Evaluation method and system for quality and quality of students

Abstract

The invention provides a student quality attainment evaluation method and system, and the method comprises the steps: obtaining multi-dimensional behavior data, network behavior deviation event data and resource margin data of a student in an online experiment platform, and generating a student capability dynamic representation vector and an information ethical weight enhancement coefficient; generating a task complexity target level based on the capability representation vector, generating an evaluation parameter adjustment instruction in combination with concurrent task load configuration, and triggering an instant feedback action and a directional resource release instruction; and at the end of a learning period, aggregating a resource release instruction and consumption trajectory data, generating a global optimization genome, reconstructing an information ethical weight and a task complexity level, and generating an initialization strategy parameter set of a next evaluation period. According to the invention, the real-time performance and personalized adaptation efficiency of student quality and nutrition evaluation are improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of educational technology, and in particular, to a method and system for evaluating students' quality and accomplishment. Background Art

[0002] With the popularity of online education platforms, the demand for assessment of students' quality and literacy is growing. Quality and literacy not only include skills such as programming and data analysis, but also involve information ethics and network behavior norms. Existing online experimental platforms need to collect students' multi-dimensional behavior data (such as programming task completion efficiency, data analysis error rate) and network behavior deviation event data in real time, and dynamically adjust assessment strategies and resource allocation in combination with teaching resource margins (such as micro-course resource capacity, concurrent task load configuration) to achieve personalized and precise quality and literacy assessment and teaching intervention.

[0003] At present, existing solutions use automated assessment systems based on rule engines to statically evaluate student behavior data through predefined rule sets (such as task completion time thresholds and error rate caps), and generate fixed resource allocation strategies based on resource margin data. This system can achieve automated assessment and resource allocation to a certain extent, but it relies on static rules and lacks dynamic adjustment capabilities.

[0004] The main drawback of the above scheme is that its static rules cannot adapt to dynamic scenarios where students' abilities fluctuate and resource loads change. For example, when students' programming task completion efficiency fluctuates due to changes in task complexity, static rules cannot adjust the evaluation strategy in real time; at the same time, resource allocation strategies cannot be dynamically optimized according to real-time resource margins and task loads, resulting in low resource utilization or uneven allocation. This static rule-driven evaluation system is difficult to meet the real-time and personalized needs of quality and literacy evaluation. Summary of the invention

[0005] The embodiment of the present application provides a method and system for evaluating students' quality and literacy, so as to solve the problems of poor real-time performance and low efficiency of personalized adaptation in the evaluation of students' quality and literacy in the prior art.

[0006] In a first aspect, the present application embodiment provides a method for evaluating students' quality and accomplishment, including: Acquire multi-dimensional behavior data generated by the interaction between students and the online experiment platform, network behavior deviation event data triggered by the online experiment platform, and resource margin data dynamically updated by the online experiment platform in the teaching environment, wherein the multi-dimensional behavior data includes real-time programming task completion efficiency data and data analysis error rate time series data, and the resource margin data includes the currently allocable micro-course resource capacity, micro-course resource consumption trajectory data, and concurrent experiment task load configuration; Based on the real-time programming task completion efficiency data and the data analysis error rate time series data, a student ability dynamic representation vector is generated, and based on the network behavior deviation event data, an information ethics weight reinforcement coefficient is generated; Based on the student ability dynamic characterization vector, a task complexity target level is generated, and the student ability dynamic characterization vector is matched with the concurrent experiment task load configuration, and an evaluation parameter adjustment instruction is generated. Based on the evaluation parameter adjustment instruction, an instant feedback action adapted to the micro-course resource capacity is triggered, and a directional resource delivery instruction is generated; At the end of the semester, the directional resource delivery instructions and the micro-course resource consumption trajectory data are aggregated to generate a global optimization chromosome set. The global optimization chromosome set generates an initialization strategy parameter set for the next evaluation cycle by reconstructing the information ethics weight reinforcement coefficient and the task complexity target level.

[0007] Optionally, the task complexity target level is generated based on the student ability dynamic characterization vector, and the student ability dynamic characterization vector is matched with the concurrent experiment task load configuration, and an evaluation parameter adjustment instruction is generated. Based on the evaluation parameter adjustment instruction, an instant feedback action adapted to the micro-course resource capacity is triggered, and a directional resource delivery instruction is generated, including: Performing task response delay slicing processing on the student ability dynamic representation vector, extracting the code submission interval sequence in the continuous time window of the real-time programming task completion efficiency data, and generating a task response delay feature vector according to the fluctuation amplitude and trend direction of the code submission interval sequence; Performing a logical fault cluster analysis on the student ability dynamic representation vector, identifying the error pattern type and distribution density of repeated correction operations in the data analysis error rate time series data, and generating a logical fault feature vector based on the error pattern type and distribution density; Performing spatiotemporal alignment and fusion of the task response delay feature vector and the logic fault feature vector to generate a task complexity target level; Calculate the dynamic matching degree between the dynamic characterization vector of the student ability and the concurrent experiment task load configuration, generate a task load adaptation index, and generate an evaluation parameter adjustment instruction based on the task load adaptation index and the micro-course resource capacity, wherein the evaluation parameter adjustment instruction includes the priority increase range of the information ethics weight reinforcement coefficient, the demotion step of the task complexity target level, and the triggering condition of the instant feedback action; According to the trigger condition, the instant feedback action is mapped to the priority queue corresponding to the micro-course resource capacity, and a directional resource delivery instruction is generated.

[0008] Optionally, the step of performing spatiotemporal alignment and fusion of the task response delay feature vector and the logic fault feature vector to generate a target level of task complexity includes: Performing time window segmentation on the task response delay feature vector, extracting the task response delay fluctuation rate in each time window, so as to generate a time dimension delay distribution matrix; Performing logical path mapping on the logical fault feature vector, identifying the association between the repeated correction operation and the logical nodes in the logical path, so as to generate a logical dimension fault topology map; Cross-dimensionally aligning the task response delay fluctuation rate in the time dimension delay distribution matrix with the association relationship in the logic dimension fault topology map; Based on the matching degree between the aligned task response delay fluctuation rate and the correlation relationship, the basic level of task complexity is calculated, and a dynamic correction coefficient is generated in combination with the distribution density in the logical dimension fault topology map, wherein the weight of the dynamic correction coefficient is obtained by dynamically adjusting the resource occupancy rate in the concurrent experimental task load configuration; The task complexity basic level and the dynamic correction coefficient are weighted and superimposed to generate a task complexity target level.

[0009] Optionally, the calculating of the dynamic matching degree between the dynamic characterization vector of the student ability and the concurrent experiment task load configuration, generating a task load adaptation index, and generating an evaluation parameter adjustment instruction based on the task load adaptation index and the micro-course resource capacity, includes: The queue waiting time and resource occupancy rate in the concurrent experimental task load configuration are normalized to generate a load pressure index, wherein the queue waiting time is obtained by dynamically correcting the priority weight of the experimental task type, and the priority weight of the experimental task type is dynamically generated based on the basic level of task complexity and the dynamic correction coefficient; Extracting the task response delay fluctuation rate and the logic fault distribution density in the student ability dynamic representation vector, and mapping them to the time dimension delay distribution matrix and the logic dimension fault topology map, respectively, to obtain the fluctuation rate characteristic scalar and the distribution density characteristic scalar; Performing weighted fusion of the volatility characteristic scalar and the distribution density characteristic scalar to generate a comprehensive capability evaluation value; Based on the matching deviation between the comprehensive capability evaluation value and the load pressure index, a dynamic matching degree is calculated, and based on the dynamic matching degree and the remaining available ratio of the micro-course resource capacity, a task load adaptation index is generated; According to the preset resource adaptation interval in which the task load adaptation index is located, the priority enhancement range of the information ethics weight reinforcement coefficient and the degradation step of the task complexity target level are generated, and the triggering condition of the instant feedback action is generated according to the task load adaptation index and the remaining available ratio of the micro-course resource capacity, wherein the preset resource adaptation interval is dynamically divided based on the ratio of the concurrent experiment task queue length in the micro-course resource capacity to the number of remaining micro-course resources; The priority enhancement range and the downgrade step size are bound to the triggering condition of the instant feedback action to generate an evaluation parameter adjustment instruction.

[0010] Optionally, the dynamic matching degree is calculated based on the matching deviation between the comprehensive capability evaluation value and the load pressure index, and the task load adaptation index is generated based on the dynamic matching degree and the remaining available ratio of the micro-course resource capacity, including: Normalizing the comprehensive ability evaluation value and the load pressure index respectively to generate a standardized ability evaluation value and a standardized load pressure value, wherein the normalization range of the standardized ability evaluation value is dynamically adjusted based on the historical extreme value of the student ability dynamic representation vector; Calculating a matching deviation between the standardized capability evaluation value and the standardized load pressure value, and calculating a dynamic matching degree based on a difference between a reciprocal of the matching deviation and a preset deviation tolerance threshold; Generate a resource capacity adaptation coefficient according to the remaining available ratio of the micro-course resource capacity, wherein the resource capacity adaptation coefficient is positively correlated with the square root value of the remaining available ratio and negatively correlated with the logarithmic decay rate of the concurrent experiment task queue length; The dynamic matching degree and the resource capacity adaptation coefficient are weighted and integrated to generate a task load adaptation index.

[0011] Optionally, generating a student ability dynamic representation vector based on the real-time programming task completion efficiency data and the data analysis error rate time series data, and generating an information ethics weight reinforcement coefficient according to the network behavior deviation event data, includes: Performing time window segmentation on the code submission interval sequence in the real-time programming task completion efficiency data to generate a time dimension volatility vector; Performing fault clustering analysis on the logic jump paths in the data analysis error rate time series data to generate a logic fault density vector; Cross-modally aligning the time dimension volatility vector with the logic fault density vector to generate a dynamic representation vector of student ability; Perform multi-level classification coding on the illegal operation types in the network behavior deviation event data, generate illegal type classification labels, count the triggering frequency of the illegal type classification labels within a preset time period, and generate an illegal frequency distribution matrix; An information ethics weight reinforcement coefficient is generated based on the proportion of high-frequency violation types and the time decay factor in the violation frequency distribution matrix, wherein the time decay factor is dynamically calculated based on the time series interval of the network behavior deviation event data.

[0012] Optionally, the aggregating the directional resource delivery instruction and the micro-course resource consumption trajectory data to generate a global optimization chromosome set includes: Multi-dimensionally encode the priority queue sorting rules in the directional resource delivery instruction, and extract the resource type label and delivery frequency distribution characteristics corresponding to the micro-course resource capacity in the priority queue sorting rules; Performing periodic cluster analysis on the resource consumption rate and time interval fluctuations in the micro-course resource consumption trajectory data to identify a switching threshold for distinguishing between a high-frequency consumption period and a low-frequency consumption period; The frequency distribution characteristics of the delivery are mapped with the switching threshold in a spatiotemporal manner to generate a resource collaborative adaptation chromosome fragment, the constraint range of the information ethics weight reinforcement coefficient is extracted with boundary features to generate an ethics weight constraint gene fragment, and the benchmark mapping rule of the task complexity target level is quantified with volatility to generate a complexity mapping gene fragment; The ethical weight constraint gene segment and the complexity mapping gene segment are cross-recombined to generate a globally optimized chromosome set, wherein the node position of the cross-recombination is dynamically selected according to the spatiotemporal correlation strength in the resource collaborative adaptation chromosome segment.

