A method for individual educational development support based on innate biological feature assessment

Abstract

This invention discloses a method for supporting individual educational development based on innate biometric assessment. The method first collects fingerprint image data, which is considered lifelong and unchanging, from individuals. Second, it preprocesses the fingerprint images and extracts quantified feature vectors. Then, it inputs these feature vectors into a pre-trained biometric-educational trait association model to obtain quantified assessment scores for the individual across multiple preset educational trait dimensions. Next, combining the assessment scores, individual age, and developmental goals, a personalized initial educational development plan is generated by matching rules through an educational plan generation rule engine. This invention uses stable biometrics as an objective benchmark and, through a computable model and rule engine, achieves a complete technical process from innate trait assessment to the generation and optimization of long-term, dynamic, and personalized educational support plans, solving the technical problems of traditional methods that rely on unstable acquired assessments and lack a long-term reference system.

Description

Technical Field

[0001] This invention relates to the fields of educational technology and personalized education, and in particular to a method for supporting individual educational development based on innate biological characteristics assessment. Background Technology

[0002] Currently, personalized education has become a shared goal of both family and school education. In practice, the realization of personalized education mainly relies on observation, psychological assessment scales, and the experience-based judgment of teachers or parents. Common existing technological means include: qualitative descriptions based on observations of students' daily behavior, periodic assessments using standardized psychological or academic ability scales, and intelligent teaching systems that recommend learning paths by analyzing students' historical academic data.

[0003] However, these existing technological solutions have several inherent flaws, making it difficult to meet the needs for precise, stable, and forward-looking educational support for individuals. First, methods relying on observation and scale assessments are highly susceptible to interference from the subject's emotional state, testing environment, examiner's influence, and social desirability. They often reflect the "current state" rather than "stable traits," resulting in educational plans based on these results lacking long-term stable reference value and potentially requiring frequent adjustments as the child's condition fluctuates, leaving families and educators bewildered. Second, recommendation systems driven by historical performance data are essentially a feedback and optimization of "acquired performance," failing to reveal the innate cognitive preferences and learning styles behind a child's achievements. For example, two children with equally excellent math scores might have vastly different strengths: one might rely on strong logical reasoning, while the other might rely on excellent visual memory. Traditional methods cannot distinguish between these two distinct "materials," causing subsequent intensive development programs to deviate from their most suitable path, preventing the maximization of potential.

[0004] A deeper problem lies in the lack of a stable "anchor point" or "benchmark reference" across the entire lifecycle of individual development. Individual educational development is a dynamic process, with continuous interaction between the environment, educational interventions, and innate foundations. Without an accurate grasp of the relatively stable innate foundation, all interventions targeting the environment are like adjustments within a moving coordinate system, making it difficult to scientifically evaluate and continuously optimize their effects. The market urgently needs a technological method that can objectively and stably assess an individual's innate cognitive and personality trait foundation, and based on this foundation, combined with developmental theories, generate long-term educational development support programs that are both personalized and feasible. Currently, there is no mature technology that can systematically and effectively link lifelong, unchanging biological characteristics with the cognitive and personality trait models required for educational development, forming a complete technological closed loop from assessment to program generation to dynamic optimization. Summary of the Invention

[0005] To achieve the above objectives, this invention provides a method for supporting individual educational development based on innate biological characteristic assessment, comprising the following steps:

[0006] S100: Collect fingerprint image data of individuals to form a fingerprint image dataset;

[0007] S200: Process each fingerprint image in the fingerprint image dataset and extract the quantized feature vector of each fingerprint;

[0008] S300: Concatenate all quantized feature vectors into a comprehensive biometric input vector, and input the comprehensive biometric input vector into a pre-trained biometric-educational trait association model to obtain the educational trait dimension evaluation score vector output by the biometric-educational trait association model;

[0009] S400: The educational trait dimension assessment score vector, the individual's current age information, and the individual's set stage development goals are input into the education plan generation rule engine. The education plan generation rule engine matches and integrates the preset educational intervention rules to generate a personalized initial education development plan.

[0010] S500: Implement an initial personalized education development plan and periodically collect individual behavioral performance data during the implementation period;

[0011] S600: Based on behavioral performance data, conduct a comparative analysis of the effects of the initial personalized education development plan, optimize and generate an optimized personalized education development plan based on the results of the comparative analysis, and update the control database used to train the biometric-educational trait association model by using the comprehensive biometric input vector and the educational trait dimension score corrected based on behavioral performance data as new samples.

