels by default, while daily routine and diet norms in the physical fitness reserve cultivation dimension and daily routine execution indicators in the habit system formation dimension are marked as second-level labels by default.

[0032] Specifically, the association relationship training and validation of the association model construction module includes: A four-dimensional sub-model is constructed using the gradient boosting tree algorithm, which learns the association rules of indicators in the dimensions of psychological structure construction, physical fitness reserve cultivation, habit system formation, and core competency improvement. An inter-dimensional collaborative mapping layer is constructed using an attention mechanism to calculate the contribution coefficient of each dimension to the improvement of core capabilities and to establish an interconnection network. During training, a five-fold cross-validation method is used, with a preset error threshold. The dataset is divided into a training set and a validation set according to a preset ratio, and the model parameters are iteratively optimized until the prediction error of the validation set is less than the error threshold. The verification process involves calling upon historical intervention data sets, setting a pre-defined fit threshold, and comparing the fit between the model's output data on core weaknesses and self-healing potential with the actual intervention effects. When the fit is greater than or equal to the fit threshold, the model is deemed to have passed verification. When the fit is less than the fit threshold, return to the data preprocessing stage to re-optimize the feature annotation rules.

[0033] Specifically, the association model construction module preprocesses and annotates the data, including splitting the dataset into four sub-datasets according to the four-dimensional framework: psychological structure construction, physical fitness reserve cultivation, habit system formation, and core competency enhancement. The preprocessing stage first performs data screening: a preset threshold of 30% for missing values ​​is set, removing samples with more than 30% missing values; outliers are corrected using linear interpolation; and then data standardization is performed: quantitative indicators are converted to a uniform level with a mean of 0 and a standard deviation of 1 using the Z-score method, and qualitative behavioral indicators are standardized using a three-level coding system. Then (first-level dimension code, second-level performance code, and third-level intensity code) are converted into computable data. In the feature annotation stage, according to the definition of the dynamic indicator library, dimension labels (such as "psychological structure construction-01"), scene labels (campus-01, family-02, practice-03, rehabilitation-04), subject labels (labeling behavior records-01, training logs-02, self-evaluation reports-03, peer evaluation results-04, teacher evaluation forms-05, professional assessment reports-06, etc., data sources) and priority labels (first-level-01, second-level-02, and third-level-03) are added to the preprocessed data. Specifically, the priority labeling rules are based on a preset indicator importance assessment system that includes dimensional impact (weight 0.4), data sensitivity (weight 0.3), and intervention relevance (weight 0.3) (the total weight is 1 and dimensional impact > data sensitivity > intervention relevance). A first-level threshold of 80 points and a second-level threshold of 60 points (out of 100 points) are set. When the comprehensive score of indicator importance is ≥80 points, it is labeled as a first-level label (core intervention indicator). When the comprehensive score is ≤60 points and <80 points, it is labeled as a second-level label (important monitoring indicator). When the comprehensive score is <60 points, it is labeled as a third-level label (routine recording indicator). Among them, the cognitive calibration of the psychological structure construction dimension and the sustainable growth performance indicators of the core ability improvement dimension are labeled as first-level labels by default. The work and rest diet norms of the physical reserve cultivation dimension and the daily norm execution indicators of the habit system formation dimension are labeled as second-level labels by default. Specifically, the association model construction module's association relationship training and validation includes using the XGBoost gradient boosting tree algorithm to construct a four-dimensional sub-model, learning the association rules of indicators in each dimension, constructing a collaborative mapping layer between dimensions through an attention mechanism, calculating the contribution coefficient of each dimension to the improvement of core capabilities in the 0-1 interval, and establishing a weighted association network. During training, a five-fold cross-validation method is used, with a preset error threshold of 5% (mean squared error). The dataset is divided into training and validation sets in a 7:3 ratio, and the model parameters are iteratively optimized until the prediction error of the validation set is <5%. In the validation stage, historical intervention data sets are called, with a preset fit threshold of 85%. The fit between the core weakness location data and self-healing potential data output by the model and the actual intervention effect is compared. If the fit is ≥85%, the model is considered to have passed validation; if it is <85%, it is returned to the data preprocessing stage to re-optimize the feature labeling rules.