[0013] In a second aspect, the present application embodiment provides a student quality assessment system, including: An acquisition module is used to acquire multi-dimensional behavior data generated by the interaction between students and the online experiment platform, network behavior deviation event data triggered by the online experiment platform, and resource margin data dynamically updated by the online experiment platform in the teaching environment, wherein the multi-dimensional behavior data includes real-time programming task completion efficiency data and data analysis error rate time series data, and the resource margin data includes the currently allocable micro-course resource capacity, micro-course resource consumption trajectory data, and concurrent experiment task load configuration; A generation module, for generating a dynamic representation vector of student ability based on the real-time programming task completion efficiency data and the data analysis error rate time series data, and generating an information ethics weight reinforcement coefficient based on the network behavior deviation event data; A matching module is used to generate a task complexity target level based on the student ability dynamic characterization vector, match the student ability dynamic characterization vector with the concurrent experiment task load configuration, generate an evaluation parameter adjustment instruction, trigger an instant feedback action adapted to the micro-course resource capacity based on the evaluation parameter adjustment instruction, and generate a directional resource delivery instruction; The reconstruction module is used to aggregate the targeted resource delivery instructions and the micro-course resource consumption trajectory data at the end of the semester to generate a global optimization chromosome set. The global optimization chromosome set generates an initialization strategy parameter set for the next evaluation cycle by reconstructing the information ethics weight reinforcement coefficient and the task complexity target level.

[0014] In a third aspect, an embodiment of the present application provides a computing device, including a processor and a memory, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute a method for assessing student quality and literacy as described in any one of the first aspects.

[0015] In a fourth aspect, an embodiment of the present application provides a computer storage medium having computer program instructions stored thereon, wherein the computer program instructions, when executed by a processor, implement a method for assessing student quality and literacy as described in any one of the first aspects.

[0016] In an embodiment of the present application, a method for evaluating students' quality and accomplishment is provided, the method comprising: obtaining multi-dimensional behavior data generated by the interaction between students and an online experiment platform, network behavior deviation event data triggered by the online experiment platform, and resource margin data dynamically updated by the online experiment platform in a teaching environment, the multi-dimensional behavior data comprising real-time programming task completion efficiency data and data analysis error rate time series data, the resource margin data comprising the currently allocable micro-class resource capacity, micro-class resource consumption trajectory data, and concurrent experiment task load configuration; based on the real-time programming task completion efficiency data and the data analysis error rate time series data, generating a dynamic characterization vector of student ability, and according to the network behavior deviation event data, generating a dynamic characterization vector of student ability, and ... Generate an information ethics weight reinforcement coefficient for the deviation event data; generate a task complexity target level based on the student ability dynamic representation vector, and match the student ability dynamic representation vector with the concurrent experimental task load configuration to generate an assessment parameter adjustment instruction; based on the assessment parameter adjustment instruction, trigger an instant feedback action adapted to the micro-course resource capacity to generate a directional resource delivery instruction; at the end of the semester, aggregate the directional resource delivery instruction and the micro-course resource consumption trajectory data to generate a global optimization chromosome set, and the global optimization chromosome set generates an initialization strategy parameter set for the next assessment cycle by reconstructing the information ethics weight reinforcement coefficient and the task complexity target level.

[0017] The technical solution of this application has the following beneficial effects: This application comprehensively collects students' behavioral data and environmental data in the online experimental platform to provide a data basis for subsequent dynamic evaluation and resource allocation. Generate capability representation vectors through real-time programming task completion efficiency and data analysis error rate, and generate ethical weight reinforcement coefficients in combination with network behavior deviation events to achieve dynamic evaluation of capability and ethics in two dimensions. Generate task complexity target levels based on capability representation vectors and concurrent task load configuration, and generate evaluation parameter adjustment instructions in combination with micro-course resource capacity to achieve dynamic adaptation of task difficulty and resource allocation. Trigger instant feedback actions according to evaluation parameter adjustment instructions, and generate targeted resource delivery instructions in combination with micro-course resource capacity to ensure the accuracy and timeliness of resource allocation. Aggregate resource delivery instructions and consumption trajectory data at the end of the semester to generate a globally optimized chromosome group, reconstruct information ethics weights and task complexity levels, generate an initialization strategy parameter set for the next evaluation cycle, and achieve global optimization and iterative updates of evaluation strategies.

[0018] Furthermore, the embodiment of the present application also performs task response delay slicing and logic fault clustering analysis on the dynamic characterization vector of student capabilities, extracts the fluctuation characteristics and error pattern distribution density of the code submission interval sequence, and generates a task response delay feature vector and a logic fault feature vector; the two are aligned and fused in time and space to generate a target level of task complexity; a task load adaptation index is generated based on the dynamic matching degree between the student capability characterization vector and the concurrent task load configuration, and an evaluation parameter adjustment instruction is generated in combination with the micro-course resource capacity; according to the trigger conditions, the instant feedback action is mapped to the priority queue of the micro-course resource capacity to generate a targeted resource delivery instruction.

[0019] Through multi-dimensional feature extraction and spatiotemporal alignment fusion, the dynamic generation and precise adaptation of task complexity target levels are achieved; based on the coordinated optimization of task load adaptation index and resource capacity, evaluation parameter adjustment instructions and targeted resource allocation instructions are generated, which improves the real-time performance of the evaluation strategy and the accuracy of resource allocation, while ensuring the timeliness and pertinence of instant feedback actions.

[0020] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0022] Figure 1A flow chart of a method for evaluating student quality and accomplishment provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of a student quality assessment system provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0024] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0025] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0026] This application constructs a dynamic representation vector of student capabilities and an information ethics weight reinforcement coefficient by collecting students' multi-dimensional behavior data (such as programming task completion efficiency, data analysis error rate), network behavior deviation event data, and teaching resource surplus data (such as micro-course resource capacity, concurrent task load configuration) in real time on the online experimental platform; generates a task complexity target level based on the capability representation vector, and generates assessment parameter adjustment instructions in combination with the concurrent task load configuration to trigger instant feedback actions and directional resource delivery instructions adapted to the micro-course resource capacity; aggregates resource delivery instructions and consumption trajectory data at the end of the semester to generate a globally optimized chromosome group, reconstruct information ethics weights and task complexity levels, and generate an initialization strategy parameter set for the next assessment cycle, thereby achieving real-time, personalized, and global optimization of student quality and literacy assessment.

[0027] Figure 1A flow chart of a method for evaluating the quality of students is provided in the embodiment of the present application, such as Figure 1 As shown, the method includes: Step 101: Acquire multi-dimensional behavior data generated by the interaction between students and an online experiment platform, network behavior deviation event data triggered by the online experiment platform, and resource margin data dynamically updated by the online experiment platform in a teaching environment.

[0028] In this step, the multi-dimensional behavior data includes real-time programming task completion efficiency data and data analysis error rate time series data. Resource margin data includes the current allocatable micro-course resource capacity, micro-course resource consumption trajectory data, and concurrent experimental task load configuration. Network behavior deviation event data refers to abnormal student network behavior events (such as data tampering, illegal access) monitored by the online experimental platform. The allocatable micro-course resource capacity refers to the number of micro-course resources currently available for allocation to students in the online experimental platform. These resources include but are not limited to programming micro-courses, data analysis cases, information ethics education videos, etc., and their capacity is limited by the total number of resources on the platform and the number of currently allocated resources. For example, in the programming experiment class, the platform has a total of 100 programming micro-course resources, 50 of which are currently allocated to other students, and the remaining 50 are the "allocatable micro-course resource capacity". If student A needs additional programming micro-course resources, the platform can allocate them from these 50 remaining resources.

[0029] In the embodiment of the present application, the log system of the online experiment platform is used to collect students' programming task completion efficiency data and data analysis error rate time series data in real time, and the network behavior deviation event data is obtained in combination with the platform monitoring module; at the same time, the micro-course resource capacity, consumption trajectory data and concurrent experiment task load configuration are extracted from the teaching resource management system. After cleaning and standardization, these data are stored in the central database to provide a basis for subsequent analysis.

[0030] For example, in a programming experiment class, when students complete a programming task, the platform records their code submission interval and task response time; in a data analysis experiment, the platform records the error type and the time it takes to correct it. At the same time, the platform detects that students attempt to illegally access other people's data, triggering a network behavior deviation event. The teaching resource management system shows that there are currently 50 remaining micro-class resources and the length of the concurrent experiment task queue is 20.

[0031] Step 102: Generate a student ability dynamic representation vector based on the real-time programming task completion efficiency data and the data analysis error rate time series data, and generate an information ethics weight reinforcement coefficient based on the network behavior deviation event data.

[0032] In this step, the student ability dynamic representation vector is used to dynamically represent the student's programming and data analysis capabilities. The information ethics weight reinforcement coefficient is a dynamically adjusted parameter used to quantify the importance of the information ethics dimension in the assessment of student quality and literacy. This coefficient is generated based on the student's network behavior deviation event data. The more deviation events there are, the higher the coefficient, and the greater the weight of the information ethics dimension in the assessment. For example, if student A triggers three data tampering events on the experimental platform, the platform will generate an information ethics weight reinforcement coefficient (such as 0.8) based on the event type and frequency, and increase the weight of the information ethics dimension in the assessment to correct his network behavior deviation.

[0033] In the embodiment of the present application, the code submission interval sequence in the real-time programming task completion efficiency data is segmented into time windows to extract the task response delay volatility and trend direction; the data analysis error rate time series data is subjected to logical fault clustering analysis to extract the error pattern type and distribution density; the two are combined to generate a dynamic representation vector of student ability. At the same time, the network behavior deviation event data is multi-level classified and encoded, the frequency of violation types is counted, and the information ethics weight reinforcement coefficient is generated.

[0034] For example, the interval sequence of students' code submissions shows that the volatility of their task response delay is 0.2, the data analysis error pattern is mainly the missing logic loop, the distribution density is 0.15, and the generated ability representation vector is [0.2, 0.15]. The platform statistics show that the frequency of data tampering in its network behavior deviation events is 3 times, and the generated information ethics weight reinforcement coefficient is 0.8.

[0035] Step 103: Based on the student ability dynamic characterization vector, generate a task complexity target level, match the student ability dynamic characterization vector with the concurrent experiment task load configuration, generate an assessment parameter adjustment instruction, and based on the assessment parameter adjustment instruction, trigger an instant feedback action adapted to the micro-course resource capacity, and generate a targeted resource delivery instruction.