[0012] S700: Repeatedly execute S500 and S600: forming a continuous iterative loop.

[0013] Preferably, the extraction of quantized feature vectors in S200 specifically includes:

[0014] S210: Perform preprocessing on the fingerprint image, including grayscale conversion, contrast enhancement, and noise filtering, to obtain the enhanced fingerprint image;

[0015] S220: On the enhanced fingerprint image, the global pattern category of the fingerprint is identified and the core point of the fingerprint is located based on the orientation field and ridge tracking algorithm. The global pattern categories include arch pattern, loop pattern and whorl pattern.

[0016] S230: Construct a polar coordinate system with the fingerprint core point as the origin, and extract multi-dimensional quantitative features composed of pattern type coding, ridge density features, ridge curvature variation features, and triangulation features in the polar coordinate system. The pattern type coding is assigned a unique value according to the global pattern type. The ridge density feature is obtained by calculating the average number of ridges per unit length on radial lines from the core point at multiple different angles. The ridge curvature variation feature is obtained by calculating the curvature variation rate statistics along the selected ridge trajectory. The triangulation feature is obtained by calculating the relative distance and azimuth angle between the core point and the triangulation point when the fingerprint is a loop or whorl pattern.

[0017] Preferably, the biometric-educational trait association model in S300 is a multi-output regression model based on a deep neural network;

[0018] The training process of the biometric-educational trait association model is as follows: A comparison database is obtained, which contains a set of comprehensive biometric input vectors of historical individuals, and a set of standardized scores of historical individuals on multiple preset educational trait dimensions obtained through long-term behavioral observation and standardized contextual assessment; The deep neural network is trained under supervision using the set of comprehensive biometric input vectors as input and the set of standardized scores as the expected output, until the model loss function converges, thus obtaining the trained biometric-educational trait association model.

[0019] The pre-defined educational trait dimensions include logical reasoning tendency, language expression tendency, spatial imagination tendency, interpersonal sensitivity tendency, focus and perseverance tendency, and activity exploration tendency.

[0020] Preferably, the education program generation rule engine in S400 stores education intervention rules represented in the form of "IF-THEN";

[0021] The "IF" part of each educational intervention rule defines a specific educational trait dimension score combination pattern, age range, and development goal keywords;

[0022] The "THEN" section of each educational intervention rule defines the specific educational intervention measures, recommended intensity, recommended frequency, and expected goals that match the "IF" section.

[0023] The process of generating a personalized initial education development plan by the education plan generation rule engine is as follows: Iterate through all education intervention rules, match the rules that meet the definition of the "IF" part based on the education trait dimension assessment score vector, current age information, and stage development goals, integrate the "THEN" part of all successfully matched education intervention rules, resolve conflicts, and prioritize them, and finally generate a structured document.

[0024] Preferably, the structured document content of the initial personalized education development plan includes: a list of recommended activities to strengthen the individual's strengths and educational traits; a list of recommended supportive activities to address the individual's educational traits that require attention; a detailed description of parent-child interaction and communication methods that match the individual's educational trait dimension assessment score vector; a guide to creating physical and psychological parameters of the learning environment that are adapted to the individual's cognitive preferences; and a timeline for key development activities and quantifiable expected development milestones within the pre-set future period.

[0025] Preferably, the behavioral performance data periodically collected in S500 includes standardized situational task data and natural situational observation record data;

[0026] Standardized situational task data is obtained by individuals periodically completing standardized assessment tasks associated with preset educational trait dimensions. The standardized assessment tasks include logic puzzle tasks, story retelling tasks, graphic space puzzle tasks, and situational response questionnaire tasks. Each standardized assessment task produces a quantitative performance score.

[0027] Natural situation observation record data is collected by individual educators through a structured electronic observation record form. The electronic observation record form has preset fields for typical behavioral event categories, frequency recording fields, and free description fields related to each educational trait dimension. Educators record their observations in the corresponding fields.

[0028] Preferably, in S600, a comparative analysis of the effectiveness of the initial personalized education development plan is conducted based on behavioral performance data, specifically including:

[0029] S611: Extract behavioral performance data collected during the current assessment period and calculate the actual values ​​of key development indicators associated with each preset educational trait dimension;

[0030] S612: Compare the actual values ​​of key development indicators with the expected development milestone indicators set in the initial plan for personalized education development, and calculate the deviation from the progress of each key development indicator.

