A Community Education Resource Scheduling System Based on Multidimensional Demand Matching

Abstract

The application discloses a community education resource scheduling system based on multi-dimensional demand matching and concretely relates to the field of education resource management, and comprises a disturbance construction module, an acquisition module, an extraction module, an evaluation module and a scheduling module. The disturbance construction module is used for constructing a disturbance task sequence of a target course. The disturbance task sequence is composed of multiple disturbance tasks. Each disturbance task comprises task content of the target course, a random incentive parameter and a task triggering time. By constructing the disturbance task to induce behavior response, extracting behavior characteristics and combining a motivation deviation evaluation function, identification of pseudo-learning behavior and dynamic adjustment of scheduling priority are realized, so that long-term occupation of high-value education resources by non-true demand users is inhibited, and the authenticity of education resource allocation and the fairness of system operation are improved.

Description

Technical Field

[0001] This invention relates to the field of educational resource management technology, and more specifically, to a community educational resource scheduling system based on multi-dimensional demand matching. Background Technology

[0002] In the current community education dispatch system, skills training courses have gradually become an important vehicle for supporting employment transformation, distributing social assistance, and incentivizing residents to relearn. In order to improve system response efficiency and user participation, most platforms have established a data-driven matching mechanism with registration priority, attendance records, and phased task completion rate as the core. They also enhance incentives by linking with the points reward system and unemployment subsidy policy. On the surface, this approach establishes the criteria and evaluation benchmarks for resource allocation and can achieve basic supply and demand matching.

[0003] However, in practical applications, such mechanisms exhibit a type of distorted behavior. Some users, driven by external interests such as obtaining course points, employment subsidies, or government service certification, frequently enroll in high-incentive courses without genuine learning or employment transition needs. After successful enrollment, their behavior exhibits typical weak execution path characteristics, including repetitive idling, periodic task avoidance, and interrupted course abandonment. These users often have a "superficially complete" data trajectory in the system records, making them difficult to identify by traditional filtering mechanisms based on single-point events. They occupy educational resource slots for a long time, seriously interfering with the teaching rhythm and the fairness of resource allocation.

[0004] Furthermore, this behavior is statistically pseudo-random and can be easily masked by existing threshold-based judgment methods. For example, some users complete task nodes alternately to create the illusion of "periodic achievement," and then continuously seek policy subsidies or community points by quickly enrolling in other courses, ultimately forming a false activity cluster within the system. This not only passively delays the ranking of users with real needs, but also causes the scheduling system to misjudge course popularity and resource demand, resulting in structural resource squeeze.

[0005] In summary, the problem with existing technologies is that current community education resource allocation systems lack the ability to identify pseudo-learning behaviors and cannot identify resource users whose goals are not learning. This results in high-value resources being continuously consumed under the guise of high utilization rates, rather than being transformed into real educational outcomes. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a community education resource scheduling system based on multidimensional demand matching. By constructing perturbation task-induced behavioral responses, extracting behavioral features, and combining them with a motivational offset evaluation function, the system can identify pseudo-learning behaviors and dynamically adjust scheduling priorities, thereby suppressing the long-term occupation of high-value educational resources by users with non-genuine needs and improving the authenticity of educational resource allocation and the fairness of system operation.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a community education resource scheduling system based on multi-dimensional demand matching, comprising a disturbance construction module, a collection module, an extraction module, an evaluation module, and a scheduling module;

[0008] The perturbation construction module is used to construct a perturbation task sequence for the target course. The perturbation task sequence consists of multiple perturbation tasks, each of which includes the task content of the target course, random incentive parameters, and task triggering time.

[0009] The data collection module is used to collect behavioral data of users who have registered for the target course in a perturbation task sequence. The behavioral data includes task entry time, task completion status, task skipping status, and stimulus response time. The behavioral data is then reorganized according to the task trigger time to construct the user's task response sequence.

[0010] The extraction module extracts several offset features based on the task response sequence to measure the consistency of user behavior. These offset features include task response delay value, task completion rate change magnitude, and stimulus response intensity. The offset features are then combined to construct a feature vector of user behavior.

