Collaborative configuration method of quality education resources in and out of school based on big data analysis

Abstract

This application provides a method for the collaborative allocation of on-campus and off-campus quality education resources based on big data analysis, involving the field of big data analysis technology. The method includes: constructing a resource library based on internal and external resource sets; collecting user group characteristic information of student user groups; filtering multiple resources with labeled information based on individual user characteristics and educational needs; invoking a resource evaluation mechanism to perform fitness evaluation analysis on the target candidate resource set and generating a list of target candidate resources; and performing collaborative resource allocation for student user groups. This application addresses the technical problem of resource mismatch or waste caused by the lack of understanding and dynamic adaptation to personalized user needs in existing technologies. By combining individual user characteristics and educational needs with big data analysis, it achieves intelligent screening and efficient allocation of quality education resources, improving the accuracy of resource allocation.

Description

Technical Field

[0001] This application relates to the field of big data analytics technology, specifically to a method for collaborative allocation of on-campus and off-campus quality education resources based on big data analytics. Background Technology

[0002] Currently, the allocation of resources for holistic education both inside and outside of schools generally adopts a static allocation method based on fixed labels (such as grade level and subject category), lacking a deep understanding of individual student differences. This coarse-grained resource matching model ignores the diversity of individual interests, abilities, learning styles, and stage-specific needs, leading to resource mismatch. Furthermore, students' educational needs are dynamic and significantly influenced by factors such as cognitive development, shifts in interests, and learning progress. Traditional resource allocation lacks the ability to perceive and respond to the dynamic evolution of students' states in real time; resource allocation is often lagging behind and difficult to adjust promptly based on student feedback, reducing the relevance and effectiveness of resource allocation and thus affecting resource utilization.

[0003] In summary, existing technologies suffer from a lack of understanding of and dynamic adaptation to users' personalized needs, leading to resource misallocation or waste, which in turn affects the utilization rate of in-school and out-of-school quality education resources. Summary of the Invention

[0004] The purpose of this application is to provide a method for the collaborative allocation of on-campus and off-campus quality education resources based on big data analysis, in order to solve the technical problems of resource misallocation or waste caused by the lack of understanding of users' personalized needs and dynamic adaptation capabilities in existing technologies.

[0005] To achieve the above objectives, this application provides a method for collaborative allocation of on-campus and off-campus quality education resources based on big data analysis. The method includes: constructing a resource library based on an internal resource set and an external resource set, wherein the resource library includes multiple resources with labeled information; collecting user group characteristic information of a student user group, wherein the user group characteristic information includes individual user characteristics and educational demand characteristics; filtering the multiple labeled resources based on the individual user characteristics and the educational demand characteristics to obtain a target candidate resource set; invoking a resource evaluation mechanism to perform fitness evaluation analysis on the target candidate resource set and generating a target candidate resource list; and performing collaborative resource allocation for the student user group based on the target candidate resource list.

[0006] Optionally, the user's individual characteristics are used as a filtering constraint to filter the multiple resources with labeled information to obtain an initial candidate resource set; the educational demand characteristics are used as a filtering constraint to filter the initial candidate resource set to obtain the target candidate resource set; wherein, the educational demand characteristics include educational goals and educational cycles.

[0007] Optionally, a first resource is extracted from the initial candidate resource set, and the first resource corresponds to first annotation information; a first quality cultivation goal is extracted from the first annotation information, and a first fit degree between the first quality cultivation goal and the educational goal is obtained by comparison; a first usage cycle mean is extracted from the first annotation information, and it is determined whether the first usage cycle mean conforms to the educational cycle; if it conforms, the first resource is added to the target candidate resource set.

[0008] Optionally, multiple sub-goals are extracted from the educational goals; the multiple sub-goals are compared with the first quality cultivation goal to obtain a first goal coverage; a first sub-goal is extracted from the multiple sub-goals and traversed through the first quality cultivation goal to obtain a first preset coefficient; the first preset coefficient is subjected to a variation weighted calculation to obtain a first goal matching degree; the average of the first goal coverage and the first goal matching degree is taken as the first fitness degree.

[0009] Optionally, the first historical usage records of the first resource are collected, and the average of multiple historical usage periods in the first historical usage records is used as the average of the first usage period.

[0010] Optionally, any candidate resource in the target candidate resource set is obtained; a predetermined evaluation feature embedded in the resource evaluation mechanism is extracted, and the arbitrary candidate resource is evaluated and analyzed based on the predetermined evaluation feature to obtain an arbitrary fitness index; the target candidate resource set is sorted in descending order based on the arbitrary fitness index to obtain the target candidate resource list.

[0011] Optionally, the predetermined evaluation characteristics include at least sustainability, professional requirements, and usage costs.