[0036] In this step, the target level of task complexity includes basic complexity and dynamic correction coefficient. The evaluation parameter adjustment instructions include the information ethics weight priority increase range, task complexity reduction step and instant feedback trigger conditions. Adaptive instant feedback action refers to the feedback action dynamically triggered according to the current micro-course resource capacity and student needs. The adaptability is reflected in the matching of the type, quantity and priority of the feedback action with the resource capacity to ensure the accuracy and efficiency of resource allocation. For example, if student A needs programming micro-course resources, and the current allocable micro-course resource capacity is 10, the platform will trigger the adaptive instant feedback action (such as pushing 2 programming micro-courses) according to its task complexity and ability level, instead of allocating too many resources at one time and causing resource waste. Targeted resource delivery instructions are used to guide the platform to deliver specific types and quantities of micro-course resources to specific students, ensure the accuracy and timeliness of resource allocation, and meet students' personalized learning needs, including resource types (such as programming micro-courses, data analysis cases), resource quantity, delivery priority, and trigger conditions.

[0037] In an embodiment of the present application, the task response delay volatility and the logic fault distribution density in the dynamic characterization vector of the student's ability are aligned and fused in time and space to generate a target level of task complexity; the task load adaptation index is generated based on the dynamic matching degree between the student's ability characterization vector and the concurrent experimental task load configuration; and the evaluation parameter adjustment instructions are generated in combination with the micro-course resource capacity to trigger instant feedback actions and generate targeted resource delivery instructions.

[0038] For example, after the student's ability representation vector [0.2, 0.15] matches the concurrent task load configuration, the generated task complexity target level is intermediate (basic complexity) + 0.1 (dynamic correction coefficient). Based on its task load adaptation index and the remaining micro-course resource capacity, the platform generates evaluation parameter adjustment instructions: the information ethics weight is increased by 0.1, the task complexity is reduced by 0.05, and the instant feedback action is triggered (pushing programming micro-courses).

[0039] Step 104: At the end of the semester, aggregate the directional resource delivery instructions and the micro-course resource consumption trajectory data to generate a global optimization chromosome set. The global optimization chromosome set generates an initialization strategy parameter set for the next evaluation cycle by reconstructing the information ethics weight reinforcement coefficient and the task complexity target level.

[0040] In this step, the initialization strategy parameter set for the next evaluation cycle includes core parameters: constraint range parameters of the information ethics weight reinforcement coefficient (weight lower limit, upper limit, adjustment step, attenuation factor), benchmark mapping rule parameters of the task complexity target level (basic complexity level, dynamic correction coefficient, adjustment threshold, mapping function), resource allocation strategy parameters (priority queue sorting rules, delivery frequency threshold, capacity adaptation interval, attenuation factor), trigger condition parameters of immediate feedback action (trigger threshold, feedback action priority, attenuation factor), and global optimization chromosome group gene sorting rule parameters (crossover node selection rule, recombination weight, update frequency). The global optimization chromosome group is used to reconstruct the evaluation strategy. Through the crossover recombination and evolution mechanism of the genetic algorithm, the initialization strategy parameter set for the next evaluation cycle is optimized to improve the global adaptability and efficiency of the evaluation strategy, including the information ethics weight constraint range gene fragment, the task complexity mapping rule gene fragment, and the resource allocation strategy gene fragment.

[0041] In an embodiment of the present application, targeted resource delivery instructions and micro-course resource consumption trajectory data are aggregated at the end of the semester to extract resource delivery patterns and consumption fluctuation characteristics; the constraint range of the information ethics weight reinforcement coefficient and the benchmark mapping rules of the task complexity target level are encoded as chromosome fragments, and a globally optimized chromosome group is generated through cross-recombination; and the initialization strategy parameter set for the next evaluation cycle is reconstructed based on the chromosome group.

[0042] For example, at the end of the semester, the platform counts the students' resource allocation instruction execution results and micro-course consumption trajectories to generate a global optimization chromosome set; through the chromosome set, the information ethics weight constraint range is reconstructed as [0.7, 1.0], and the task complexity mapping rule is "ability representation vector [0.2, 0.15]→intermediate+0.1", generating the initialization strategy parameter set for the next evaluation cycle.

[0043] This application dynamically generates student ability representation vectors and information ethics weight reinforcement coefficients by collecting student behavior data and environmental data in real time; generates task complexity target levels and assessment parameter adjustment instructions based on the ability representation vectors and task load configuration, triggering immediate feedback actions and targeted resource delivery instructions; at the end of the semester, the assessment strategy is reconstructed through global optimization of the chromosome group to generate the initialization parameter set for the next cycle, thereby achieving real-time, personalization and global optimization of student quality assessment, while improving the efficiency and accuracy of dynamic allocation of teaching resources.

[0044] In order to solve the problem of dynamic adaptation of task complexity and precise allocation of resources in the assessment of students' quality and literacy, this application performs task response delay slicing on the dynamic representation vector of students' abilities, extracts the code submission interval sequence in the real-time programming task completion efficiency data, analyzes its fluctuation amplitude and trend direction, and generates a task response delay feature vector; at the same time, the error pattern type and distribution density in the data analysis error rate time series data are identified through logical fault clustering analysis to generate a logical fault feature vector; the task response delay feature vector and the logical fault feature vector are aligned and fused in time and space to generate a target level of task complexity; the task load adaptation index is generated based on the dynamic matching degree between the student ability representation vector and the concurrent experimental task load configuration, and the assessment parameter adjustment instructions (including the information ethics weight priority increase amplitude, the task complexity degradation step and the instant feedback trigger condition) are generated in combination with the micro-course resource capacity; the instant feedback action is mapped to the priority queue of the micro-course resource capacity according to the trigger condition, and a directional resource delivery instruction is generated, thereby realizing the coordinated optimization of dynamic adaptation of task complexity, precise resource allocation and instant feedback.

[0045] In some embodiments, the step 103 generates a target level of task complexity based on the student ability dynamic characterization vector, matches the student ability dynamic characterization vector with the concurrent experiment task load configuration, generates an evaluation parameter adjustment instruction, triggers an instant feedback action adapted to the micro-course resource capacity based on the evaluation parameter adjustment instruction, and generates a directional resource delivery instruction, including: Step 201: Perform task response delay segmentation processing on the student ability dynamic characterization vector, extract the code submission interval sequence in the continuous time window of the real-time programming task completion efficiency data, and generate a task response delay feature vector according to the fluctuation amplitude and trend direction of the code submission interval sequence.

[0046] In step 201, the task response delay feature vector is used to characterize the response delay fluctuation characteristics of students in programming tasks. The code submission interval sequence in a continuous time window refers to the time interval sequence of students submitting codes in a continuous time period (such as a time window of 5 minutes) in a programming task. For example, the time interval for student A to submit code in time window 1 is [10 seconds, 15 seconds, 20 seconds], and in time window 2 it is [12 seconds, 18 seconds, 22 seconds]. The task response delay feature vector is used to characterize the response delay characteristics of students in programming tasks. For example, the vector [0.2, -0.1] indicates that the task response delay fluctuation rate is 0.2, and the trend direction is negative (i.e., the delay gradually decreases).

[0047] In the embodiment of the present application, the student ability dynamic characterization vector is subjected to task response delay segmentation processing, and the code submission interval sequence in the continuous time window of the real-time programming task completion efficiency data is extracted; according to the fluctuation amplitude and trend direction of the code submission interval sequence, a task response delay feature vector is generated. Specifically, the code submission interval sequence is extracted by time window segmentation, and its fluctuation amplitude (such as standard deviation) and trend direction (such as linear regression slope) are calculated to generate a task response delay feature vector.

[0048] Step 202: Perform logical fault clustering analysis on the student ability dynamic representation vector to identify the error pattern type and distribution density of repeated correction operations in the data analysis error rate time series data, and generate a logical fault feature vector based on the error pattern type and distribution density.

[0049] In step 202, the error pattern type of repeated correction operation refers to the error type that students repeatedly make in data analysis tasks, for example, the missing logic loop is caused by the incorrect use of the loop structure, resulting in incomplete data processing; redundant conditional judgment is caused by the use of unnecessary conditional judgment, resulting in complicated logic; data format error is caused by incorrect processing of the data format, resulting in analysis failure. Distribution density refers to the distribution frequency of error patterns in time or tasks. For example, the missing logic loop occurs 3 times in 10 correction operations, and its distribution density is 0.3. The logical fault feature vector is used to characterize the logical fault characteristics of students in data analysis tasks. For example, the vector [0.3, 0.2] indicates that the distribution density of the missing logic loop is 0.3, and the distribution density of the redundant conditional judgment is 0.2.

[0050] In the embodiment of the present application, a logical fault clustering analysis is performed on the student ability dynamic characterization vector to identify the error pattern type (such as missing logic loops, redundant conditional judgments) and distribution density (such as error frequency and time distribution) of repeated correction operations in the data analysis error rate time series data; a logical fault feature vector is generated based on the error pattern type and distribution density. Specifically, the error pattern type is identified through a clustering algorithm, and its distribution density is counted to generate a logical fault feature vector.

[0051] Step 203: perform spatiotemporal alignment and fusion of the task response delay feature vector and the logic fault feature vector to generate a target level of task complexity.

[0052] In step 203, the task complexity target level includes a basic complexity level and a dynamic correction coefficient.

[0053] In the embodiment of the present application, the task response delay feature vector and the logic fault feature vector are temporally and spatially aligned and fused to generate a task complexity target level. Specifically, the response delay feature of the programming task and the logic fault feature of the data analysis are fused through the temporal and spatial alignment technology to generate a task complexity target level including a basic complexity level and a dynamic correction coefficient.

[0054] Step 204: Calculate the dynamic matching degree between the student ability dynamic characterization vector and the concurrent experiment task load configuration, generate a task load adaptation index, and generate an evaluation parameter adjustment instruction based on the task load adaptation index and the micro-course resource capacity.

[0055] In step 204, the evaluation parameter adjustment instruction includes the priority increase range of the information ethics weight reinforcement coefficient, the downgrade step size of the task complexity target level, and the triggering condition of the immediate feedback action. The task load adaptation index is used to quantify the adaptation relationship between student ability and system load. For example, an index of 0.8 indicates that the student ability is highly adapted to the system load, and adaptation refers to the degree of match between the student ability and the system load. For example, when the student ability is stronger and the system load is lower, the adaptation index is higher; conversely, when the student ability is weaker and the system load is higher, the adaptation index is lower. The triggering conditions of the immediate feedback action include the task complexity exceeding the threshold, the resource capacity being lower than the critical value, and the information ethics weight reinforcement coefficient exceeding the upper limit.

[0056] In the embodiment of the present application, the dynamic matching degree between the dynamic characterization vector of the student ability and the concurrent experiment task load configuration is calculated to generate a task load adaptation index; based on the task load adaptation index and the micro-course resource capacity, an evaluation parameter adjustment instruction is generated. Specifically, the task load adaptation index is generated by matching calculation, and the evaluation parameter adjustment instruction is generated in combination with the resource capacity.

[0057] Step 205: According to the trigger condition, the instant feedback action is mapped to the priority queue corresponding to the micro-course resource capacity, and a directional resource delivery instruction is generated.