[0031] S613: Analyze the changing trends of behavioral performance data, identify the behavioral performance data segments corresponding to each specific educational intervention measure in the initial personalized education development plan, and determine whether the behavioral performance data segment shows a positive changing trend, no obvious changing trend, or a negative changing trend within the evaluation period.

[0032] Preferably, in S600, a personalized education development optimization plan is generated based on the results of effect comparison analysis, specifically including:

[0033] S621: For educational interventions that show a positive trend of change identified in the effect comparison analysis, the recommended implementation intensity and frequency of the educational interventions shall be maintained or increased in the personalized education development optimization program;

[0034] S622: For educational intervention measures that show no obvious trend of change in the effect comparison analysis, adjust the specific implementation strategy description of the educational intervention measures in the personalized education development optimization plan, or call the alternative equivalent educational intervention measures from the rule base of the education plan generation rule engine to replace them.

[0035] S623: For educational intervention measures that show a negative trend in the effect comparison analysis, the educational intervention measures shall be suspended or deleted in the personalized education development optimization plan;

[0036] S624: Based on the natural growth of an individual's age, update the stage-by-stage development goals and trigger the education program generation rule engine to generate a personalized education development optimization plan for the next assessment cycle based on the updated information.

[0037] Preferably, in S600, the comprehensive biometric input vector and the educational trait dimension scores corrected based on behavioral performance data are used as new samples to update the control database, specifically including:

[0038] S631: Based on the behavioral performance data collected during the current assessment period, especially the performance scores of standardized situational task data, the assessment score vector of the educational trait dimension obtained in S300 is calibrated to obtain the corrected educational trait dimension score vector. The calibration method is to take a weighted average of the assessment scores and the inferred scores based on behavioral data.

[0039] S632: Use the comprehensive biometric input vector of the individual as the input sample and the corrected educational trait dimension score vector as the output label to form a new training sample pair;

[0040] S633: Add new training sample pairs to the control database;

[0041] S634: When the cumulative number of newly added training sample pairs in the control database reaches the preset batch update threshold, the updated control database is used to perform incremental training or full retraining of the biometric-educational trait association model.

[0042] Preferably, the continuous iteration cycle in S700 is carried out with a fixed time period. Each complete iteration cycle includes a scheme execution and data collection phase and an effect evaluation and optimization phase. The duration of the scheme execution and data collection phase is shorter than that of the effect evaluation and optimization phase. The cycle length of the iteration cycle is dynamically adjusted according to the individual's age group. The younger the age, the shorter the cycle length of the iteration cycle.

[0043] The beneficial effects of this invention are:

[0044] 1. This invention uses lifelong, unchanging fingerprint biometrics as the starting point for analysis, transforming it into a quantitative assessment of an individual's cognitive and behavioral traits through a well-trained "biometric-educational trait association model." This assessment result is unaffected by temporary individual emotions, testing environment, or subjective will, fundamentally overcoming the instability of traditional psychological and behavioral assessment methods. It establishes a reliable "innate material map" for each individual throughout their entire developmental cycle, freeing long-term educational planning from the fluctuations and uncertainties caused by assessment benchmark variations, and providing a solid and consistent scientific basis for all subsequent educational interventions.

[0045] 2. This invention automatically transforms abstract assessment scores into structured, actionable, and personalized educational development plans through an "educational plan generation rule engine," thus realizing the operational implementation of the "teaching according to aptitude" concept. More importantly, the system continuously collects behavioral performance data to construct a complete feedback loop of "execution-assessment-optimization." This loop not only dynamically adjusts an individual's personalized educational plan based on their actual feedback, making it more aligned with their developmental response, but also continuously optimizes the core correlation model using accumulated new data. This design upgrades educational support from "one-off static suggestions" to an intelligent service system that "dynamically optimizes as the child grows."