[0012] Optionally, the student user group's behavior tracking records are collected; a first behavior in the behavior tracking records is analyzed to obtain a first learning effect score, wherein the first behavior corresponds to a first occurrence time; a score time series is formed by combining the correspondence between the first occurrence time and the first learning effect score; a comprehensive effect index is obtained by weighting the time-domain feature parameters of the score time series; wherein the comprehensive effect index is used to quantitatively evaluate the effect of the student user group using the target candidate resource list.

[0013] Optionally, an arbitrary user is obtained, and an arbitrary feature vector of the arbitrary user is collected; an arbitrary student user in the student user group is extracted, and an arbitrary user feature vector of the arbitrary student user is obtained; the arbitrary feature vector is compared with the arbitrary user feature vector to obtain a feature distance value; with the minimum feature distance value as the objective, the target student user is obtained by traversing and optimizing; the target record of the target student user in the behavior tracking record is analyzed to obtain a target index; the target index is used as the effect prediction result of the arbitrary user using the target candidate resource list.

[0014] Optionally, the arbitrary feature vector and the arbitrary user feature vector are sequentially processed into curves to obtain an arbitrary curve and an arbitrary user curve, respectively; the minimum value of the maximum distance between the arbitrary curve and the arbitrary user curve is obtained by comparison and used as the feature distance value.

[0015] The technical solution provided in this application has at least the following technical effects or advantages:

[0016] By constructing a resource library based on internal and external resource sets, wherein the resource library includes multiple resources with labeled information; collecting user group characteristic information of student users, wherein the user group characteristic information includes individual user characteristics and educational needs characteristics; filtering the multiple labeled resources based on the individual user characteristics and educational needs characteristics to obtain a target candidate resource set; invoking a resource evaluation mechanism to perform fitness evaluation analysis on the target candidate resource set and generating a target candidate resource list; and performing collaborative resource configuration for the student user group based on the target candidate resource list. In other words, by using big data analysis of mobile user group characteristic information, filtering from the resource library to obtain a target candidate resource set, using a resource evaluation mechanism to perform fitness evaluation analysis to obtain a target candidate resource list, and performing collaborative resource configuration, the accuracy and efficiency of resource configuration are improved.

[0017] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the collaborative allocation method for on-campus and off-campus quality education resources based on big data analysis, as proposed in this application.

[0020] Figure 2 This is a flowchart illustrating the process of determining the target candidate resource list in the collaborative allocation method for on-campus and off-campus quality education resources based on big data analysis in this application. Detailed Implementation

[0021] This application addresses the technical problem of resource misallocation or waste caused by the lack of understanding and dynamic adaptation to users' personalized needs in existing technologies by providing a method for collaborative allocation of on-campus and off-campus quality education resources based on big data analysis. By analyzing the characteristics of mobile phone user groups through big data analysis, a set of target candidate resources is obtained by filtering from a resource database. An adaptation evaluation mechanism is then used to obtain a list of target candidate resources, which are then collaboratively allocated, improving the accuracy and efficiency of resource allocation.

[0022] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0023] For examples, please refer to the appendix. Figure 1 This application provides a method for collaborative allocation of on-campus and off-campus quality education resources based on big data analysis, wherein the method specifically includes the following steps:

[0024] S100: Construct a resource library based on an internal resource set and an external resource set, wherein the resource library includes multiple resources with annotation information.

[0025] Specifically, the system integrates internal and external resource sets to create a resource repository containing all information on extracurricular education resources, helping users quickly filter and allocate resources. The internal resource set comprises extracurricular education resources provided within the school, including but not limited to curriculum materials, laboratory equipment, teacher resources, extracurricular activities, and sports facilities; these are provided by the school itself and used for daily teaching and extracurricular education activities. The external resource set includes extracurricular education resources provided outside the school, including resources from social partners, non-profit organizations, industry institutions, and corporate collaborations, such as external lectures, social practice bases, art training courses, and cultural exchange activities. The resource repository is an integrated platform for storing and managing various educational resources, containing basic information, annotation information, and possible use cases for resources, facilitating rapid retrieval, filtering, and allocation.

[0026] We comprehensively collected all on-campus resources for holistic education (such as courses, extracurricular activities, and practical training projects) and off-campus resources (such as social practice bases and training courses provided by partner companies). Each resource was categorized according to its type (such as courses, activities, experiments, and lectures) for subsequent labeling and management. Each collected resource needed to be labeled according to specific characteristics. For example, for course resources, we labeled the applicable grade level (e.g., primary, middle, and high school), educational objectives (e.g., improving logical thinking and cultivating practical skills), and teaching duration (e.g., 2 hours per class period). For social practice activities, we labeled the activity location (e.g., community center, museum), applicable student group (e.g., middle school students), and activity objectives (e.g., increasing social practice experience and enhancing teamwork skills).