[0034] Understandably, by splitting the dataset according to the four dimensions of the system, removing invalid samples based on a preset threshold for the proportion of missing values, correcting outlier data using interpolation, and combining Z-score standardization of quantitative indicators with preset coding rules to transform qualitative indicators, data unification and computability are achieved. Furthermore, through multi-dimensional labeling and explicit weight allocation and threshold priority rules (including default label settings), the importance of indicators and data attributes are accurately distinguished. Simultaneously, a gradient boosting tree algorithm is used to construct a four-dimensional sub-model, an attention mechanism is used to construct a collaborative mapping layer, and five-fold cross-validation (split the dataset according to a preset ratio and iteratively optimize until the prediction error is below the threshold), fit verification (comparing model output with actual intervention effects), and a closed-loop optimization mechanism are employed to effectively improve data quality and labeling accuracy, strengthen the correlation analysis of indicators across dimensions and the synergy between dimensions, ensure the reliability of model training and the accuracy of output results (core weakness location, self-healing potential data), provide scientific and precise data support for subsequent personalized intervention path planning, and ultimately improve the targeting and effectiveness of growth assessment and intervention. Specifically, the intervention path planning module allocates implementation entities based on data sources, including: When the data comes from classroom status records and group activity participation data in a campus setting, the implementing entity is set as the teacher, and the collaborating entity is configured as the school psychological counselor. The teacher is responsible for the implementation of the intervention project and daily records, while the psychological counselor provides psychological technical support. When the data comes from the daily routine records and self-management behavior data of the family scene, the implementing entity is set as the parent and the collaborating entity is set as the class teacher. The parent takes the lead in the implementation of intervention in the family scene, and the class teacher regularly updates the school performance and provides guidance. When the data comes from social participation feedback and skills application performance data in practical scenarios, the implementing entity is set as the practical activity mentor, and the collaborating entity is configured as professionals in relevant fields. The mentor is responsible for on-site intervention and guidance, and provides skills adaptation suggestions to professionals in relevant fields. When the data comes from behavioral improvement records and symptom relief data in rehabilitation scenarios, the implementing entity is set as professional intervention personnel, and the collaborating entity is configured as rehabilitation institution technicians. Professional intervention personnel develop rehabilitation intervention plans, and technicians are responsible for assisting in the implementation of training. All implementing entities and collaborating entities intervene in the progress in real time through the system data interface, forming a closed-loop implementation mechanism of primary responsibility execution and collaborative support.

[0035] Understandably, accurately allocating implementing and collaborating entities based on data sources and corresponding scenario characteristics, clarifying the primary responsibilities of teachers, parents, practical activity mentors, and professional intervention personnel, and the collaborative support responsibilities of school counselors and homeroom teachers, and achieving real-time synchronization of intervention progress through system data interfaces, constructs a closed-loop implementation mechanism of primary responsibility execution and collaborative support. This ensures the professionalism and adaptability of intervention execution in different scenarios, avoids the limitations of single-entity intervention and the disconnect between scenario-based intervention, strengthens the standardization and collaboration of intervention execution, effectively improves the efficiency and effectiveness of intervention program implementation, and provides reliable execution guarantees for the entire closed-loop process of growth assessment intervention. Specifically, qualitative behavioral indicators are converted into computable data according to preset coding rules, including: Qualitative behavioral indicators are divided into four categories of coding objects according to the four-start system dimensions: psychological tendency, physical performance, habit execution, and ability application. Each category of coding object corresponds to a preset behavioral description dictionary. A three-level coding hierarchy is established: first-level dimension code, second-level representation code, and third-level strength code; A fixed weight relationship is assigned to each level: the weight of the second-level performance code is greater than the weight of the third-level strength code, which is greater than the weight of the first-level dimension code. The formula for calculating the encoding value remains unchanged. The formula for calculating the encoding value is: the encoding value is equal to the sum of the first-level dimension code multiplied by the weight of the first-level dimension code, the second-level performance code multiplied by the weight of the second-level performance code, and the third-level strength code multiplied by the weight of the third-level strength code, multiplied by 10, and the result is rounded to the nearest integer. Establish a coding mapping table: psychological tendency class corresponds to first-level dimension code 1, physical performance class corresponds to first-level dimension code 2, habit execution class corresponds to first-level dimension code 3, and ability application class corresponds to first-level dimension code 4. Coding validity verification: A consistency threshold is preset, 10% of the qualitative data samples are randomly selected, and two professional evaluators independently code the data. The coding consistency coefficient is calculated, and the coding rule takes effect when the consistency coefficient is greater than or equal to the consistency threshold. If the consistency coefficient is less than the consistency threshold, the fuzzy description of the behavior description dictionary is revised, typical behavior examples are added, and the code is recoded until the consistency requirements are met. The revised coding rules are then stored in the dynamic indicator library as the basis for updates.