[0046] 3. This invention organically integrates multiple technical aspects, including biometric identification, data modeling, rule-based reasoning, and effectiveness evaluation, into a coherent and standardized technical process. The method produces a structured solution that includes specific measures, implementation parameters, and expected goals, making the educational support process plannable, executable, and monitorable. Simultaneously, by defining key developmental indicators and comparing behavioral data, the intervention's effectiveness can be quantitatively analyzed and verified, changing the traditional reliance on subjective experience in family education and the difficulty in assessing its effects. This promotes family education towards a more scientific, systematic, and verifiable direction. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in this invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, those skilled in the art can obtain other drawings based on these drawings without creative effort.

[0048] Figure 1 This is a flowchart of the steps of the method of the present invention. Detailed Implementation

[0049] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should also be noted that, to make the embodiments more comprehensive, the following embodiments are the best and preferred embodiments, and those skilled in the art can use other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0050] Please see Figure 1 This invention provides a method for supporting individual educational development based on innate biometric assessment. Using an industry-certified live fingerprint scanner under uniform white light illumination, it acquires planar rolling fingerprint images of all ten fingers of an individual. During acquisition, the individual is guided to fully align the pads of each finger with the acquisition window, ensuring that the obtained image contains the complete and valid area from the center of the fingertip to the flexor crease of the first interphalangeal joint. Each image is digitally stored at a resolution of at least 500 DPI and bound to the individual's unique identification code, forming a structured raw fingerprint image dataset. The standardized, high-quality images generated in this step form the basis for all subsequent analyses.

[0051] The process of processing fingerprint images and extracting quantized feature vectors involves a series of automated operations. First, an adaptive histogram equalization algorithm is used to enhance each grayscale fingerprint image, improving the contrast between ridges and valleys. Then, a Gabor filter bank is used to enhance the orientation field and smooth noise. On the preprocessed image, a thinning algorithm is used to extract the skeleton structure of the fingerprint ridges, and the Poincaré exponent method is used to locate singular points such as core points and triangulation points. Finally, a region of interest is defined around the core point, and within this region, ridge counts in different fan-shaped areas are statistically analyzed, and the local consistency of ridge orientation is calculated, generating a high-dimensional feature vector composed of nearly a hundred statistical measures. This vector objectively represents the morphological properties of the fingerprint.

[0052] The feature vectors of all ten fingers are concatenated in a predetermined order and then input into a biometric-educational trait association model composed of a five-layer fully connected neural network. This model is pre-trained on a database containing 100,000 pairs of "fingerprint feature-behavioral trait labels." The model outputs a six-dimensional vector, with each component corresponding to a standardized score for traits such as logical reasoning and verbal expression. This step transforms biometric features, which are difficult to understand intuitively, into quantifiable and comparable psychological trait indicators, providing objective, data-driven insights for educational planning.

[0053] The generation of an initial personalized education development plan relies on a knowledge base containing thousands of production rules. The rule engine matches the received trait scores, age, and target keywords with the logical patterns in the rule antecedents. For example, if "the logical reasoning score is above the threshold and the age is 6-8 years old," a series of recommended measures containing keywords such as "logic games" and "structured blocks" are triggered. The engine aggregates and resolves conflicts in all triggered rule consequents, ultimately generating a structured PDF document. This document not only lists activity suggestions but also details the logical connection between each suggestion and the child's traits, specific operational steps, and precautions.

[0054] The implementation and data collection phases are conducted through a dedicated mobile application. This application regularly sends standardized observation tasks and situational assessments to parents. For example, it might send a short story each week that the child needs to verbally retell to assess language expression, or record the child's emotional reactions when faced with unfinished tasks to assess perseverance. Parents can record text descriptions, select ratings, or upload video clips through the application interface. All data is timestamped and tagged with activity tags, encrypted, and uploaded to a cloud server, forming a continuous time-series of behavioral performance data.

[0055] Effectiveness evaluation and optimization is a data-driven decision-making process. The system analyzes collected behavioral data quarterly, calculating the rate of change of key indicators relative to the baseline. By comparing actual changes with expected outcomes, the system automatically categorizes interventions into three types: "effective," "neutral," and "needs adjustment." The optimization algorithm dynamically combines these labels with the child's latest age stage from a rule base to generate the next cycle's optimization plan. Simultaneously, the "feature vector - calibrated trait score" data pairs from the current cycle are quality-labeled and added to the model's retraining pool.

[0056] The continuous iterative cycle runs automatically in three-month standard cycles. At the end of each cycle, the system automatically generates a family report that includes a review of the effects and a plan for the next phase. As the child grows, the system not only adjusts the content of the activities but also adjusts the focus of data collection and some parameters of the assessment model based on developmental psychology theories. This accompanying and adaptive working mode ensures that educational support keeps pace with the child's dynamic development, achieving long-term and consistent support.