[0027] The resource repository includes multiple resources with annotation information, meaning each resource comes with detailed annotation details. Through the integration and annotation of internal and external resources, resource management becomes more standardized, enabling resources to be searched and filtered based on multiple dimensions.

[0028] S200: Collect user group characteristic information of the student user group, wherein the user group characteristic information includes individual user characteristics and educational demand characteristics.

[0029] Specifically, basic student information is collected from campus management systems or questionnaires, including static information such as grade, gender, and date of birth, as well as dynamic information such as interests and personality. A student user group is a group of students who act as resource recipients, typically referring to a group of students with similar needs, backgrounds, or characteristics. Each student belongs to a larger group, and these students may share commonalities in grade, interests, abilities, etc. User group characteristic information encompasses various aspects of the student user group, used to describe the overall characteristics of the student group, including but not limited to personal background, interests, ability levels, learning habits, and educational needs.

[0030] User individual characteristics refer to a student's basic information and personalized features, including personal information such as age, gender, interests, and current ability level. Educational needs characteristics refer to a student's requirements for educational resources, including educational goals and educational cycles. Educational goals include learning preferences, subject-specific needs (such as improving math and language skills), and non-subject-specific needs (such as developing creativity and improving physical fitness). The collection of educational needs is typically done by combining data with students' learning progress, historical grades, and learning feedback. It can also be obtained through school hours, parental feedback, and student self-assessment. By organizing this data, a profile of students' educational needs is created. For example, some students may need to improve their math skills, while others may want to enhance their English listening and speaking abilities.

[0031] As students' learning progress and needs change, user group characteristics are dynamically updated. By collecting and analyzing the characteristics of student user groups, the needs of each student can be accurately identified, and the most suitable quality education resources can be intelligently recommended, improving the accuracy of resource allocation and avoiding resource mismatch.

[0032] S300: The user's individual characteristics and the educational needs characteristics are used to filter the multiple resources with labeled information to obtain a target candidate resource set.

[0033] Furthermore, this application S300 includes:

[0034] S310: Using the individual user characteristics as a filtering constraint, filter the multiple resources with annotation information to obtain an initial candidate resource set; S320: Using the educational demand characteristics as a filtering constraint, filter the initial candidate resource set to obtain the target candidate resource set; wherein, the educational demand characteristics include educational goals and educational cycles.

[0035] Specifically, user individual characteristics are used as filtering constraints. Multiple resources with labeled information in the resource library are filtered to obtain an initial candidate resource set. That is, all educational resources in the resource library (such as courses, activities, experimental equipment, etc.) are traversed, and the labeled information is matched with user individual characteristics. All resources in the resource library are initially filtered to select those that match the student's individual characteristics, generating the initial candidate resource set. The initial candidate resource set is a set of resources obtained after the first round of filtering; these resources conform to certain individual characteristics (such as grade level, interests, etc.) and are the result of filtering all resources in the resource library based on individual characteristics.

[0036] Using educational needs characteristics as screening constraints, the initial candidate resource set is further filtered to obtain the target candidate resource set. Educational needs characteristics include educational goals (the specific learning objectives students hope to achieve) and educational cycles (the expected timeframe for achieving the goals). The competency goals labeled in the initial candidate resources are extracted from the resource library. These labeled competency development goals must match the students' educational goals, and the course cycle cannot exceed the user's educational cycle.

[0037] For example, assume the user's individual characteristics are: age 13, first year of junior high school, current ability at a basic level (test score: 48 / 100), interest in public speaking and stand-up comedy, preference for in-person courses + video supplementary resources; educational goal is to master three types of structured speech skills (opening / body / closing), and the educational cycle is 30 days. Screening criteria include: age matching 12-15 years old, interest matching voice expression / public speaking, ability matching including resource difficulty of basic or beginner level, and format matching with preference for offline courses and support for video supplementation. After initial screening, the initial candidate resource set is shown in Table 1.

[0038] Table 1. Sample data of the initial candidate resource set

[0039]

[0040] The educational objectives were structurally decomposed, including: opening design ability, main structure expression ability, and closing conclusion ability. A1's objective coverage was calculated to be 100%, objective matching degree to be 95%, and final fit degree to be 97.5%, so it was retained. A2's quality development objectives did not match the average usage period, resulting in an objective coverage of 0%, so it was excluded. A3's quality development objectives only included structural training of the educational objectives, with an objective coverage of 33.3%, an objective matching degree to be 30%, and a final fit degree to be 31.65%, so it was excluded. A4's quality development objectives did not match the educational objectives, resulting in an objective coverage of 0%, and the usage period did not match the educational cycle, so it was excluded. A5's quality development objectives lacked a main structure, with an objective coverage of 66.7%, an objective matching degree to be 64%, and a final fit degree to be 65.35%, so it was retained pending further evaluation. Therefore, the most suitable candidate resource set for objectives was A1.