[0036] Specifically, qualitative behavioral indicators are converted into calculable data according to preset coding rules. This includes dividing the qualitative behavioral indicators into four categories based on the four-dimensional system: psychological tendency (corresponding to a preset behavioral description dictionary containing 20 core behavioral descriptions such as "proactive," "passive avoidance," and "emotional stability"), physical performance (a dictionary containing 15 descriptions such as "endurance achievement," "motor coordination," and "recovery rate"), habit execution (a dictionary containing 18 descriptions such as "timely completion," "consistent persistence," and "self-planning"), and ability application (a dictionary containing 22 descriptions such as "skill transfer," "problem solving," and "collaboration"). A three-level coding hierarchy is set: a first-level dimension code, a second-level performance code (subdivided into codes ranging from 01 to 20 according to behavioral type), and a third-level intensity code (1-5 points corresponding to "extremely weak" to "extremely strong"). Fixed weights are assigned to each level (second-level performance code weight 0.5, third-level intensity code weight 0.3, first-level dimension code weight 0.2, satisfying second-level > third-level > first-level). The coding value calculation formula is: Coding value = (first-level dimension code × 0.2 + second-level performance code × 0.2) The result is calculated as 0.5 + Level 3 strength code × 0.3) × 10, rounded to the nearest integer. A coding mapping table is established: psychological tendency corresponds to Level 1 dimension code 1, physical performance corresponds to Level 1 dimension code 2, habit execution corresponds to Level 1 dimension code 3, and ability application corresponds to Level 1 dimension code 4. During coding validity verification, a consistency threshold of 0.8 (Kappa coefficient) is preset. 10% of the qualitative data samples are randomly selected and independently coded by two professionals with over 3 years of assessment experience in the relevant field. The coding consistency coefficient is calculated using Cohen's Kappa coefficient formula. The coding rules take effect when the consistency coefficient is ≥ 0.8. If the consistency coefficient is < 0.8, the vague expressions in the behavior description dictionary are revised (e.g., defining "proactive" as "actively initiating interactions and proactively undertaking tasks"), and typical behavioral examples are added (e.g., adding "no obvious emotional loss of control when encountering setbacks, recovering calm within 10 minutes" to the psychological tendency category "emotional stability"). The coding is then re-coded until the consistency coefficient reaches the target. The revised coding rules are simultaneously stored in the dynamic indicator database as a basis for subsequent updates. Understandably, by dividing qualitative behavioral indicators into coding objects according to the four-dimensional framework and configuring a pre-defined behavioral description dictionary, setting a three-level coding hierarchy, clarifying fixed weight relationships and quantitative calculation formulas and coding mapping tables, and coupled with an effectiveness verification mechanism of "randomly selecting 10% of samples + independent coding by two professional assessors + consistency coefficient verification" (including closed-loop optimization of fuzzy description revision, supplementation of typical examples, and rule archiving updates), the standardized and precise transformation of qualitative behavioral indicators into calculable data has been achieved. This ensures the uniformity, consistency, and reliability of coding, while fully preserving key information such as the dimensional attributes, performance characteristics, and intensity differences of behavior. It effectively solves the problem of the difficulty in quantifying and analyzing qualitative data, providing high-quality and reusable quantitative data support for correlation model training, core weakness identification, and personalized intervention path planning, thereby improving the scientificity and accuracy of growth assessment and intervention.