[0057] In one possible implementation, the extraction of quantized feature vectors is carried out as follows: During the image preprocessing stage, the CLAHE algorithm is used for contrast-limited adaptive histogram equalization, which effectively enhances the ridge details in blurred areas while suppressing excessive noise. Pattern classification is achieved by calculating the consistency pattern of the global orientation field of the fingerprint. For example, whorl patterns will exhibit more than two core points and complex spiral structures, and the system automatically classifies them using a pattern matching algorithm.

[0058] After locating the core point, the system constructs a polar coordinate grid with the core point as the origin. Ridge density features are obtained by measuring the number of ridge intersections within a fixed distance from the core point to the edge in 72 equally divided angular directions. Ridge curvature variation features are quantified by tracing the ridge skeleton and calculating the standard deviation of the variation difference values ​​of the direction angles of its continuous line segments. For fingerprints with triangular points, the system precisely measures the pixel distance between the core point and the triangular points, as well as the angle between the connecting line and the horizontal axis. All features are Z-score normalized after extraction to eliminate the scale effect caused by differences in acquisition devices. This series of refined measurements transforms the visual pattern into precise coordinates in a high-dimensional data space.

[0059] In one possible implementation, the biometric-educational trait association model is implemented as follows: The model employs a fully connected neural network structure with one input layer, three hidden layers, and one output layer. The number of neurons in the input layer is consistent with the dimension of the concatenated integrated biometric input vector. Each hidden layer is followed by a Dropout layer to prevent overfitting, and the ReLU function is used as the activation function. The output layer uses a linear activation function to directly output the predicted scores of the six educational traits.

[0060] The model's training relies on a continuously updated control database. Each record in this database contains an individual's fingerprint feature vector and a multidimensional behavioral trait questionnaire score that has been tested for reliability and validity. Before training, the data is cleaned and augmented. During training, mean squared error is used as the loss function, the Adam optimizer is used for backpropagation, and early stopping is employed to determine the optimal number of training epochs. The trained model can capture the complex, non-linear mapping relationship between fingerprint features and behavioral traits. The model's advantage lies in its end-to-end predictive capability, avoiding the subjectivity and limitations of manually designed association rules.

[0061] In one possible implementation, the education program generation rule engine is implemented as follows: The rule base is stored in JSON format, each rule has a unique ID, and its applicable age range and target keywords are clearly defined. The "IF" part of the rule consists of multiple atomic conditions connected by logical "AND" statements. Each atomic condition defines a score threshold range for a specific educational trait dimension. The "THEN" part is a structured list of actions, with each action including an activity name, resource link, suggested duration, implementation points, and theoretical basis.

[0062] The matching process employs the Rete algorithm for efficient pattern matching. Upon arrival of input, the engine constructs a fact network in memory, rapidly activating all relevant rules. For each activated rule, the engine calculates a comprehensive confidence score based on its weight factor (based on historical valid statistical data) and the degree of match between the current input and the rule conditions. Subsequently, the output actions are merged and sorted to generate the final solution. For example, for a combination of high logical reasoning and moderate language expression, the solution will prioritize activities that promote logical language, such as "programming story creation."

[0063] In one possible implementation, the structured document content of the initial personalized education development plan is generated through a dynamic template. The document header is an abstract that briefly describes the core traits and development direction. The section on strengthening strengths not only lists the activity names but also provides a brief explanation of how the activity stimulates the corresponding neuropsychological functions. Support suggestions for traits requiring attention are more strategic, such as breaking down the task of "improving concentration" into specific steps such as "starting with short-duration, high-frequency training."

[0064] The parent-child interaction section provides dialogue script examples, emotional response strategies, and communication methods to avoid. The learning environment guide suggests the setup of the physical environment (such as reducing visual distractions) and the creation of a positive psychological environment (such as using growth-oriented praise). The key developmental activity timeline is presented in Gantt chart format, clearly indicating the start time, frequency, and responsible person for each activity. The expected developmental milestones are all observable and recordable behavioral descriptions, such as "able to independently complete instructions involving three steps." This document translates abstract principles into a detailed, actionable blueprint for the family.