[0058] In step 205, the priority queue refers to a queue structure that prioritizes instant feedback actions (such as pushing micro-course resources) according to certain sorting rules.

[0059] In the embodiment of the present application, according to the trigger condition, the instant feedback action is mapped to the priority queue corresponding to the micro-course resource capacity to generate a directional resource delivery instruction. Specifically, the instant feedback action is matched with the resource capacity through the priority queue sorting rule to generate a directional resource delivery instruction.

[0060] Here is a specific example: In the programming experiment class, when student A completed the programming task, the platform recorded the interval sequence of his code submission, extracted the fluctuation amplitude of 0.2, the trend direction was negative, and generated the task response delay feature vector [0.2,-0.1]; in the data analysis experiment, the platform recorded the error mode type as the missing logic loop, the distribution density was 0.15, and generated the logic fault feature vector [0.15]. The two were aligned and fused in time and space, and the target level of task complexity was generated as intermediate (basic complexity) + 0.1 (dynamic correction coefficient). The platform calculated the dynamic matching degree between student A's ability representation vector and the concurrent task load configuration, and generated the task load adaptation index of 0.8; based on this index and the remaining micro-course resource capacity, the evaluation parameter adjustment instruction was generated: the information ethics weight was increased by 0.1, the task complexity was reduced by 0.05, and the instant feedback action (pushing programming micro-courses) was triggered. According to the triggering conditions, the platform mapped the instant feedback action to the priority queue of the micro-course resource capacity, and generated a directional resource delivery instruction: push 2 programming micro-courses with high priority.

[0061] Through the generation and fusion of task response delay feature vectors and logical fault feature vectors, dynamic adaptation of the target level of task complexity is achieved; based on the coordinated optimization of task load adaptation index and micro-course resource capacity, assessment parameter adjustment instructions and targeted resource allocation instructions are generated, which improves the real-time performance of the assessment strategy and the accuracy of resource allocation, while ensuring the timeliness and pertinence of instant feedback actions, thereby achieving personalization and efficiency in the assessment of students' quality and literacy.

[0062] In order to further improve the accuracy and dynamic adaptability of task complexity calculation, the present application divides the task response delay feature vector into time windows, extracts the task response delay volatility in each time window, and generates a time dimension delay distribution matrix; at the same time, performs logical path mapping on the logical fault feature vector, identifies the association between repeated correction operations and logical nodes, and generates a logical dimension fault topology map; cross-dimensionally aligns the task response delay volatility in the time dimension delay distribution matrix with the association in the logical dimension fault topology map; calculates the basic level of task complexity based on the matching degree of the aligned volatility and the association, and generates a dynamic correction coefficient in combination with the distribution density in the logical dimension fault topology map, and its weight is dynamically adjusted according to the resource occupancy rate in the concurrent experimental task load configuration; finally, the basic level of task complexity and the dynamic correction coefficient are weighted and superimposed to generate a target level of task complexity, thereby realizing dynamic adaptation and precise adjustment of task complexity.

[0063] In some embodiments, the step 203 of performing spatiotemporal alignment and fusion of the task response delay feature vector and the logic fault feature vector to generate a task complexity target level includes: Step 301: segment the task response delay feature vector into time windows, extract the task response delay fluctuation rate in each time window, and generate a time dimension delay distribution matrix.

[0064] In step 301, the task response delay volatility refers to the degree of fluctuation of the code submission interval of the student in the programming task, which is usually calculated by statistical methods (such as standard deviation). For example, a volatility of 0.2 means that the fluctuation of the code submission interval is large and the task response time is unstable. The time dimension delay distribution matrix is ​​used to characterize the task response delay volatility in different time windows. For example, the matrix [0.2, 0.15] indicates that the volatility in time window 1 is 0.2 and the volatility in time window 2 is 0.15.

[0065] In an embodiment of the present application, the task response delay feature vector is segmented into time windows, and the task response delay volatility in each time window is extracted to generate a time dimension delay distribution matrix. Specifically, the code submission interval sequence is divided into continuous time windows (such as a window every 5 minutes) through a time window segmentation technique (such as a sliding window method), and the task response delay volatility (such as a standard deviation) is calculated in each window; the volatility of each window is arranged in chronological order to generate a time dimension delay distribution matrix. For example, if the volatility of time window 1 is 0.2 and the volatility of time window 2 is 0.15, then the matrix [0.2, 0.15] is generated.

[0066] Step 302: Perform logical path mapping on the logical fault feature vector, identify the association between the repeated correction operation and the logical nodes in the logical path, so as to generate a logical dimension fault topology map.

[0067] In step 302, the logical dimension fault topology graph is used to characterize the association between the repeated correction operation and the logical node. For example, the nodes in the graph represent the logical operation (such as loop, conditional judgment), and the edges represent the association between the error mode and the logical node (such as the association between the missing logical loop and the loop node).

[0068] In an embodiment of the present application, the logical fault feature vector is mapped to a logical path, and the association between the repeated correction operation and the logical node in the logical path is identified to generate a logical dimension fault topology map. Specifically, the error pattern type (such as logical loop missing, conditional judgment redundancy) is associated with the logical node (such as loop node, conditional node) through the graph structure modeling technology (such as directed acyclic graph) to generate a logical dimension fault topology map. For example, if the logical loop missing is associated with the loop node, and the conditional judgment redundancy is associated with the conditional node, then a topology map containing these nodes and edges is generated.

[0069] Step 303: cross-dimensionally align the task response delay fluctuation rate in the time dimension delay distribution matrix with the association relationship in the logic dimension fault topology map.

[0070] In step 303, cross-dimensional alignment refers to matching and fusing the task response delay fluctuation rate in the time dimension delay distribution matrix with the association relationship in the logic dimension fault topology map.

[0071] In an embodiment of the present application, the task response delay volatility in the time dimension delay distribution matrix is ​​aligned across dimensions with the association relationship in the logic dimension fault topology map. Specifically, the response delay characteristics of the programming task are fused with the logical fault characteristics of the data analysis through time-space alignment technology (such as dynamic time warping) to generate aligned data. For example, the volatility of time window 1 of 0.2 is aligned with the association relationship of the missing logic loop, and the volatility of time window 2 of 0.15 is aligned with the association relationship of the conditional judgment redundancy.

[0072] Step 304: Based on the matching degree between the aligned task response delay fluctuation rate and the association relationship, the basic level of task complexity is calculated, and a dynamic correction coefficient is generated in combination with the distribution density in the logical dimension fault topology diagram, wherein the weight of the dynamic correction coefficient is obtained by dynamically adjusting the resource occupancy rate in the concurrent experimental task load configuration.

[0073] In step 304, the basic level of task complexity is a static task difficulty level generated based on the aligned task response delay volatility and the association relationship, which is used to characterize the inherent complexity of the task. For example, a basic level of intermediate indicates that the inherent difficulty of the task is moderate. The basic level of task complexity includes a basic complexity level and an association relationship matching degree. The dynamic correction coefficient is used to correct the task complexity in real time. For example, a dynamic correction coefficient of 0.1 indicates that the task complexity needs to be adjusted upward by 10%.

[0074] In the embodiment of the present application, the basic level of task complexity is calculated based on the matching degree between the aligned task response delay fluctuation rate and the association relationship, and a dynamic correction coefficient is generated in combination with the distribution density in the logical dimension fault topology diagram, wherein the weight of the dynamic correction coefficient is obtained by dynamically adjusting the resource occupancy rate in the concurrent experimental task load configuration. Specifically, the basic level of task complexity (such as intermediate) is generated by matching degree calculation (such as cosine similarity), and a dynamic correction coefficient (such as 0.1) is generated by distribution density (such as the distribution density of logical loop missing is 0.3) and resource occupancy rate (such as 70%).

[0075] Step 305: Perform weighted superposition of the task complexity basic level and the dynamic correction coefficient to generate a task complexity target level.

[0076] In the embodiment of the present application, the task complexity base level and the dynamic correction coefficient are weighted and superimposed to generate the task complexity target level. Specifically, the base level (such as intermediate) and the dynamic correction coefficient (such as 0.1) are combined through the weighted superposition technology to generate the final task complexity target level (such as intermediate + 0.1). For example, if the base level is intermediate and the dynamic correction coefficient is 0.1, the generated task complexity target level is intermediate + 0.1.

[0077] Here is a specific example: In the programming experiment class, the code submission interval sequence of student A showed that the volatility of its task response delay was 0.2, and the time dimension delay distribution matrix [0.2, 0.15] was generated; in the data analysis experiment, the platform recorded its error mode type as a logic loop missing, with a distribution density of 0.15, and a logic dimension fault topology map was generated. The two were aligned across dimensions to generate aligned data; based on the aligned data, the basic level of task complexity was calculated as intermediate, and the dynamic correction coefficient 0.1 was generated in combination with the distribution density, and its weight was dynamically adjusted to 0.8 according to the resource occupancy rate. Finally, the basic level and the dynamic correction coefficient were weighted and superimposed to generate a target level of task complexity of intermediate + 0.08.

[0078] Through the generation and cross-dimensional alignment of the time dimension delay distribution matrix and the logic dimension fault topology map, the accurate calculation of the basic level of task complexity is achieved; the dynamic correction coefficient is generated based on the distribution density and resource occupancy rate, and the target level of task complexity is generated in combination with the basic level, which realizes the dynamic adaptation and real-time adjustment of task complexity, thereby improving the personalization and efficiency of students' quality and literacy assessment.

[0079] In order to further improve the accuracy of task load adaptation and resource allocation, this application generates a load pressure index by normalizing the queue waiting time and resource occupancy rate in the concurrent experimental task load configuration, where the queue waiting time is dynamically corrected according to the priority weight of the experimental task type, and the priority weight is dynamically generated based on the basic level of task complexity and the dynamic correction coefficient; extract the task response delay volatility and logical fault distribution density in the dynamic representation vector of student ability, and map them to the time dimension delay distribution matrix and the logical dimension fault topology map respectively, to generate volatility feature scalar and distribution density feature scalar; weighted fusion of the two generates a comprehensive ability index. Combined evaluation value; dynamic matching degree is calculated based on the matching deviation between the comprehensive evaluation value of the ability and the load pressure index, and the task load adaptation index is generated in combination with the remaining available ratio of the micro-course resource capacity; according to the preset resource adaptation interval of the task load adaptation index, the priority increase amplitude of the information ethics weight reinforcement coefficient and the degradation step size of the task complexity target level are generated, and the triggering conditions for the instant feedback action are generated in combination with the task load adaptation index and the remaining available ratio of the micro-course resource capacity; finally, the priority increase amplitude and the degradation step size are bound to the triggering conditions, and the evaluation parameter adjustment instructions are generated, so as to realize the coordinated optimization of task load adaptation, resource allocation and instant feedback.

[0080] In some embodiments, the step 204 calculates the dynamic matching degree between the student ability dynamic representation vector and the concurrent experiment task load configuration, generates a task load adaptation index, and generates an evaluation parameter adjustment instruction based on the task load adaptation index and the micro-course resource capacity, including: Step 401: Normalize the queue waiting time and resource occupancy rate in the concurrent experimental task load configuration to generate a load pressure index, wherein the queue waiting time is obtained by dynamically correcting the priority weight of the experimental task type, and the priority weight of the experimental task type is dynamically generated based on the basic level of task complexity and the dynamic correction coefficient.