[0037] Specifically, the evaluation methods for the self-healing potential data output by the correlation model building module include: The self-healing potential assessment is determined in three dimensions: willingness to adjust autonomously, ability to repair through dimensional synergy, and environmental adaptability, which correspond to the core driving factors of the four-start system. Collect basic data on self-healing potential: including natural changes in indicators of target subjects during periods without intervention, records of proactive improvement behaviors, cross-scenario behavior transfer performance, and evaluation data on external environmental support; Calculate the self-healing potential score for each dimension, set the self-healing potential level threshold, which is divided into high potential threshold, medium potential threshold, and low potential threshold. The correlation model outputs targeted self-healing potential data according to the level. If there is no historical data for the target object during the intervention period, the baseline value of self-healing potential in the same dimension of the same group is used. The deviation rate between the target object's initial indicator data and the group mean is used for correction. The deviation rate is equal to the difference between the target object's initial value and the group mean divided by the group mean. The correction formula is: the corrected self-healing potential value is equal to the sum of the baseline value multiplied by 1 and the deviation rate multiplied by 0.5. Finally, the corrected self-healing potential level and data are output.

[0038] Specifically, the evaluation method for the self-healing potential data output by the association model construction module includes determining three dimensions for self-healing potential assessment: willingness to adjust autonomously (corresponding to the core driving factor of psychological structure construction), synergistic repair ability (corresponding to the synergistic driving factor of physical reserve cultivation and habit system formation), and environmental adaptability (corresponding to the driving factor of core ability improvement). Basic data on self-healing potential is collected, including natural changes in indicators recorded daily by the target subject during a 30-day no-intervention period, the frequency and quality of proactive improvement behaviors, the success rate of behavioral transfer across three scenarios: campus, family, and practice, and external environmental support evaluation data converted to a percentage system using a 5-point scale (1-5 points correspond to "very weak" to "very strong"). The self-healing potential score for each dimension is calculated (each dimension has a maximum score of 100 points, calculated using a weighted summation of basic data, with a weight allocation of 0.4 for willingness to adjust autonomously). The self-healing potential level thresholds are set as follows: high potential threshold 80 points, medium potential threshold 60 points, and low potential threshold 60 points (excluding 60 points). The correlation model outputs the corresponding level and specific score of self-healing potential data based on the comprehensive score. If there is no historical data of the target object during the intervention period, the self-healing potential benchmark value of the same age group and the same group (student / adult) in the same dimension (preset as 65 points) is used. The deviation rate is corrected by combining the initial index data of the target object with the group mean of more than 300 groups of historical data of the group. The deviation rate = (initial value of the target object - group mean) / group mean (the value range is [-0.5, 0.5], and the value outside the range is ±0.5). The correction formula is: corrected self-healing potential value = benchmark value × (1 + deviation rate × 0.5). Finally, the corrected self-healing potential level (high / medium / low) and specific score (rounded to the nearest integer) are output.

[0039] Understandably, by clearly defining the three dimensions of self-healing potential assessment (willingness to adjust autonomously, ability to repair through synergy of dimensions, and adaptability to the environment) and corresponding to the core driving factors of the four-stage system, collecting multiple types of basic data (natural changes in indicators during periods without intervention, proactive improvement behaviors, etc.), setting grading thresholds to achieve precise level classification, and using a combination of baseline values ​​from the same group and deviation rate correction (quantitative adaptation through correction formulas) for scenarios without historical data, a comprehensive and scientific assessment of self-healing potential data is achieved. This not only covers the adaptability of different scenarios with / without historical data but also accurately captures the core influencing factors of self-healing potential, providing a differentiated basis for intervention path planning (such as optimizing intervention intensity for high potential and strengthening targeted measures for low potential), avoiding over- or under-intervention, and further improving the comprehensiveness of growth assessment and the personalization and adaptability of intervention plans.

[0040] Specifically, it also includes a data security and privacy protection module, which is configured to conduct full lifecycle security management of all data involving personal information of the target object obtained by the data collection module and the post-intervention data collection module, and to encrypt the data during transmission using an asymmetric encryption algorithm.

[0041] Understandably, the data security and privacy protection module implements full lifecycle security management of the target's personal information data. It combines asymmetric encryption algorithms to ensure data transmission security, role-based access control models to standardize operational permissions, and a dedicated guardian authorization access mechanism for students to strengthen the protection of special subjects. This effectively prevents security risks such as data leakage and misuse, strictly complies with relevant regulations on personal information protection, fully protects the privacy rights of students and adults, and enhances the credibility and compliance of system data use. It provides a safe and reliable data environment to support growth assessment and intervention work based on the four-stage system.