[0011] The evaluation module is used to input the feature vector into the motivation offset function, calculate the user's motivation offset value, measure the user's goal consistency under the perturbation task sequence through the motivation offset value, and compare the motivation offset value with a preset threshold.

[0012] The scheduling module includes: when the motivation offset value exceeds a preset threshold, executing scheduling restrictions, the scheduling restrictions including reducing the user's scheduling priority for incentive courses, suspending their registration qualification, or activating additional verification, until the user's motivation offset value falls back to within the preset offset threshold and the scheduling restrictions are restored.

[0013] In a preferred embodiment, within the perturbation construction module, the task flow of the target course is structurally decomposed, and each task content in the task flow is extracted and labeled as the first task. Task content of each task ,in , Indicates the total number of tasks;

[0014] For each Set a random excitation parameter and combined with time windows Construct the excitation function :

[0015] ;

[0016] in This is the actual completion time. Excitation decay coefficient; time window express The expected completion time;

[0017] For each Introducing a disturbance time offset term And reset the task trigger time. for: ; Indicates the first The perturbation offset for each task; Indicates the interval Uniform distribution on; This represents the maximum disturbance amplitude;

[0018] based on , and , integrated into the first A disturbance task ;

[0019] All Arrange the tasks according to the original task flow to generate a perturbed task sequence: ;

[0020] The perturbation task sequence It is embedded in the task execution process of the target course, replacing the original task set.

[0021] In a preferred embodiment, the acquisition module extracts the perturbation task sequence that each user who has registered for the target course participates in. ;

[0022] For each Record the corresponding user's behavioral data, including the task entry time. Task completion status Task skipped status Incentive response time ;

[0023] based on Composition of behavioral data units ;

[0024] Based on task trigger time For all behavioral data units Sort in ascending order to obtain a sequence of behavioral data ordered by time. ;

[0025] behavioral data sequences With the corresponding task trigger time set Combine and construct a task response sequence .

[0026] In a preferred embodiment, the task entry time Represented as:

[0027] ;

[0028] The task completion status Represented as:

[0029] ;

[0030] ;

[0031] The skip behavior state Represented as:

[0032] ;

[0033] The excitation response time Represented as:

[0034] ;

[0035] Task entry time Indicates the first The entry time of each task; Indicates task content The complexity of the task content The complexity includes the number of steps or interactions;

[0036] Task completion status Indicates whether the task is completed. Indicates the task completion status; otherwise, it is... ; This indicates the decay rate of the excitation function value over time offset; This refers to the actual completion time. This indicates the upper limit of the allowed response time threshold for this task;

[0037] in Indicates whether to skip the task. This indicates that the user skipped the task; otherwise, it means... .

[0038] In a preferred embodiment, the task response latency value is calculated in the extraction module. :

[0039] ;

[0040] in Indicates the first The time delay from when a task is triggered to when it actually responds; It is an exponential function;

[0041] The task completion rate change is calculated in the extraction module. :

[0042] ;

[0043] Calculate the excitation response intensity in the extraction module. :

[0044] ;

[0045] in Indicates the interval of task duration The maximum value of the rate of change of the excitation function is selected internally; This represents the rate of change of the excitation function; This represents the overall cumulative intensity of the task's activation function;

[0046] In the extraction module, the three offset features are combined into a feature vector of user behavior:

[0047] ;

[0048] in A feature vector representing user behavior.

[0049] In a preferred embodiment, a motivational offset function is constructed in the evaluation module:

[0050] ;

[0051] ;

[0052] in Indicates user The motivation offset value; Indicates allocation to user Disturbance mission The total number;

[0053] Among users The eigenvectors are represented as , Indicates user The average task response delay for all perturbation tasks, in time; This represents the maximum allowed latency threshold across all perturbed tasks; This represents the normalized delay value; Indicates user The magnitude of change in task completion rate in a perturbed task sequence; Indicates user The average intensity of the stimulus response across all tasks;

[0054] in Indicates time Next, the One assigned to user Excitation function of the perturbation task The value; Indicates to Perform sigmoid mapping; This represents the time from time 0 to the actual end time of each task. Integrate the stimulus response between them; It is the hyperbolic tangent function.