[0081] In step 401, the queue waiting time refers to the length of time the task waits in the queue for execution after the student submits the experimental task. For example, after student A submits the task, the task waits in the queue for 10 seconds before it starts to execute. Resource occupancy rate refers to the proportion of resources currently occupied in the online experiment platform. For example, if the total platform resources are 100% and 70% are currently occupied, the resource occupancy rate is 70%. The load pressure index is used to quantify the system load pressure. For example, an index of 0.75 indicates that the system load pressure is high.

[0082] In an embodiment of the present application, the queue waiting time and resource occupancy rate in the concurrent experimental task load configuration are normalized to generate a load pressure index. Specifically, the queue waiting time and resource occupancy rate are standardized to the same dimension through normalization technology, and the queue waiting time is dynamically corrected according to the priority weight of the experimental task type (dynamically generated based on the basic level of task complexity and the dynamic correction coefficient) to generate a load pressure index. For example, if the queue waiting time is 10 seconds, the resource occupancy rate is 70%, and the priority weight is 0.8, a load pressure index of 0.75 is generated.

[0083] Step 402: extract the task response delay fluctuation rate and the logic fault distribution density in the student ability dynamic representation vector, and map them to the time dimension delay distribution matrix and the logic dimension fault topology map respectively to obtain the fluctuation rate characteristic scalar and the distribution density characteristic scalar.

[0084] In step 402, the volatility feature scalar is used to characterize the fluctuation characteristics of the student's response delay in the programming task. For example, a scalar value of 0.2 indicates that the task response delay fluctuates greatly. The distribution density feature scalar is used to characterize the distribution characteristics of the student's logic faults in the data analysis task. For example, a scalar value of 0.15 indicates that the distribution of logic faults is relatively concentrated.

[0085] In the embodiment of the present application, the task response delay volatility and the logical fault distribution density in the dynamic characterization vector of the student's ability are extracted and mapped to the time dimension delay distribution matrix and the logical dimension fault topology map, respectively, to generate a volatility feature scalar and a distribution density feature scalar. Specifically, the volatility and distribution density are quantified into scalar values ​​through mapping technology. For example, the volatility feature scalar is 0.2, and the distribution density feature scalar is 0.15.

[0086] Step 403: weighted fusion of the volatility characteristic scalar and the distribution density characteristic scalar to generate a comprehensive capability evaluation value.

[0087] In step 403, the comprehensive ability evaluation value is used to quantify the comprehensive ability level of the student. For example, an evaluation value of 0.18 indicates that the student has a strong comprehensive ability.

[0088] In an embodiment of the present application, the volatility characteristic scalar and the distribution density characteristic scalar are weighted and fused to generate a comprehensive capability evaluation value. Specifically, the volatility characteristic scalar (such as 0.2) and the distribution density characteristic scalar (such as 0.15) are combined through weighted fusion technology (such as weighted average) to generate a comprehensive capability evaluation value (such as 0.18). For example, if the volatility characteristic scalar weight is 0.6 and the distribution density characteristic scalar weight is 0.4, the comprehensive capability evaluation value is 0.2×0.6+0.15×0.4=0.18. The weighting coefficient is dynamically adjusted according to the association strength of the logical nodes in the logical dimension fault topology diagram.

[0089] Step 404: Based on the matching deviation between the comprehensive capability evaluation value and the load pressure index, a dynamic matching degree is calculated, and based on the dynamic matching degree and the remaining available ratio of the micro-course resource capacity, a task load adaptation index is generated.

[0090] In step 404, the dynamic matching degree is used to quantify the adaptation relationship between the student's ability and the system load. The matching deviation degree refers to the difference between the comprehensive evaluation value of the student's ability and the system load pressure index, which is used to quantify the adaptation relationship between the student's ability and the system load. For example, if the comprehensive evaluation value of the ability is 0.18 and the load pressure index is 0.75, the matching deviation degree is |0.18-0.75|=0.57, indicating that the student's ability is relatively adapted to the system load.

[0091] In an embodiment of the present application, a dynamic matching degree is calculated based on the matching deviation between the comprehensive capability evaluation value and the load pressure index; based on the dynamic matching degree, combined with the remaining available ratio of the micro-course resource capacity, a task load adaptation index is generated. Specifically, a dynamic matching degree (such as |0.18-0.75|=0.57) is generated by calculating the matching deviation (such as the absolute value difference), and a task load adaptation index (such as 0.85) is generated in combination with the remaining resource ratio. For example, if the remaining resource ratio is 30%, the task load adaptation index is 0.85.

[0092] Step 405: Based on the preset resource adaptation interval of the task load adaptation index, generate the priority increase range of the information ethics weight reinforcement coefficient and the degradation step of the task complexity target level, and generate the trigger condition of the instant feedback action based on the task load adaptation index and the remaining available ratio of the micro-course resource capacity.

[0093] In step 405, the preset resource adaptation interval refers to an interval dynamically divided based on the ratio of the concurrent experiment task queue length in the micro-course resource capacity to the number of remaining micro-course resources, which is used to guide the resource allocation strategy. The priority increase range refers to the priority increase ratio of the information ethics weight reinforcement coefficient in the evaluation. For example, an increase of 0.1 means that the information ethics weight priority is increased by 10%. The downgrade step size refers to the reduction range of the task complexity target level in the evaluation. For example, a downgrade step size of 0.05 means that the task complexity is reduced by 5%.

[0094] In an embodiment of the present application, the priority increase amplitude of the information ethics weight reinforcement coefficient and the degradation step size of the task complexity target level are generated according to the preset resource adaptation interval in which the task load adaptation index is located, and the trigger condition of the instant feedback action is generated according to the task load adaptation index and the remaining available ratio of the micro-course resource capacity. Specifically, the priority increase amplitude (such as 0.1) and the degradation step size (such as 0.05) are generated by dynamically dividing the resource adaptation interval, and the trigger condition (such as triggered when the resource capacity is less than 20%) is generated in combination with the remaining resource ratio. For example, if the task load adaptation index is 0.85 and the remaining resource ratio is 30%, a priority increase amplitude of 0.1 and a degradation step size of 0.05 are generated.

[0095] Step 406: Bind the priority enhancement amplitude and the degradation step size to the triggering condition of the instant feedback action, and generate an evaluation parameter adjustment instruction.

[0096] In an embodiment of the present application, the priority increase amplitude and the degradation step size are bound to the trigger condition of the instant feedback action to generate an evaluation parameter adjustment instruction. Specifically, the priority increase amplitude (such as 0.1) and the degradation step size (such as 0.05) are combined with the trigger condition through the binding technology to generate an evaluation parameter adjustment instruction. For example, if the trigger condition is that the resource capacity is less than 20%, an evaluation parameter adjustment instruction is generated: the priority increase amplitude is 0.1, the degradation step size is 0.05, and the trigger condition is that the resource capacity is less than 20%.

[0097] Here is a specific example: In the programming experiment class, the code submission interval sequence of student A showed that the volatility of its task response delay was 0.2, generating a volatility characteristic scalar of 0.2; in the data analysis experiment, the platform recorded its logical fault distribution density as 0.15, generating a distribution density characteristic scalar of 0.15. The two are weighted and fused to generate a comprehensive capability evaluation value of 0.17; based on the matching deviation between the comprehensive capability evaluation value and the load pressure index of 0.75, the dynamic matching degree of 0.8 is calculated, and the task load adaptation index of 0.85 is generated in combination with the remaining micro-course resource ratio. According to the resource adaptation interval where the task load adaptation index is located, the priority increase amplitude of 0.1 and the degradation step size of 0.05 are generated, and the trigger condition is generated in combination with the remaining resource ratio (such as triggering when the resource capacity is less than 20%). Finally, the priority increase amplitude and degradation step size are bound to the trigger condition to generate the evaluation parameter adjustment instruction.

[0098] By generating a load pressure index and calculating a comprehensive capability assessment value, a dynamic match between student capabilities and system load is achieved. Based on the task load adaptation index and the proportion of remaining resources, the priority increase range, downgrade step size and trigger conditions are generated, and the binding technology is used to generate assessment parameter adjustment instructions, thus achieving dynamic adjustment of assessment strategies and precise adaptation of resource allocation, thereby improving the real-time and efficiency of student quality assessment.

[0099] In order to further improve the accuracy and dynamic adaptability of the task load adaptation index, the present application generates a standardized capability evaluation value and a standardized load pressure value by normalizing the comprehensive capability evaluation value and the load pressure index respectively, wherein the normalization range of the standardized capability evaluation value is dynamically adjusted based on the historical extreme value of the student's capability dynamic characterization vector; the matching deviation degree of the standardized capability evaluation value and the standardized load pressure value is calculated, and the dynamic matching degree is calculated based on the difference between the inverse of the matching deviation degree and the preset deviation tolerance threshold; the resource capacity adaptation coefficient is generated according to the remaining available ratio of the micro-course resource capacity, wherein the resource capacity adaptation coefficient is positively correlated with the square root value of the remaining available ratio, and negatively correlated with the logarithmic decay rate of the concurrent experimental task queue length; the dynamic matching degree and the resource capacity adaptation coefficient are weightedly fused to generate a task load adaptation index, thereby realizing dynamic matching of student capabilities and system load and precise adaptation of resource allocation.

[0100] In some embodiments, the step 404 calculates a dynamic matching degree based on the matching deviation between the comprehensive capability evaluation value and the load pressure index, and generates a task load adaptation index based on the dynamic matching degree and the remaining available ratio of the micro-course resource capacity, including: Step 501: Normalize the comprehensive ability evaluation value and the load pressure index respectively to generate a standardized ability evaluation value and a standardized load pressure value, wherein the normalization range of the standardized ability evaluation value is dynamically adjusted based on the historical extreme value of the student ability dynamic representation vector.

[0101] In step 501, normalization is used to eliminate dimensional differences to facilitate subsequent calculations. In an embodiment of the present application, the comprehensive ability evaluation value and the load pressure index are normalized respectively to generate a standardized ability evaluation value and a standardized load pressure value. Specifically, the comprehensive ability evaluation value and the load pressure index are standardized to values ​​within the range of [0,1] through normalization techniques (such as minimum-maximum normalization), wherein the normalization range of the standardized ability evaluation value is dynamically adjusted based on the historical extreme values ​​of the student's ability dynamic characterization vector. For example, if the comprehensive ability evaluation value is 0.18 and the historical extreme value is [0.1,0.3], then the standardized ability evaluation value is (0.18-0.1) / (0.3-0.1)=0.4.

[0102] Step 502: Calculate the matching deviation between the standardized capability evaluation value and the standardized load pressure value, and calculate the dynamic matching degree based on the difference between the inverse of the matching deviation and a preset deviation tolerance threshold.