[0065] In one possible implementation, behavioral performance data is collected through a dual-channel mechanism. Standardized situational task data is pushed to the system weekly, with task design following psychometric principles and having norm references. For example, a spatial imagination task might involve completing a specific multiple-choice question about rotating shapes within a time limit, with the system automatically recording the accuracy rate and reaction time. This task data provides objective and comparable quantitative indicators.

[0066] Natural situation observation data is collected through the structured log function on the parent's end. The system sends a short log reminder every evening. The log template includes preset positive behavior options (e.g., "Shared toys today") and behaviors to be observed (e.g., "Did you give up easily when encountering difficulties?"). Parents can check the boxes and add brief descriptions. The system uses natural language processing technology to extract keywords and perform sentiment analysis on the text descriptions, transforming them into semi-structured data. This dual-channel design balances data objectivity and ecological validity, comprehensively reflecting the individual's developmental status.

[0067] In one possible implementation, the specific implementation method of the effect comparison analysis is as follows: The system first extracts data segments directly related to the execution plan of the current period from the time series of behavioral performance data. For standardized task data, the system calculates the mean score and initial score of each relevant task within the period to determine the improvement rate. For observation record data, the system statistically analyzes the frequency changes of specific behavioral keywords.

[0068] Next, the system compares the calculated progress with the expected progress range set in the initial personalized education development plan. It calculates the deviation from the target range using the percentage deviation method and automatically categorizes it based on preset thresholds (e.g., deviation within ±20% is considered satisfactory). Simultaneously, the system uses time series analysis models (such as ARIMA) to fit the trends of key behavioral indicators, determining whether they are in a significant upward, stable, or downward trend. This analysis process transforms vague subjective perceptions into clear, data-driven charts of results.

[0069] In one possible implementation, the specific logic for generating personalized educational development optimization plans is as follows: For measures deemed "effective" by the analysis, the system marks them as core measures in the optimization plan and may suggest appropriately increasing the frequency or difficulty based on "dosage-effect" data. For "neutral" measures, the system will initiate attribution analysis to check whether the problem is due to insufficient implementation frequency, lack of interest from the child, or inappropriate activity format, and provide an A / B test option, i.e., providing two variations for the next cycle to try.

[0070] For measures marked "to be adjusted," the system doesn't simply delete them. Instead, it searches the rule base for alternative rules with similar educational goals but different activity formats. For example, it might replace "training concentration through reading" with "training concentration by building blocks according to instructions." Before generating a new plan, the system calls a consistency check module to ensure that the new plan doesn't have conflicting activity times or contradictory goals. Finally, combined with the new age tags, the plan automatically incorporates theoretical and activity recommendations that align with the new developmental stage.

[0071] In one possible implementation, the specific method for updating the control database and model is as follows: The data calibration step employs the Bayesian update principle. The initial output evaluation score of the model is considered as the prior probability, and the trait score inferred based on the behavioral data of the current period is considered as the likelihood. The posterior probability is calculated using the Bayesian formula as the corrected score. This calibration method assigns appropriate weights to the new behavioral data.

[0072] New training sample pairs undergo quality verification before being added to the database, eliminating low-quality data resulting from incomplete or contradictory data collection. Upon addition, samples are labeled with their source individual and period. When the accumulated new samples reach a preset threshold (e.g., 1000 pairs), the system automatically initiates incremental training of the model. Incremental training employs transfer learning, fixing the parameters of the first few layers and primarily fine-tuning parameters closer to the output layer, maximizing the retention of original generalization ability while absorbing new knowledge. This mechanism allows the model to evolve and become increasingly accurate over time.

[0073] In one possible implementation, the dynamic adjustment mechanism for the iterative cycle is as follows: the system has a built-in cycle configuration table related to age. For example, for children aged 3-6, the program execution and data collection phase is set to 8 weeks, the evaluation and optimization phase is set to 2 weeks, and the total cycle is 10 weeks. For children aged 7-12, the execution phase is extended to 12 weeks, and the evaluation phase is 3 weeks. The cycle adjustment is based on the stability and rate of change of children's psychological development; the younger the age, the faster the development and changes, and the more frequent the evaluation and adjustment are required.