[0055] The technical effects and advantages of this invention are as follows:

[0056] 1. This invention constructs dynamic behavioral trajectories and extracts behavioral response features by introducing a perturbation task mechanism, thereby achieving automatic identification of pseudo-learning motivation under unsupervised conditions. This solves the problems of difficulty in identifying users with non-genuine learning intentions and false activity masking resource consumption in existing scheduling systems, ensuring relatively efficient allocation of resources to users with genuine needs.

[0057] 2. This invention embeds incentive-response perturbation tasks into the course task process, inducing users to exhibit differentiated responses at unexpected time points, thereby forming high-resolution behavioral trajectories that can be used to assess their behavioral consistency and motivational stability, breaking through the limitations of traditional task data modeling which only focuses on whether the task is completed or not.

[0058] 3. By collecting and analyzing multi-dimensional data on the response time, completion status, and stimulus sensitivity of disturbance tasks, a feature vector reflecting the degree of motivational deviation was constructed, enabling the system to establish a highly robust and irregular data discrimination channel, thereby improving the identification rate and screening accuracy of potential resource disturbance behaviors.

[0059] 4. The motivational deviation evaluation function adopts a nonlinear combination of multiple behavioral features and constructs a dynamic threshold judgment strategy with a recovery mechanism. This enables the system to not only have the ability to react quickly to abnormal behavior, but also to support time-evolution-based behavioral state repair and strategy adjustment, thereby enhancing the system's adaptability and strategy flexibility. Attached Figure Description

[0060] Figure 1 This is a schematic diagram of the system modules of the present invention. Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] Refer to the instruction manual appendix Figure 1 An embodiment of the present invention provides a community education resource scheduling system based on multidimensional demand matching, comprising a disturbance construction module, a collection module, an extraction module, an evaluation module, and a scheduling module.

[0063] The perturbation construction module is used to construct a perturbation task sequence for the target course. The perturbation task sequence consists of multiple perturbation tasks. Each perturbation task includes the task content of the target course, random incentive parameters, and task trigger time, which are used to be embedded in the task flow of the target course to form a non-fixed task execution path based on the perturbation task sequence.

[0064] The data collection module is used to collect behavioral data of users who have registered for the target course in a perturbation task sequence. The behavioral data includes task entry time, task completion status, task skipping status, and stimulus response time. The behavioral data is then reorganized according to the task trigger time to construct the user's task response sequence.

[0065] The extraction module extracts several offset features based on the task response sequence to measure the consistency of user behavior. The offset features include task response delay value, task completion rate change magnitude and incentive response intensity. The offset features are combined to construct a feature vector of user behavior. The feature vector is used to quantify the stability of user behavior under perturbed task sequences.

[0066] The evaluation module is used to input the feature vector into the motivation offset function, calculate the user's motivation offset value, measure the user's goal consistency under the perturbation task sequence through the motivation offset value, and compare the motivation offset value with a preset threshold.

[0067] The scheduling module includes: when the motivation offset value exceeds a preset threshold, executing scheduling restrictions, the scheduling restrictions including reducing the user's scheduling priority for high-incentive courses, suspending their registration qualification, or activating additional verification, until the user's motivation offset value falls back to within the preset offset threshold and the scheduling restrictions are restored.

[0068] It should be noted that in the formula structure involved in this scheme, dimensionless terms can be used as proportional or structural adjustment factors. When combined with quantities with units, they only play a role in numerical scaling and do not introduce new physical dimensions. Therefore, they will not change or confuse the overall unit system. This combination of "dimensionless terms and terms with units" can be understood as a composite structural expression commonly used in mathematical physics modeling. It conforms to the principle of dimensional consistency and has a clear physical interpretation basis.