[0103] In step 502, the matching deviation is dynamically modified according to the spatiotemporal correlation between the growth trend of the standardized capability evaluation value and the decay slope of the standardized load pressure value.

[0104] In the embodiment of the present application, the matching deviation between the standardized capacity evaluation value and the standardized load pressure value is calculated, and the dynamic matching degree is calculated based on the difference between the reciprocal of the matching deviation and the preset deviation tolerance threshold. Specifically, the matching deviation (such as |0.4-0.75|=0.35) is generated by the deviation calculation (such as the absolute value difference), and the dynamic matching degree (such as 1 / 0.35-0.2=2.66) is generated by the difference between the reciprocal and the deviation tolerance threshold.

[0105] Step 503: Generate a resource capacity adaptation coefficient according to the remaining available ratio of the micro-course resource capacity, wherein the resource capacity adaptation coefficient is positively correlated with the square root value of the remaining available ratio and negatively correlated with the logarithmic decay rate of the concurrent experiment task queue length.

[0106] In step 503, the resource capacity adaptation coefficient is used to quantify the adaptability of resource allocation. The coefficient comprehensively considers the influence of the remaining resource ratio and the length of the concurrent experimental task queue, and reflects the rationality and efficiency of the current resource allocation.

[0107] In the embodiment of the present application, a resource capacity adaptation coefficient is generated according to the remaining available ratio of the micro-course resource capacity. Specifically, the resource capacity adaptation coefficient is modeled by a composite relationship of square root and logarithmic decay, where the coefficient is positively correlated with the square root value of the remaining available ratio and negatively correlated with the logarithmic decay rate of the concurrent experimental task queue length. For example, if the remaining available ratio is 30% and the queue length is 20, the resource capacity adaptation coefficient is √0.3×(1-log(20))=0.55.

[0108] Step 504: weighted fusion of the dynamic matching degree and the resource capacity adaptation coefficient to generate a task load adaptation index.

[0109] In step 504, the weight of the weighted fusion is dynamically adjusted based on the association strength of the logical nodes in the logical dimension fault topology diagram and the volatility of the time dimension delay distribution matrix.

[0110] In the embodiment of the present application, the dynamic matching degree and the resource capacity adaptation coefficient are weighted and fused to generate a task load adaptation index. Specifically, the dynamic matching degree (such as 2.66) and the resource capacity adaptation coefficient (such as 0.55) are combined through a weighted fusion technology (such as weighted average) to generate a task load adaptation index (such as 2.66×0.6+0.55×0.4=1.8).

[0111] Here is a specific example: In the programming experiment class, the comprehensive ability evaluation value of student A is 0.18, and the historical extreme value is [0.1, 0.3], generating a standardized ability evaluation value of 0.4; the load pressure index is 0.75, generating a standardized load pressure value of 0.75. Calculate the matching deviation degree |0.4-0.75|=0.35, generate a dynamic matching degree 1 / 0.35-0.2=2.66; based on the remaining available ratio 30% and the queue length 20, generate a resource capacity adaptation coefficient of 0.55; weighted fusion of the dynamic matching degree 2.66 and the resource capacity adaptation coefficient 0.55 to generate a task load adaptation index of 1.8.

[0112] Through normalization processing, standardized ability assessment values ​​and load pressure values ​​are generated, and dynamic matching degree is generated by combining deviation calculation. The task load adaptation index is generated through the resource capacity adaptation coefficient, which realizes the dynamic matching of student ability and system load and the precise adaptation of resource allocation, thereby improving the real-time and efficiency of student quality assessment.

[0113] In order to further improve the accuracy of dynamic representation of students' abilities and reinforcement of information ethics weights, this application performs time window segmentation on the code submission interval sequence in the real-time programming task completion efficiency data, extracts the task response delay volatility, and generates a time dimension volatility vector; at the same time, performs fault clustering analysis on the logical jump paths in the data analysis error rate time series data, extracts the error pattern type and distribution density, and generates a logical fault density vector; cross-modally aligns the time dimension volatility vector with the logical fault density vector to generate a dynamic representation vector of students' abilities; performs multi-level classification encoding on the types of illegal operations in the network behavior deviation event data, generates violation type classification labels, and counts the triggering frequency within a preset time period to generate a violation frequency distribution matrix; generates an information ethics weight reinforcement coefficient based on the proportion of high-frequency violation types and the time decay factor, wherein the time decay factor is dynamically calculated based on the time series interval of the network behavior deviation event data, thereby realizing a two-dimensional dynamic evaluation of students' abilities and information ethics.

[0114] In some embodiments, step 102: generating a student ability dynamic representation vector based on the real-time programming task completion efficiency data and the data analysis error rate time series data, and generating an information ethics weight reinforcement coefficient based on the network behavior deviation event data, includes: Step 601: Perform time window segmentation on the code submission interval sequence in the real-time programming task completion efficiency data to generate a time dimension volatility vector. The span of the time window is dynamically adjusted according to the error frequency in the data analysis error rate time series data. The code submission interval sequence refers to the time interval sequence of students continuously submitting codes in programming tasks, such as [10 seconds, 15 seconds, 20 seconds]. By performing time window segmentation on the sequence, the volatility (such as standard deviation) in each time window is extracted to generate a "time dimension volatility vector". For example, if the volatility of time window 1 is 0.2 and the volatility of time window 2 is 0.15, then a vector [0.2, 0.15] is generated.

[0115] In step 601, the time dimension volatility vector is used to characterize the fluctuation characteristics of the student's response delay in the programming task.

[0116] In an embodiment of the present application, the code submission interval sequence in the real-time programming task completion efficiency data is segmented into time windows to generate a time dimension volatility vector. Specifically, the code submission interval sequence is divided into continuous time windows (such as a window every 5 minutes) through a time window segmentation technique (such as a sliding window method), and the task response delay volatility (such as a standard deviation) is calculated in each window to generate a time dimension volatility vector. For example, if the volatility of time window 1 is 0.2 and the volatility of time window 2 is 0.15, a vector [0.2, 0.15] is generated.

[0117] Step 602: Perform fault clustering analysis on the logic jump paths in the data analysis error rate time series data to generate a logic fault density vector.

[0118] In step 602, the logic fault density vector is used to characterize the distribution characteristics of the students' logic faults in the data analysis task.

[0119] In the embodiment of the present application, fault clustering analysis is performed on the logic jump path in the data analysis error rate time series data to generate a logic fault density vector. Specifically, the error pattern type (such as logic loop missing, conditional judgment redundancy) in the logic jump path is identified through a clustering algorithm (such as K-means), and its distribution density (such as error frequency and time distribution) is counted to generate a logic fault density vector. For example, if the distribution density of logic loop missing is 0.3 and the distribution density of conditional judgment redundancy is 0.2, then a vector [0.3, 0.2] is generated.

[0120] Step 603: Perform cross-modal alignment on the time dimension volatility vector and the logic fault density vector to generate a dynamic representation vector of student ability.

[0121] In step 603, the weight of the cross-modal alignment is dynamically adjusted based on the distribution ratio of the violation types in the network behavior deviation event data.

[0122] In an embodiment of the present application, the time dimension volatility vector and the logical fault density vector are cross-modally aligned to generate a student ability dynamic representation vector. Specifically, the response delay characteristics of the programming task and the logical fault characteristics of the data analysis are fused through cross-modal alignment technology (such as dynamic time warping) to generate a student ability dynamic representation vector. For example, if the time dimension volatility vector is [0.2, 0.15] and the logical fault density vector is [0.3, 0.2], then the student ability dynamic representation vector [0.2, 0.15, 0.3, 0.2] is generated.

[0123] Step 604: Perform multi-level classification coding on the illegal operation types in the network behavior deviation event data, generate violation type classification labels, count the triggering frequency of the violation type classification labels within a preset time period, and generate a violation frequency distribution matrix.

[0124] In step 604, the violation frequency distribution matrix is ​​used to characterize the distribution of students' violation behaviors within a preset time period. Violation operation types include data tampering, illegal access, privacy leakage and network attack. Violation type classification labels are used to quantify the types and frequencies of violation behaviors. For example, the label of data tampering is 1, and the label of illegal access is 2.

[0125] In the embodiment of the present application, the illegal operation types in the network behavior deviation event data are multi-level classified and encoded to generate illegal type classification labels, and the triggering frequency of the illegal type classification labels within a preset time period is counted to generate a violation frequency distribution matrix. Specifically, the illegal operation types (such as data tampering, illegal access) are encoded as classification labels through classification coding technology (such as unique hot coding), and their triggering frequencies within a preset time period (such as one week) are counted to generate a violation frequency distribution matrix. For example, if the frequency of data tampering is 3 times and the frequency of illegal access is 2 times, then the matrix [3,2] is generated.

[0126] Step 605: Generate an information ethics weight reinforcement coefficient based on the proportion of high-frequency violation types in the violation frequency distribution matrix and the time decay factor, wherein the time decay factor is dynamically calculated based on the time series interval of the network behavior deviation event data.

[0127] In step 605, the proportion of high-frequency violation types refers to the proportion of the triggering frequency of a certain type of violation operation to the total violation frequency within a preset time period. For example, if the total violation frequency is 5 times, of which the frequency of data tampering is 3 times, the proportion of high-frequency violation types of data tampering is 60%. By counting the triggering frequency of various types of violation operations within a preset time period, the proportion of various types of violation operations is calculated. If the proportion of a certain type of violation operation exceeds a preset threshold (such as 50%), it is judged to be a high-frequency violation type. For example, if the frequency of data tampering is 3 times and the total violation frequency is 5 times, the proportion of data tampering is 60%, which exceeds the threshold of 50%, and it is judged to be a high-frequency violation type. The time decay factor is used to reflect the timeliness of the violation. For example, if the time interval between the violation events is short, the time decay factor is higher; conversely, if the time interval is long, the time decay factor is lower.

[0128] In the embodiment of the present application, the information ethics weight reinforcement coefficient is generated according to the high-frequency violation type proportion and the time decay factor of the violation frequency distribution matrix. Specifically, the information ethics weight reinforcement coefficient is generated by the high-frequency violation type proportion (such as data tampering proportion of 60%) and the time decay factor (such as the inverse of the time interval of the violation event). For example, if the high-frequency violation type proportion is 60% and the time decay factor is 0.8, the information ethics weight reinforcement coefficient is generated as 0.6×0.8=0.48.

[0129] Here is a specific example: In the programming experiment class, the code submission interval sequence of student A showed that the volatility of his task response delay was 0.2, generating a time dimension volatility vector [0.2, 0.15]; in the data analysis experiment, the platform recorded its logic fault distribution density as 0.3 (logic loop missing) and 0.2 (conditional judgment redundancy), generating a logic fault density vector [0.3, 0.2]. The two were cross-modally aligned to generate a student ability dynamic representation vector [0.2, 0.15, 0.3, 0.2]; in network behavior monitoring, the platform recorded student A's data tampering frequency as 3 times and illegal access frequency as 2 times, generating a violation frequency distribution matrix [3, 2]; based on the high-frequency violation type accounting for 60% and the time decay factor of 0.8, the information ethics weight reinforcement coefficient was generated as 0.48.