[0042] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program goods. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program goods embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0043] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program goods according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, are configured to implement the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0044] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0045] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus to be configured to implement a process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0046] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A growth assessment intervention system based on a four- start system, characterized by, Comprise: The index library establishment module is configured to establish a four-dimensional dynamic index library covering psychological structure construction, physical reserve cultivation, habit system formation, and core ability improvement. The dynamic index library includes quantitative indicators and qualitative behavior indicators. An update trigger mechanism is set for semester evaluation nodes, intervention implementation completion nodes, index system related policy update nodes, and self-healing potential verification nodes. The index entries are added or deleted according to the preset rules, the index recording rules are adjusted, and a standardized four-dimensional index system is output. The data collection module is configured to collect four-dimensional data of target objects in need of behavior intervention or ability improvement in student and adult groups in multiple scenarios based on the four-dimensional index system, and form a data set by summarizing the data. The correlation model construction module is configured to receive the data set and construct a correlation model between four-dimensional data and individual growth trajectory. The data is preprocessed, feature labeled, correlation relationship trained and verified according to the index definition of the dynamic index library, and four-dimensional development state, cooperative adaptation degree, core ability short board and self-healing potential data are output. The intervention path planning module is configured to determine the intervention key and cross-dimensional content based on the output data of the correlation model construction module, develop short-term and long-term intervention plans, assign implementation subjects according to data sources, and form an intervention path table containing intervention projects, scenarios, and time limits. The post-intervention data collection module is configured to collect post-intervention data at fixed intervals, intervention completion nodes, and before the next round of evaluation, and maintain the same data collection standards as the data collection module. The post-intervention data set is formed by matching and organizing the data according to the index library dimensions. The iterative optimization module is configured to receive the post-intervention data set, extract index adaptability evaluation data and intervention effectiveness evaluation data, and iteratively update the four-dimensional dynamic index library and the intervention path table.

2. The four-gate based growth assessment intervention system of claim 1, wherein, After the iterative optimization module iteratively updates the four-dimensional dynamic index library and the intervention path table, it further comprises: Based on the updated four-dimensional dynamic index library, the data collection module optimizes the collection dimension weight. Different data collection priorities are set for different target objects of students and adults and different scene characteristics of campus and rehabilitation. The dynamic adjustment threshold is configured for the short-term intervention plan. The real-time intervention data collected by the post-intervention data collection module is compared with the dynamic adjustment threshold, and the intervention path planning module is triggered to adjust the intervention project time limit and implementation frequency. The iterative correlation model is synchronized to the data collection module with mapping parameters to guide the collection accuracy of core ability short board related data. The index adaptability evaluation results are fed back to the implementation subject of the target object's scene to provide special guidance for index understanding and data recording.

3. The four-gate based growth assessment intervention system of claim 2, wherein, When the dynamic adjustment threshold is configured for the short-term intervention plan and the intervention path adjustment is triggered, it comprises: Based on the group type and core ability short board dimension of the target object, the basic adjustment threshold of the short-term intervention plan is determined. The basic adjustment threshold is set for the four dimensions of psychological structure construction, physical reserve cultivation, habit system formation, and core ability improvement. The intervention response efficiency of the same target object in historical intervention data is combined to calibrate the basic adjustment threshold, and a dynamic adjustment threshold interval of the individual is obtained, the dynamic adjustment threshold interval is divided into a lower threshold and an upper threshold; The intervention participation, behavior change amplitude and training execution compliance rate data collected by the post-intervention data acquisition module in real time are compared with the dynamic adjustment threshold interval: When the real-time data is less than the lower threshold, the intervention path planning module is triggered to shorten the single duration of the intervention project and increase the implementation frequency; When the real-time data is greater than the upper threshold, the intervention path planning module is triggered to extend the intervention project interval period and optimize the cross-dimension content ratio; When the real-time data is between the lower threshold and the upper threshold, the original intervention plan parameters remain unchanged, and the data fluctuation trend is recorded synchronously as the basis for the next round of threshold calibration.