[0074] At the start of each cycle, the system automatically adjusts the frequency and density of data collection based on the new cycle length and dynamically generates a corresponding time schedule. This non-fixed-cycle design allows the entire support method to intelligently adapt to the rhythms of different developmental stages of an individual, increasing sensitivity when rapid follow-up is needed and providing families with more time to execute when the developmental stage is relatively stable, thus achieving an adaptive match between the method's rhythm and the individual's developmental patterns.

[0075] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0076] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for supporting individual educational development based on innate biological characteristics assessment, characterized in that, Includes the following steps: S100: Collect fingerprint image data of individuals to form a fingerprint image dataset; S200: Process each fingerprint image in the fingerprint image dataset and extract the quantized feature vector of each fingerprint; S300: Concatenate all quantized feature vectors into a comprehensive biometric input vector, and input the comprehensive biometric input vector into a pre-trained biometric-educational trait association model to obtain the educational trait dimension evaluation score vector output by the biometric-educational trait association model; S400: The educational trait dimension assessment score vector, the individual's current age information, and the individual's set stage development goals are input into the education plan generation rule engine. The education plan generation rule engine matches and integrates the preset educational intervention rules to generate a personalized initial education development plan. S500: Implement an initial personalized education development plan and periodically collect individual behavioral performance data during the implementation period; S600: Based on behavioral performance data, conduct a comparative analysis of the effects of the initial personalized education development plan, optimize and generate an optimized personalized education development plan based on the results of the comparative analysis, and update the control database used to train the biometric-educational trait association model by using the comprehensive biometric input vector and the educational trait dimension score corrected based on behavioral performance data as new samples. S700: Repeatedly execute S500 and S600: forming a continuous iterative loop.

2. The individual educational development support method based on innate biological feature evaluation according to claim 1, characterized by, Extracting quantized feature vectors from S200 specifically includes: S210: Perform preprocessing on the fingerprint image, including grayscale conversion, contrast enhancement, and noise filtering, to obtain the enhanced fingerprint image; S220: On the enhanced fingerprint image, the global pattern category of the fingerprint is identified and the core point of the fingerprint is located based on the orientation field and ridge tracking algorithm. The global pattern categories include arch pattern, loop pattern and whorl pattern. S230: Construct a polar coordinate system with the fingerprint core point as the origin, and extract multi-dimensional quantitative features composed of pattern type coding, ridge density features, ridge curvature variation features, and triangulation features in the polar coordinate system. The pattern type coding is assigned a unique value according to the global pattern type. The ridge density feature is obtained by calculating the average number of ridges per unit length on radial lines from the core point at multiple different angles. The ridge curvature variation feature is obtained by calculating the curvature variation rate statistics along the selected ridge trajectory. The triangulation feature is obtained by calculating the relative distance and azimuth angle between the core point and the triangulation point when the fingerprint is a loop or whorl pattern.

3. The individual educational development support method based on innate biological characteristic evaluation according to claim 1, characterized by, The biometric-educational trait association model in S300 is a multi-output regression model based on deep neural networks; The training process of the biometric-educational trait association model is as follows: A comparison database is obtained, which contains a set of comprehensive biometric input vectors of historical individuals, and a set of standardized scores of historical individuals on multiple preset educational trait dimensions obtained through long-term behavioral observation and standardized contextual assessment; The deep neural network is trained under supervision using the set of comprehensive biometric input vectors as input and the set of standardized scores as the expected output, until the model loss function converges, thus obtaining the trained biometric-educational trait association model. The pre-defined educational trait dimensions include logical reasoning tendency, language expression tendency, spatial imagination tendency, interpersonal sensitivity tendency, focus and perseverance tendency, and activity exploration tendency.

4. The individual educational development support method based on innate biological characteristic evaluation according to claim 1, characterized by, The education program generation rule engine in S400 stores education intervention rules represented in the form of "IF-THEN". The "IF" part of each educational intervention rule defines a specific educational trait dimension score combination pattern, age range, and development goal keywords; The "THEN" section of each educational intervention rule defines the specific educational intervention measures, recommended intensity, recommended frequency, and expected goals that match the "IF" section. The process of generating a personalized initial education development plan by the education plan generation rule engine is as follows: Iterate through all education intervention rules, match the rules that meet the definition of the "IF" part based on the education trait dimension assessment score vector, current age information, and stage development goals, integrate the "THEN" part of all successfully matched education intervention rules, resolve conflicts and prioritize them, and finally generate a structured document.