[0069] Secondly, in the formula structure of this scheme, if multiple variables with different physical units are involved, including but not limited to time, mass or energy variables, their joint appearance is to express the collaborative modeling relationship of multiple physical mechanisms. Each variable can form a unified structure through function mapping, ratio combination or normalization adjustment, with clear units and clear meaning. The overall expression conforms to the principle of dimensional consistency and the conventional formula of engineering modeling.

[0070] In this solution, constants, weights, adjustment factors, threshold parameters, proportional coefficients, etc., are all adjustable control parameters for different application environments. Their values ​​depend on the target equipment configuration, data input characteristics, and performance optimization goals. During the implementation phase, they are set to converge within a reasonable range through model verification, performance constraints, or engineering calibration. Although these parameters do not have a unique preset value, they have clear adjustment logic and calculation paths. They belong to the deterministic setting process in engineering implementation. The purpose of this setting is to ensure that the solution is both universally adaptable and reproducible and operable, without affecting its technical clarity and feasibility.

[0071] In the perturbation construction module, the task flow of the target course is structurally decomposed, and each task content in the task flow is extracted and marked as the first task. Task content of each task ,in , Indicates the total number of tasks;

[0072] For each Set a random excitation parameter and combined with time windows Construct the excitation function :

[0073] ;

[0074] in This is the actual completion time. The excitation decay coefficient is used to dynamically generate excitation feedback for different completion points; among the values ​​of the excitation decay coefficient, when the task needs to approach... When a higher incentive is awarded for completion, a larger incentive should be chosen. Value, making exist If a steep drop in the vicinity can provide effective incentives while allowing for a more lenient completion timeframe, then the smaller one should be chosen. Value, to slow down the rate of excitation decay, therefore The size reflects the system's emphasis on "on-time completion" and should be determined by combining the course pace and incentive strategy objectives, through experience setting or fitting historical behavioral data, and should meet the following requirements. That's all; These represent the minimum and maximum values ​​of the excitation parameters, respectively; the time window. express The expected completion time;

[0075] For each Introducing a disturbance time offset term And reset the task trigger time. for: ; Indicates the first The perturbation offset for each task, which is used to change the trigger time of that task; It means that it "follows" a certain probability distribution; Indicates the interval The uniform distribution on the surface means that the disturbance offset is within this range. The probability of any value appearing within the range is the same; The maximum disturbance amplitude is set by the system and is used to limit the range of disturbance time; disturbance time offset term. The whole represents the first Disturbance offset for each task It is in the interval Values ​​obtained by random uniform sampling within the area;

[0076] based on , and , integrated into the first A disturbance task ;

[0077] All Arrange the tasks according to the original task flow to generate a perturbed task sequence: ;

[0078] The perturbation task sequence It is embedded in the task execution process of the target course, replacing the original task set.

[0079] In the data acquisition module, for each user who has registered for the target course, the sequence of perturbation tasks they participated in is extracted. ;

[0080] For each Record the corresponding user's behavioral data, including the task entry time. Task completion status Task skipped status Incentive response time ;

[0081] based on Composition of behavioral data units ;

[0082] Based on task trigger time For all behavioral data units Sort in ascending order to obtain a sequence of behavioral data ordered by time. ;

[0083] behavioral data sequences With the corresponding task trigger time set Combine and construct a task response sequence .

[0084] The task entry time Represented as:

[0085] ;

[0086] The task completion status Represented as:

[0087] ;

[0088] ;

[0089] The skip behavior state Represented as:

[0090] ;

[0091] The excitation response time Represented as:

[0092] ;

[0093] Task entry time Indicates the first The entry time of each task; Indicates task content The complexity of the task content The complexity includes the number of steps or the number of interactions. The number of interactions refers to the number of explicit behavioral interactions between the user and the task system during the completion of the perturbation task. The number of steps refers to the minimum number of operation flow nodes that the user must sequentially complete according to the task logic during the completion of the perturbation task; each node represents a specific execution step in the task's progression. In the expression, the stronger the incentive, that is The larger the value, the faster the task enters a response.