4. The four-gate based growth assessment intervention system of claim 3, wherein, The intervention response efficiency of the same target object in historical intervention data is combined to calibrate the basic adjustment threshold, and a dynamic adjustment threshold interval of the individual is obtained, the dynamic adjustment threshold interval is divided into a lower threshold and an upper threshold; The same target object is classified according to group type and core capability short board dimension, an intervention response database of the same object is constructed, and the intervention response database includes historical basic threshold, intervention period compliance rate, short board improvement rate and cross-dimension collaborative improvement degree data; An intervention response efficiency evaluation index is defined, the evaluation index includes short-term intervention target achievement rate, data fluctuation stability and dimension adaptation fit degree, and a comprehensive response efficiency value is calculated according to a preset weight; An optimal response threshold and a qualified response threshold are preset, and the calibration grade is divided according to the comprehensive response efficiency value: When the comprehensive response efficiency value is greater than or equal to the optimal response threshold, the calibration coefficient is 1.2; When the comprehensive response efficiency value is greater than or equal to the qualified response threshold and less than the optimal response threshold, the calibration coefficient is 1.0; When the comprehensive response efficiency value is less than the qualified response threshold, the calibration coefficient is 0.8; The calibrated dynamic threshold of each dimension of the four-start system is calculated: the lower threshold is equal to the basic lower threshold multiplied by the calibration coefficient, and the upper threshold is equal to the basic upper threshold multiplied by the calibration coefficient; If the sample size of the same target object historical data is greater than or equal to the preset sample threshold, the weighted average method is used to correct the calibration coefficient, and the weight factor is allocated according to the evaluation priority of the short-term intervention target achievement rate, the data fluctuation stability and the dimension adaptation fit degree; If the sample size is less than the preset sample threshold, the dimension adaptation calibration benchmark value preset by the system is called to supplement the correction, and the dynamic adjustment threshold interval of the individual is finally determined.

5. The four-gate based growth assessment intervention system of claim 1, wherein, The data preprocessing and feature labeling of the correlation model construction module include: The data set is divided into four sub-data sets of psychological structure construction, physical reserve cultivation, habit system formation and core capability improvement according to the four-start system dimension; Data screening is performed in the preprocessing link: a preset missing value proportion threshold is set, sample data with a missing value proportion greater than the missing value proportion threshold is removed, and abnormal data is corrected by interpolation method; Data standardization is performed: quantitative indicators are converted into a unified magnitude by Z-score method, and qualitative behavior indicators are converted into calculable data according to a preset coding rule; The feature labeling link adds dimension labels, scene labels, subject labels and priority labels to the preprocessed data according to the dynamic index library definition: the dimension labels correspond to four-dimensional classification, the scene labels distinguish between campus, family, practice and rehabilitation scenes, the subject labels label the data sources, and the priority labels are divided into first-level labels, second-level labels and third-level labels according to the importance of the indicators; Among them, the data sources include behavior records, training logs, self-evaluation reports, mutual evaluation results, teacher evaluation forms and professional evaluation reports.

6. The four-gate based growth assessment intervention system of claim 5, wherein, The division rules of the priority labels include: a preset indicator importance evaluation system, the evaluation system includes three dimensions of dimension influence degree, data sensitivity and intervention correlation, and each dimension is assigned a preset calculation weight, and the weight distribution rule is that the dimension influence degree is greater than the data sensitivity, and the intervention correlation is greater than 1; Set a first threshold and a second threshold, wherein the first threshold is greater than the second threshold, and the priority label is divided according to the comprehensive score: When the indicator importance comprehensive score is greater than or equal to the first threshold, it is labeled as a first-level label, corresponding to a core intervention indicator; When the indicator importance comprehensive score is greater than or equal to the second threshold and less than the first threshold, it is labeled as a second-level label, corresponding to an important monitoring indicator; When the indicator importance comprehensive score is less than the second threshold, it is labeled as a third-level label, corresponding to a regular record indicator; Among them, the cognitive calibration of the psychological structure construction dimension and the sustainable growth performance indicator of the core ability improvement dimension are labeled as first-level labels by default, and the rest of the indicators are labeled as second-level labels by default.