5. A method of individual educational development support based on innate biological feature assessment according to claim 4, characterized in that, The structured document for the initial personalized education development plan includes: a list of recommended activities to strengthen the individual's strengths and educational traits; a list of recommended supportive activities to address the individual's educational traits that require attention; a detailed explanation of parent-child interaction and communication methods that match the individual's educational trait dimension assessment score vector; a guide to creating physical and psychological parameters of the learning environment that are adapted to the individual's cognitive preferences; and a timeline for key development activities and quantifiable expected development milestones within the pre-set future period.

6. The individual educational development support method based on innate biometric evaluation according to claim 1, characterized in that, The behavioral performance data collected periodically in S500 includes standardized situational task data and natural situational observation record data; Standardized situational task data is obtained by individuals periodically completing standardized assessment tasks associated with preset educational trait dimensions. The standardized assessment tasks include logic puzzle tasks, story retelling tasks, graphic space puzzle tasks, and situational response questionnaire tasks. Each standardized assessment task produces a quantitative performance score. Natural situation observation record data is collected by individual educators through a structured electronic observation record form. The electronic observation record form has preset fields for typical behavioral event categories, frequency recording fields, and free description fields related to each educational trait dimension. Educators record their observations in the corresponding fields.

7. The individual educational development support method based on innate biometric evaluation according to claim 1, characterized in that, S600 uses behavioral performance data to conduct a comparative analysis of the effectiveness of initial personalized education development programs, specifically including: S611: Extract behavioral performance data collected during the current assessment period and calculate the actual values ​​of key development indicators associated with each preset educational trait dimension; S612: Compare the actual values ​​of key development indicators with the expected development milestone indicators set in the initial plan for personalized education development, and calculate the deviation from the progress of each key development indicator. S613: Analyze the changing trends of behavioral performance data, identify the behavioral performance data segments corresponding to each specific educational intervention measure in the initial personalized education development plan, and determine whether the behavioral performance data segment shows a positive changing trend, no obvious changing trend, or a negative changing trend within the evaluation period.

8. A method of individual educational development support based on innate biometric assessment according to claim 7, characterized in that, S600 optimizes personalized education development plans based on the results of effect comparison analysis, specifically including: S621: For educational interventions that show a positive trend of change identified in the effect comparison analysis, the recommended implementation intensity and frequency of the educational interventions shall be maintained or increased in the personalized education development optimization program; S622: For educational intervention measures that show no obvious trend of change in the effect comparison analysis, adjust the specific implementation strategy description of the educational intervention measures in the personalized education development optimization plan, or call the alternative equivalent educational intervention measures from the rule base of the education plan generation rule engine to replace them. S623: For educational intervention measures that show a negative trend in the effect comparison analysis, the educational intervention measures shall be suspended or deleted in the personalized education development optimization plan; S624: Based on the natural growth of an individual's age, update the stage-by-stage development goals and trigger the education program generation rule engine to generate a personalized education development optimization plan for the next assessment cycle based on the updated information.

9. The individual educational development support method based on innate biometric evaluation according to claim 1, characterized in that, S600 uses a comprehensive biometric input vector and educational trait dimension scores corrected based on behavioral performance data as new samples to update the control database, specifically including: S631: Based on the behavioral performance data collected during the current assessment period, especially the performance scores of standardized situational task data, the assessment score vector of the educational trait dimension obtained in S300 is calibrated to obtain the corrected educational trait dimension score vector. The calibration method is to take a weighted average of the assessment scores and the inferred scores based on behavioral data. S632: Use the comprehensive biometric input vector of the individual as the input sample and the corrected educational trait dimension score vector as the output label to form a new training sample pair; S633: Add new training sample pairs to the control database; S634: When the cumulative number of newly added training sample pairs in the control database reaches the preset batch update threshold, the updated control database is used to perform incremental training or full retraining of the biometric-educational trait association model.

10. The individual educational development support method based on innate biological characteristic assessment according to claim 1, characterized in that, In S700, continuous iteration cycles are performed at fixed time periods. Each complete iteration cycle includes a scheme execution and data collection phase and an effect evaluation and optimization phase. The duration of the scheme execution and data collection phase is shorter than that of the effect evaluation and optimization phase. The cycle length of the iteration cycle is dynamically adjusted according to the individual's age group. The younger the age, the shorter the cycle length of the iteration cycle.

Images