[0094] Task completion status Indicates whether the task is completed. Indicates the task completion status; otherwise, it is... ; It represents the rate at which the excitation function value decays over time. It is used to indicate how quickly the excitation function value decreases as the user completes the task with a delay relative to the task trigger time. The rate at which the excitation function value decreases with a delay means the degree to which the excitation function value decreases faster or slower the user completes the task later. This refers to the actual completion time. This indicates the upper limit of the allowed response time threshold for this task;

[0095] in Indicates whether to skip the task. This indicates that the user skipped the task; otherwise, it means... ;

[0096] Among them, the excitation response time The expression enters the time by the task With excitation function The reciprocal of the task execution cycle The sum of inner integrals reflects the cumulative response delay caused by the diminishing incentive during completion. Therefore, the overall response time represents the actual effective response time of the incentive after the user completes the task with low incentive perception. The incentive function... The value reflects the intensity of the excitation.

[0097] Calculate the task response latency value in the extraction module. :

[0098] ;

[0099] in Indicates the first The time delay from when a task is triggered to when it actually responds; It is an exponential function; The "delay penalty coefficient" is used to reflect the task under low-stimulation conditions; that is, the lower the stimulus, the lower the tolerance for response delay. The overall reflection of users' "weighted delayed behavior" under the influence of incentives;

[0100] The task completion rate change is calculated in the extraction module. :

[0101] ;

[0102] exist In the formula, is used to calculate the magnitude of the change in the task completion state in the perturbed task sequence, where Indicates the first The completion status of each task is represented by averaging the squared differences between the completion statuses of two adjacent tasks, indicating the degree of fluctuation in the completion rate of user behavior under a perturbation sequence; if the completion status changes frequently, The larger the value, the smaller the value, reflecting the stability of the user's task completion behavior; if If the completion rate of the task remains unchanged, its contribution to the overall change is zero, indicating that the behavior is stable during this stage.

[0103] Calculate the excitation response intensity in the extraction module. :

[0104] ;

[0105] in Indicates the interval of task duration The maximum value of the rate of change of the excitation function is selected internally; This represents the rate of change of the excitation function; This represents the overall cumulative intensity of the task's activation function; in The product of the two in the calculation process represents the degree of coupling between the maximum stimulus response rate and the overall stimulus perception, which is used to reflect the driving effect of stimulus on user behavior in the time domain;

[0106] In the extraction module, the three offset features are combined into a feature vector of user behavior:

[0107] ;

[0108] in A feature vector representing user behavior.

[0109] Construct the motivation offset function in the evaluation module:

[0110] ;

[0111] ;

[0112] in Indicates user The motivation offset value, It is the final evaluation result, used to measure the degree of consistency of the user's target behavior under a perturbed task sequence; Indicates allocation to user Disturbance mission The total number, i.e. the number of tasks completed by the user;

[0113] Among users The eigenvectors are represented as , Indicates user The average task response delay for all perturbation tasks, in time; This represents the maximum allowed delay threshold across all perturbed tasks, used to normalize time delay. This represents the normalized lag value, used to convert the unit to dimensionless, making it easier to measure in a unified manner with other indicators. Indicates user The range of change in task completion rate in a perturbed task sequence, i.e. the degree of fluctuation in completion quality, is dimensionless. Indicates user The average of the stimulus response intensity across all tasks, where and The units are all dimensionless because The unit is a dimensionless "point" or "amplitude of incentive";

[0114] in Indicates time Next, the One assigned to user Excitation function of the perturbation task The value; Indicates to Perform a sigmoid mapping to normalize its value to the range (0, 1), which is used to reflect the "sensitivity" or "instantaneous feedback strength" of the response; This represents the time from time 0 to the actual end time of each task. The excitation responses between the given points are integrated to calculate the overall response. Indicates to the user The average of the integral values ​​of the excitation responses of all perturbation tasks is used to represent the "mean" of the overall excitation response; For hyperbolic tangent function, in the above formula, the hyperbolic tangent function means that the mean of the response is projected onto the interval (1,1), which enhances the nonlinear discrimination ability, making high response more concentrated and low response more compressed.