7. The four-gate based growth assessment intervention system of claim 5, wherein, The association relationship training and verification of the association model construction module includes: adopting gradient boosting tree algorithm to construct four-dimensional sub-models to learn the indicator association rules of the psychological structure construction, physical reserve cultivation, habit system formation and core ability improvement dimensions respectively; Through the attention mechanism, a dimension intercoordination mapping layer is constructed to calculate the contribution coefficient of each dimension to the core ability improvement and establish an association network; In the training process, a five-fold cross-validation method is used, a preset error threshold is set, the data set is divided into a training set and a validation set according to the preset proportion, and the model parameters are iteratively optimized until the prediction error of the validation set is less than the error threshold; In the verification link, a historical intervention data set is called, a preset fit degree threshold is set, the fit degree of the core short board positioning data and the self-healing potential data output by the model and the actual intervention effect is compared, and when the fit degree is greater than or equal to the fit degree threshold, it is determined that the model verification is passed; When the fit degree is less than the fit degree threshold, return to the data preprocessing link to optimize the feature labeling rules again.

8. The four-gate based growth assessment intervention system of claim 1, wherein, The qualitative behavior indicators are converted into calculable data according to the preset coding rules, including: Divide the qualitative behavior indicators into four categories of coding objects according to the four-dimensional system dimensions, namely psychological tendency, physical performance, habit execution and ability application, and each category of coding object corresponds to a preset behavior description dictionary; Set three coding levels: first-level dimension code, second-level performance code and third-level intensity code; Assign a fixed weight relationship to each level: the weight of the second-level performance code is greater than the weight of the third-level intensity code, which is greater than the weight of the first-level dimension code. The encoding value calculation formula remains unchanged, and the encoding value calculation formula is: the encoding value is equal to the product of the first-level dimension code, the weight of the first-level dimension code, the second-level performance code, the weight of the second-level performance code, the third-level intensity code and the weight of the third-level intensity code, and the sum is multiplied by 10. The result is kept as an integer; Establish an encoding mapping table: the psychological tendency class corresponds to the first-level dimension code 1, the physical performance class corresponds to the first-level dimension code 2, the habit execution class corresponds to the first-level dimension code 3, and the ability application class corresponds to the first-level dimension code 4; Coding validity verification: a preset consistency threshold is randomly selected, and ten percent of the qualitative data samples are independently coded by two professional evaluators. The coding consistency coefficient is calculated. When the consistency coefficient is greater than or equal to the consistency threshold, the coding rule takes effect. If the consistency coefficient is less than the consistency threshold, revise the fuzzy expression of the behavior description dictionary, supplement typical behavior examples, and recode until the consistency requirement is met. The revised coding rule is stored in the dynamic index library as an update basis.

9. The four-gate based growth assessment intervention system of claim 1, wherein, The evaluation method of the self-healing potential data output by the correlation model construction module includes: Determine the three dimensions of self-healing potential evaluation: self-adjustment willingness, dimension collaborative repair ability and environmental adaptation degree, which correspond to the core driving factors of the four-start system respectively; Collect self-healing potential basic data: including target object's natural change data of indicators in the non-intervention period, active improvement behavior record, cross-scene behavior migration performance and external environment support evaluation data; Calculate the self-healing potential score of each dimension, set the self-healing potential level threshold, and the self-healing potential level threshold is divided into high potential threshold, medium potential threshold and low potential threshold. The correlation model outputs targeted self-healing potential data according to the level; If the target object has no intervention period historical data, the same group and dimension self-healing potential benchmark value is used, and the deviation rate is corrected according to the deviation of the target object's initial indicator data and the group mean value. The deviation rate is equal to the difference between the target object's initial value and the group mean value divided by the group mean value. The correction formula is: the corrected self-healing potential value is equal to the benchmark value multiplied by 1 plus the deviation rate multiplied by 0.

5. Finally, the corrected self-healing potential level and data are output.

10. The four-gate based growth assessment intervention system of claim 1, wherein, Also includes: The data security and privacy protection module is configured to perform full-life-cycle security management and control on all data related to the target object's personal information collected by the data collection module and the post-intervention data acquisition module, and to perform encryption processing on the data in the transmission process using an asymmetric encryption algorithm.

Images