[0130] Through time window segmentation and fault clustering analysis, the time dimension volatility vector and logical fault density vector are generated, and through cross-modal alignment, the student ability dynamic representation vector is generated, which realizes the dynamic evaluation of students' comprehensive ability; the information ethics weight reinforcement coefficient is generated by the violation frequency distribution matrix and the time attenuation factor, which realizes the dynamic adjustment of the information ethics dimension, thereby improving the accuracy and real-time performance of students' quality and literacy assessment.

[0131] In order to further improve the global optimization capability of resource allocation strategy and evaluation strategy, this application extracts resource type labels and delivery frequency distribution characteristics corresponding to micro-course resource capacity by multi-dimensionally encoding the priority queue sorting rules in the directional resource delivery instructions; performs periodic clustering analysis on the resource consumption rate and time interval fluctuations in the micro-course resource consumption trajectory data, and identifies the switching thresholds between high-frequency consumption periods and low-frequency consumption periods; performs spatiotemporal correlation mapping of delivery frequency distribution characteristics and switching thresholds to generate resource collaborative adaptation chromosome fragments; extracts boundary features of the constraint range of the information ethics weight reinforcement coefficient to generate ethical weight constraint gene fragments, and quantifies the volatility of the benchmark mapping rules of the task complexity target level to generate complexity mapping gene fragments; cross-recombines the ethical weight constraint gene fragments with the complexity mapping gene fragments to generate a globally optimized chromosome group, wherein the node positions of the cross-recombination are dynamically selected according to the spatiotemporal correlation strength in the resource collaborative adaptation chromosome fragments, thereby achieving global optimization and dynamic adaptation of resource allocation strategy and evaluation strategy.

[0132] In some embodiments, the step 104 of aggregating the directional resource delivery instruction and the micro-course resource consumption trajectory data to generate a global optimized chromosome set includes: Step 701: multi-dimensionally encode the priority queue sorting rules in the directional resource delivery instruction, and extract the resource type label and delivery frequency distribution characteristics corresponding to the micro-course resource capacity in the priority queue sorting rules.

[0133] In step 701, the resource type tag is used to distinguish different types of micro-course resources (such as programming micro-courses and data analysis cases). The delivery frequency distribution feature is used to characterize the frequency and priority of resource delivery. Multidimensional encoding refers to the process of quantifying the features of multiple dimensions (such as resource type and delivery frequency) into numerical vectors. For example, if the resource type is a programming micro-course (encoded as [1,0,0]) and the delivery frequency is 3 times (encoded as [3]), the multidimensional encoding result is [1,0,0,3].

[0134] In the embodiment of the present application, the priority queue sorting rules in the directional resource delivery instruction are multi-dimensionally encoded, and the resource type label and delivery frequency distribution characteristics corresponding to the micro-course resource capacity in the priority queue sorting rules are extracted. Specifically, the resource type and delivery frequency are quantified into feature vectors through multi-dimensional encoding technology (such as unique hot encoding), and the resource type label and delivery frequency distribution characteristics are generated. For example, if the resource type is a programming micro-course and the delivery frequency is 3 times, then the resource type label 1 and the delivery frequency distribution characteristics [3] are generated.

[0135] Step 702: Perform periodic clustering analysis on the resource consumption rate and time interval fluctuations in the micro-course resource consumption trajectory data to identify a switching threshold for distinguishing between high-frequency consumption periods and low-frequency consumption periods.

[0136] In step 702, the process of switching the threshold is: in the high-frequency consumption period [a1, a2], the resource consumption rate of each time node is greater than the switching threshold. For example, if the switching threshold is 0.75, and the consumption rate in the period [a1, a2] is [0.8, 0.9, 1.0], then the consumption rate of each time node is greater than 0.75, which belongs to the high-frequency consumption period. In the low-frequency consumption period [b1, b2], the resource consumption rate of each time node is less than the switching threshold. For example, if the switching threshold is 0.75, and the consumption rate in the period [b1, b2] is [0.5, 0.6, 0.7], then the consumption rate of each time node is less than 0.75, which belongs to the low-frequency consumption period.

[0137] In an embodiment of the present application, a periodic clustering analysis is performed on the resource consumption rate and time interval fluctuations in the micro-course resource consumption trajectory data to identify a switching threshold for distinguishing high-frequency consumption periods from low-frequency consumption periods. Specifically, the resource consumption rate and time interval fluctuations are divided into two categories, high frequency and low frequency, through a clustering algorithm to generate a switching threshold. For example, if the resource consumption rate in the high-frequency consumption period is 1 time per minute and the low-frequency consumption period is 0.5 times per minute, a switching threshold of 0.75 is generated.

[0138] Step 703: Perform spatiotemporal correlation mapping between the delivery frequency distribution characteristics and the switching threshold to generate resource collaborative adaptation chromosome fragments, extract boundary features of the constraint range of the information ethics weight reinforcement coefficient to generate ethical weight constraint gene fragments, and quantify the volatility of the benchmark mapping rules of the task complexity target level to generate complexity mapping gene fragments.

[0139] In step 703, the resource cooperative adaptation chromosome segment is used to characterize the cooperative adaptability of the resource allocation strategy, the ethical weight constraint gene segment is used to characterize the boundary characteristics of the ethical weight, and the complexity mapping gene segment is used to characterize the fluctuation characteristics of the task complexity.

[0140] In the embodiment of the present application, the delivery frequency distribution characteristics and the switching threshold are subjected to spatiotemporal correlation mapping to generate resource collaborative adaptation chromosome fragments; the boundary characteristics of the constraint range of the information ethics weight reinforcement coefficient are extracted to generate ethical weight constraint gene fragments; and the benchmark mapping rules of the task complexity target level are subjected to volatility quantification to generate complexity mapping gene fragments. Specifically, the delivery frequency distribution characteristics and the switching threshold are fused through spatiotemporal correlation mapping technology to generate resource collaborative adaptation chromosome fragments; the ethical weight constraint gene fragments are generated through boundary feature extraction technology; and the complexity mapping gene fragments are generated through volatility quantification technology. For example, if the delivery frequency distribution characteristics are [3] and the switching threshold is 0.75, the resource collaborative adaptation chromosome fragment [3,0.75] is generated.

[0141] Step 704: Cross-recombining the ethical weight constraint gene segment with the complexity mapping gene segment to generate a globally optimized chromosome set, wherein the node position of the cross-recombination is dynamically selected according to the spatiotemporal correlation strength in the resource collaborative adaptation chromosome segment.

[0142] In step 704, a specific embodiment is given for the dynamic selection of spatiotemporal correlation strength. For example, if the resource collaborative adaptation chromosome segment is [3, 0.75] and the spatiotemporal correlation strength is 0.8, the node position with higher correlation strength (such as the first node) is selected for crossover recombination. The specific steps are as follows: Chromosome segment 1 (ethical weight constraint gene segment): [0.7, 1.0], Chromosome segment 2 (complexity mapping gene segment): [0.2, 0.15], Crossover recombination node position: Select the first node (with higher correlation strength) for crossover recombination. Recombination result: Generate a new chromosome set [0.2, 1.0, 0.7, 0.15].

[0143] In the embodiment of the present application, the ethical weight constraint gene fragment and the complexity mapping gene fragment are cross-recombined to generate a global optimized chromosome set, wherein the node position of the cross-recombination is dynamically selected according to the spatiotemporal correlation intensity in the resource collaborative adaptation chromosome fragment. Specifically, the ethical weight constraint gene fragment is combined with the complexity mapping gene fragment by the cross-recombination technology to generate a global optimized chromosome set. For example, if the ethical weight constraint gene fragment is [0.7, 1.0] and the complexity mapping gene fragment is [0.2, 0.15], then a global optimized chromosome set [0.7, 1.0, 0.2, 0.15] is generated.

[0144] Here is a specific example: In the programming experiment class, in the directional resource delivery instruction of student A, the resource type is programming micro-course, the delivery frequency is 3 times, and the resource type label 1 and the delivery frequency distribution characteristics [3] are generated; the platform records the micro-course resource consumption trajectory data, and generates a switching threshold of 0.75 through cluster analysis; the delivery frequency distribution characteristics [3] and the switching threshold 0.75 are mapped in time and space to generate a resource collaborative adaptation chromosome fragment [3, 0.75]; based on the constraint range of the information ethics weight reinforcement coefficient [0.7, 1.0], the ethical weight constraint gene fragment [0.7, 1.0] is generated, and based on the benchmark mapping rule of the task complexity target level, the complexity mapping gene fragment [0.2, 0.15] is generated; the ethical weight constraint gene fragment [0.7, 1.0] and the complexity mapping gene fragment [0.2, 0.15] are cross-recombined to generate a global optimized chromosome group [0.7, 1.0, 0.2,0.15].

[0145] Through multidimensional coding and periodic clustering analysis, resource type labels, delivery frequency distribution characteristics and switching thresholds are generated, and through spatiotemporal correlation mapping, resource collaborative adaptation chromosome fragments are generated. Based on boundary feature extraction and volatility quantification, ethical weight constraint gene fragments and complexity mapping gene fragments are generated. Through cross-recombination, a globally optimized chromosome group is generated, which realizes the global optimization and dynamic adaptation of resource allocation strategies and evaluation strategies, thereby improving the accuracy and efficiency of students' quality and literacy assessment.

[0146] Figure 2 A structural diagram of a student quality assessment system is provided for the present application embodiment, such as Figure 2 As shown, the system includes: An acquisition module 21 is used to acquire multi-dimensional behavior data generated by the interaction between students and the online experiment platform, network behavior deviation event data triggered by the online experiment platform, and resource margin data dynamically updated by the online experiment platform in the teaching environment, wherein the multi-dimensional behavior data includes real-time programming task completion efficiency data and data analysis error rate time series data, and the resource margin data includes the currently allocable micro-course resource capacity, micro-course resource consumption trajectory data, and concurrent experiment task load configuration; A generating module 22, for generating a dynamic representation vector of student ability based on the real-time programming task completion efficiency data and the data analysis error rate time series data, and generating an information ethics weight reinforcement coefficient according to the network behavior deviation event data; A matching module 23 is used to generate a task complexity target level based on the student ability dynamic characterization vector, match the student ability dynamic characterization vector with the concurrent experiment task load configuration, generate an evaluation parameter adjustment instruction, trigger an instant feedback action adapted to the micro-course resource capacity based on the evaluation parameter adjustment instruction, and generate a directional resource delivery instruction; The reconstruction module 24 is used to aggregate the directional resource delivery instructions and the micro-course resource consumption trajectory data at the end of the semester to generate a global optimization chromosome set. The global optimization chromosome set generates an initialization strategy parameter set for the next evaluation cycle by reconstructing the information ethics weight reinforcement coefficient and the task complexity target level.

[0147] Figure 2 The student quality assessment system can be implemented Figure 1 The implementation principle and technical effect of the method for evaluating the quality and accomplishment of students described in the embodiment shown are not described in detail. The specific way in which each module and unit performs operations in the evaluation system for the quality and accomplishment of students in the above embodiment has been described in detail in the embodiment of the method, and will not be described in detail here.