[0115] It should be noted that this solution was developed based on an in-depth analysis of the shortcomings of the current community education resource scheduling system in identifying and responding to dynamic changes in user motivation. In particular, it addresses the critical issue that the inability to promptly identify, respond to, and adjust scheduling strategies for differences in user behavior during participation in educational tasks is a key problem. The solution introduces the "perturbation task sequence" mechanism as an innovative entry point.

[0116] The perturbation task sequence consists of several perturbation tasks containing target course task content, incentive parameters, and trigger times. This sequence is not a fixed static setting, but is formed by adding random perturbation based on a set time offset interval to the original task flow. This makes the specific task order and incentive feedback rhythm faced by each user when completing the course task uncertain, thus helping to expose the stability of user behavior and the persistence of motivation in the face of sudden or changing conditions. The perturbation construction module decomposes the original task flow in a structured way and configures a dynamic incentive feedback function for each task. The degree of attenuation of the incentive feedback reflects the response preference setting for the completion time.

[0117] Subsequently, a maximum perturbation amplitude limit is introduced to control the range of perturbation time variation, thereby relatively avoiding confusion in the course logic. All perturbation tasks are uniformly embedded in the course task flow to form a new schedulable execution path. This process ensures the non-linear expression of the influence of incentives on task behavior and the generation of behavioral profiles at the structural level.

[0118] The data acquisition module collects behavioral data for each user based on the above-mentioned perturbation task sequence, including multi-dimensional information such as whether the task was entered in time, whether it was completed, whether it was skipped, and the time of response incentives. By reorganizing the data units according to the actual trigger time of the perturbation task, a task response sequence with time process characteristics is formed, ensuring that the behavioral data has complete and traceable temporal sequence under the background of perturbation.

[0119] Based on this, the extraction module constructs an offset feature vector by calculating three indicators: response latency, completion rate fluctuation, and incentive response intensity. These feature dimensions describe the behavioral patterns of users in the task sequence from three perspectives: time response, behavioral stability, and incentive perception response capability.

[0120] The response latency metric reflects the user's speed of task entry under low-motivation conditions; the completion rate fluctuation expresses the consistent trend of behavior changes across tasks; and the incentive response intensity reflects the coupling degree between incentive intensity and behavior completion effect. After the three metrics are combined into a feature vector, they are input into the motivation offset function constructed by the evaluation module to output the user's current motivation offset value under the background of the perturbed task. The motivation offset function uniformly processes the normalized values ​​of various features and embeds a nonlinear mapping mechanism to distinguish the magnitude of behavioral differences under different motivational states, thereby ensuring that the offset value has strong discriminative power in terms of numerical value.

[0121] Based on the above determination of motivation offset values, the scheduling module sets up a behavior response mechanism: when a user's motivation offset value exceeds a preset threshold, a series of restrictive response measures will be triggered, including downgrading the scheduling priority, restricting registration, or additional verification. At the same time, the module does not adopt a one-way freeze mode, but automatically lifts the restrictions when the motivation offset value recovers to within the threshold, giving it dynamic control and self-recovery capabilities.

[0122] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1.A community education resource scheduling system based on multi-dimensional demand matching, comprising a disturbance construction module, a collection module, an extraction module, an evaluation module, and a scheduling module, characterized in that: The disturbance construction module is used to construct a disturbance task sequence of a target course, wherein the disturbance task sequence is composed of a plurality of disturbance tasks, and each disturbance task includes task content of the target course, a random incentive parameter, and a task triggering time. The collection module is used to collect behavior data of a user who signs up for the target course in the disturbance task sequence, wherein the behavior data includes a task entry time, a task completion status, a task skipping status, and an incentive response time; and the behavior data is reorganized according to the task triggering time in sequence to construct a task response sequence of the user. The extraction module extracts a plurality of offset features for measuring the consistency of the user's behavior based on the task response sequence, wherein the offset features include a task response delay value, a task completion rate change amplitude, and an incentive response strength; and the offset features are combined to construct a feature vector of the user's behavior. The evaluation module is used to input the feature vector into a motivation offset function to calculate a motivation offset value of the user, measure the target consistency of the user under the disturbance task sequence through the motivation offset value, and compare the motivation offset value with a preset threshold. The scheduling module includes: when the motivation offset value exceeds the preset threshold, a scheduling restriction is executed, wherein the scheduling restriction includes reducing the scheduling priority of the user for the incentive course, suspending the registration qualification of the user, or activating additional verification, until the motivation offset value of the user falls within the preset offset threshold to restore the scheduling restriction. In the disturbance construction module, the task flow of the target course is structurally decomposed, and each task content in the task flow is extracted and marked as the task content of the first th task , wherein , the total number of tasks is represented; For each , a random excitation parameter is set , combined with a time window , an excitation function is constructed ; wherein is the actual completion time, is the incentive decay coefficient; time window denotes the expected completion time point; denote the minimum and maximum values of the incentive parameter, respectively; for each introducing a perturbation time offset term , and resetting the task trigger time to: ; denotes the perturbation offset for the th task; denotes a uniform distribution over the interval ; is the maximum perturbation amplitude; Based on , and , the integration is the first disturbance task ;​ Arrange all Arrange according to the task order in the original task flow, generate the disturbance task sequence: ; The sequence of perturbation tasks is determined The tasks are embedded in the target course's task execution flow, replacing the original set of tasks. 2.The community education resource scheduling system based on multi-dimensional demand matching according to claim 1, characterized in that: In the collection module, for each user who registers to participate in a target course, the sequence of disturbance tasks participated by the user is extracted ; For each , record the behavior data of the corresponding user, the behavior data including task entering time , task completion state , task skipping state , and incentive response time ; based on constituting a behavior data unit ; Task-triggered time-based All behavior data units are sorted in ascending order to obtain a time-ordered behavior data sequence All behavior data units are sorted in ascending order to obtain a time-ordered behavior data sequence ; combining the behavior data sequence with the corresponding set of task trigger times to construct a task response sequence . 3.The community education resource scheduling system based on multi-dimensional demand matching according to claim 2, characterized in that: The task entry time is represented as: ; the task completion status is represented as: ; ; The skip behavior state is represented as: ; The excitation response time is represented as: ; wherein the task entry time represents the entry time of the th task; represents the complexity of the content of the task, the complexity of the content of the task including the number of steps or the number of interactions; wherein the task completion status indicates whether the task is completed, if indicates the task completion status, otherwise ; indicates the decay rate of the incentive function value at the time offset; is the actual completion time; indicates the upper limit of the response time threshold allowed for the task; wherein indicates whether the task was skipped, if indicates that the user skipped the task, otherwise . 4.The community education resource scheduling system based on multi-dimensional demand matching according to claim 3, characterized in that: Computing task response latency values in an extraction module : ; wherein represents the time delay from triggering to actual response of the nth task; is an exponential function; Computing a task completion rate change magnitude in an extraction module : ; Calculating excitation response strength in an extraction module : ; wherein represents the maximum value of the rate of change of the motivation function selected within the interval of the task duration; represents the rate of change of the motivation function; represents the overall cumulative intensity of the motivation function for the task;​ In the extraction module, the three offset features are combined into the feature vector of the user's behavior. ; wherein a feature vector representing user behavior. 5.The community education resource scheduling system based on multi-dimensional demand matching according to claim 4, characterized in that: In the evaluation module, the motivation offset function is constructed. ; ; wherein represents a user motivation offset value; represents a total number of disturbance tasks assigned to a user ; wherein the feature vector of the user is represented as , represents the average value of the task response latency values of the user for all perturbation tasks, in time units; represents the maximum latency threshold allowed in all perturbation tasks; represents the normalized latency value; represents the variation amplitude of the task completion rate of the user in the sequence of perturbation tasks; represents the average value of the intensity of the motivational response of the user in all tasks; wherein denotes the value of the incentive function for the perturbation task assigned to the user at time denotes a sigmoid mapping of denotes the integration of the incentive response from time 0 to the actual termination time of the task is the hyperbolic tangent function.​​​​​​

Images