[0148] In one possible design, Figure 2 The student quality assessment system of the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32; The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0149] The processing component 32 is described above Figure 1 The embodiment provides a method for evaluating students' quality and accomplishment.

[0150] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0151] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0152] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0153] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.

[0154] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0155] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0156] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 A method for evaluating students' quality and accomplishments in the illustrated embodiment.

[0157] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0158] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0159] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0160] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for evaluating students' quality and accomplishment, characterized in that: include: Acquire multi-dimensional behavior data generated by the interaction between students and the online experiment platform, network behavior deviation event data triggered by the online experiment platform, and resource margin data dynamically updated by the online experiment platform in the teaching environment, wherein the multi-dimensional behavior data includes real-time programming task completion efficiency data and data analysis error rate time series data, and the resource margin data includes the currently allocable micro-course resource capacity, micro-course resource consumption trajectory data, and concurrent experiment task load configuration; Based on the real-time programming task completion efficiency data and the data analysis error rate time series data, a student ability dynamic representation vector is generated, and based on the network behavior deviation event data, an information ethics weight reinforcement coefficient is generated; Based on the student ability dynamic characterization vector, a task complexity target level is generated, and the student ability dynamic characterization vector is matched with the concurrent experiment task load configuration, and an evaluation parameter adjustment instruction is generated. Based on the evaluation parameter adjustment instruction, an instant feedback action adapted to the micro-course resource capacity is triggered, and a directional resource delivery instruction is generated; At the end of the semester, the directional resource delivery instructions and the micro-course resource consumption trajectory data are aggregated to generate a global optimization chromosome set. The global optimization chromosome set generates an initialization strategy parameter set for the next evaluation cycle by reconstructing the information ethics weight reinforcement coefficient and the task complexity target level.

2. The method according to claim 1, characterized in that The method generates a target level of task complexity based on the dynamic characterization vector of student ability, matches the dynamic characterization vector of student ability with the concurrent experiment task load configuration, generates an evaluation parameter adjustment instruction, triggers an instant feedback action adapted to the micro-course resource capacity based on the evaluation parameter adjustment instruction, and generates a directional resource delivery instruction, including: Performing task response delay slicing processing on the student ability dynamic representation vector, extracting the code submission interval sequence in the continuous time window of the real-time programming task completion efficiency data, and generating a task response delay feature vector according to the fluctuation amplitude and trend direction of the code submission interval sequence; Performing a logical fault cluster analysis on the student ability dynamic representation vector, identifying the error pattern type and distribution density of repeated correction operations in the data analysis error rate time series data, and generating a logical fault feature vector based on the error pattern type and distribution density; Performing spatiotemporal alignment and fusion of the task response delay feature vector and the logic fault feature vector to generate a task complexity target level; Calculate the dynamic matching degree between the dynamic characterization vector of the student ability and the concurrent experiment task load configuration, generate a task load adaptation index, and generate an evaluation parameter adjustment instruction based on the task load adaptation index and the micro-course resource capacity, wherein the evaluation parameter adjustment instruction includes the priority increase range of the information ethics weight reinforcement coefficient, the demotion step of the task complexity target level, and the triggering condition of the instant feedback action; According to the trigger condition, the instant feedback action is mapped to the priority queue corresponding to the micro-course resource capacity, and a directional resource delivery instruction is generated.

3. The method according to claim 2, characterized in that The step of performing spatiotemporal alignment and fusion of the task response delay feature vector and the logic fault feature vector to generate a task complexity target level includes: Performing time window segmentation on the task response delay feature vector, extracting the task response delay fluctuation rate in each time window, so as to generate a time dimension delay distribution matrix; Performing logical path mapping on the logical fault feature vector, identifying the association between the repeated correction operation and the logical nodes in the logical path, so as to generate a logical dimension fault topology map; Cross-dimensionally aligning the task response delay fluctuation rate in the time dimension delay distribution matrix with the association relationship in the logic dimension fault topology map; Based on the matching degree between the aligned task response delay fluctuation rate and the correlation relationship, the basic level of task complexity is calculated, and a dynamic correction coefficient is generated in combination with the distribution density in the logical dimension fault topology map, wherein the weight of the dynamic correction coefficient is obtained by dynamically adjusting the resource occupancy rate in the concurrent experimental task load configuration; The task complexity basic level and the dynamic correction coefficient are weighted and superimposed to generate a task complexity target level.

4. The method according to claim 2, characterized in that: The calculating of the dynamic matching degree between the dynamic characterization vector of the student ability and the concurrent experiment task load configuration, generating a task load adaptation index, and generating an evaluation parameter adjustment instruction based on the task load adaptation index and the micro-course resource capacity, comprises: The queue waiting time and resource occupancy rate in the concurrent experimental task load configuration are normalized to generate a load pressure index, wherein the queue waiting time is obtained by dynamically correcting the priority weight of the experimental task type, and the priority weight of the experimental task type is dynamically generated based on the basic level of task complexity and the dynamic correction coefficient; Extracting the task response delay fluctuation rate and the logic fault distribution density in the student ability dynamic representation vector, and mapping them to the time dimension delay distribution matrix and the logic dimension fault topology map, respectively, to obtain the fluctuation rate characteristic scalar and the distribution density characteristic scalar; Performing weighted fusion of the volatility characteristic scalar and the distribution density characteristic scalar to generate a comprehensive capability evaluation value; Based on the matching deviation between the comprehensive capability evaluation value and the load pressure index, a dynamic matching degree is calculated, and based on the dynamic matching degree and the remaining available ratio of the micro-course resource capacity, a task load adaptation index is generated; According to the preset resource adaptation interval in which the task load adaptation index is located, the priority enhancement range of the information ethics weight reinforcement coefficient and the degradation step of the task complexity target level are generated, and the triggering condition of the instant feedback action is generated according to the task load adaptation index and the remaining available ratio of the micro-course resource capacity, wherein the preset resource adaptation interval is dynamically divided based on the ratio of the concurrent experiment task queue length in the micro-course resource capacity to the number of remaining micro-course resources; The priority enhancement range and the downgrade step size are bound to the triggering condition of the instant feedback action to generate an evaluation parameter adjustment instruction.

5. The method according to claim 4, characterized in that The method of calculating the dynamic matching degree based on the matching deviation between the comprehensive capability evaluation value and the load pressure index, and generating the task load adaptation index based on the dynamic matching degree and the remaining available ratio of the micro-course resource capacity includes: Normalizing the comprehensive ability evaluation value and the load pressure index respectively to generate a standardized ability evaluation value and a standardized load pressure value, wherein the normalization range of the standardized ability evaluation value is dynamically adjusted based on the historical extreme value of the student ability dynamic representation vector; Calculating a matching deviation between the standardized capability evaluation value and the standardized load pressure value, and calculating a dynamic matching degree based on a difference between a reciprocal of the matching deviation and a preset deviation tolerance threshold; Generate a resource capacity adaptation coefficient according to the remaining available ratio of the micro-course resource capacity, wherein the resource capacity adaptation coefficient is positively correlated with the square root value of the remaining available ratio and negatively correlated with the logarithmic decay rate of the concurrent experiment task queue length; The dynamic matching degree and the resource capacity adaptation coefficient are weighted and integrated to generate a task load adaptation index.

6. The method according to claim 1, characterized in that The method of generating a student ability dynamic representation vector based on the real-time programming task completion efficiency data and the data analysis error rate time series data, and generating an information ethics weight reinforcement coefficient based on the network behavior deviation event data, includes: Performing time window segmentation on the code submission interval sequence in the real-time programming task completion efficiency data to generate a time dimension volatility vector; Performing fault clustering analysis on the logic jump paths in the data analysis error rate time series data to generate a logic fault density vector; Cross-modally aligning the time dimension volatility vector with the logic fault density vector to generate a dynamic representation vector of student ability; Perform multi-level classification coding on the illegal operation types in the network behavior deviation event data, generate illegal type classification labels, count the triggering frequency of the illegal type classification labels within a preset time period, and generate an illegal frequency distribution matrix; An information ethics weight reinforcement coefficient is generated based on the proportion of high-frequency violation types and the time decay factor in the violation frequency distribution matrix, wherein the time decay factor is dynamically calculated based on the time series interval of the network behavior deviation event data.

7. The method according to claim 1, characterized in that The aggregating the directional resource delivery instruction and the micro-course resource consumption trajectory data to generate a global optimization chromosome set includes: Multi-dimensionally encode the priority queue sorting rules in the directional resource delivery instruction, and extract the resource type label and delivery frequency distribution characteristics corresponding to the micro-course resource capacity in the priority queue sorting rules; Performing periodic cluster analysis on the resource consumption rate and time interval fluctuations in the micro-course resource consumption trajectory data to identify a switching threshold for distinguishing between a high-frequency consumption period and a low-frequency consumption period; The frequency distribution characteristics of the delivery are mapped with the switching threshold in a spatiotemporal manner to generate a resource collaborative adaptation chromosome fragment, the constraint range of the information ethics weight reinforcement coefficient is extracted with boundary features to generate an ethics weight constraint gene fragment, and the benchmark mapping rule of the task complexity target level is quantified with volatility to generate a complexity mapping gene fragment; The ethical weight constraint gene segment and the complexity mapping gene segment are cross-recombined to generate a globally optimized chromosome set, wherein the node position of the cross-recombination is dynamically selected according to the spatiotemporal correlation strength in the resource collaborative adaptation chromosome segment.

8. A student quality assessment system, characterized by: include: An acquisition module is used to acquire multi-dimensional behavior data generated by the interaction between students and the online experiment platform, network behavior deviation event data triggered by the online experiment platform, and resource margin data dynamically updated by the online experiment platform in the teaching environment, wherein the multi-dimensional behavior data includes real-time programming task completion efficiency data and data analysis error rate time series data, and the resource margin data includes the currently allocable micro-course resource capacity, micro-course resource consumption trajectory data, and concurrent experiment task load configuration; A generation module, for generating a dynamic representation vector of student ability based on the real-time programming task completion efficiency data and the data analysis error rate time series data, and generating an information ethics weight reinforcement coefficient based on the network behavior deviation event data; A matching module is used to generate a task complexity target level based on the student ability dynamic characterization vector, match the student ability dynamic characterization vector with the concurrent experiment task load configuration, generate an evaluation parameter adjustment instruction, trigger an instant feedback action adapted to the micro-course resource capacity based on the evaluation parameter adjustment instruction, and generate a directional resource delivery instruction; The reconstruction module is used to aggregate the targeted resource delivery instructions and the micro-course resource consumption trajectory data at the end of the semester to generate a global optimization chromosome set. The global optimization chromosome set generates an initialization strategy parameter set for the next evaluation cycle by reconstructing the information ethics weight reinforcement coefficient and the task complexity target level.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for evaluating students' quality and literacy as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, a method for evaluating students' quality and accomplishment as described in any one of claims 1 to 7 is implemented.

Images