[0041] Through two rounds of screening, the final recommended resources are ensured to be highly matched with students' needs, avoiding resource waste and reducing the time students spend searching for suitable resources. Personalized, tailored resource recommendations are provided to students, improving the efficiency and effectiveness of resource allocation. Every student can access educational resources that meet their individual needs, enhancing their learning motivation and outcomes.

[0042] Furthermore, this application also includes the following steps:

[0043] S321: Extract the first resource from the initial candidate resource set, and the first resource corresponds to the first annotation information; S322: Extract the first quality cultivation goal from the first annotation information, and compare the first quality cultivation goal with the first educational goal to obtain the first fit; S323: Extract the first usage cycle average from the first annotation information, and determine whether the first usage cycle average conforms to the educational cycle; S324: If it conforms, add the first resource to the target candidate resource set.

[0044] Collect the first historical usage records of the first resource, and take the average of multiple historical usage periods in the first historical usage records as the average of the first usage period.

[0045] Extract multiple sub-goals from the educational goals; compare the multiple sub-goals with the first quality cultivation goal to obtain a first goal coverage; extract the first sub-goal from the multiple sub-goals and traverse the first sub-goal in the first quality cultivation goal to obtain a first preset coefficient; perform a variation weighted calculation on the first preset coefficient to obtain a first goal matching degree; take the average of the first goal coverage and the first goal matching degree as the first fitness degree.

[0046] Specifically, a first resource is extracted from the initial candidate resource set, and a resource is randomly selected as the first resource. This first resource includes first annotation information. The first annotation information contains specific details such as the resource's educational objectives, usage period, target audience, and course difficulty. The first competency development objective is extracted from the first annotation information; this is the main educational objective (such as improving a specific ability or quality) indicated by the first resource. This objective is usually closely related to the course content, such as language expression skills or logical thinking skills.

[0047] Educational goals can typically be broken down into multiple specific sub-goals, such as improving pronunciation clarity or enhancing logical expression. For example, improving language organization skills can be broken down into three sub-goals: structured thinking, clear expression, and logical rigor. Compare these sub-goals with the primary competency development goal to determine whether each sub-goal is covered by the primary competency development goal. Based on the degree of overlap, the coverage of the primary goal is determined. For example, if two out of three sub-goals overlap with the primary competency development goal, then the coverage of the primary goal is 2 / 3.

[0048] Extract the first sub-goal from a set of sub-goals, and iterate through it within the first competency development goal. Calculate the matching degree of the first sub-goal within the first competency development goal to obtain a first preset coefficient. The first preset coefficient reflects the consistency or similarity between the first sub-goal and the first competency development goal in the resources; the higher the value, the higher the matching degree. For example, if the first sub-goal of improving pronunciation clarity has a 70% matching degree with the first competency development goal of language organization ability, then the first preset coefficient is 0.7. Repeat this calculation process for the other sub-goals to obtain the preset coefficient for each sub-goal.

[0049] The primary goal matching degree is obtained by weighting the preset coefficients based on multiple sub-goals of the educational objective. More important sub-goals are given higher weights to ensure that the final matching degree better aligns with students' learning priorities. The primary goal matching degree is a quantified result of the degree of matching between the educational objective and the primary quality development objective, representing the degree of fit between resources and students' educational goals. A higher matching degree indicates a higher degree of goal fulfillment by the resources.

[0050] The mean of the first target coverage and the first target matching degree is calculated to obtain the first fit degree. For example, if the target coverage is 90% and the target matching degree is 76%, the final fit degree is 83%. The first fit degree is used to evaluate the fit between the first quality cultivation goal and the educational goal, and is used to assess the degree of matching between educational resources and educational goals.

[0051] Goal coverage refers to how many sub-goals of the educational goals are covered by the primary quality cultivation goals of the primary resource, with a focus on quantity; goal matching degree refers to the specific degree (i.e., quality) of matching between the primary quality cultivation goals of the primary resource and the sub-goals of the educational goals, with a focus on the depth and accuracy of the matching.

[0052] Extract the first historical usage record of the first resource, which is the record generated during the past use of the first resource, including the time, frequency, and specific usage period of the resource. Extract multiple historical usage periods from the first historical usage record, and calculate the average of multiple historical usage periods to obtain the first usage period average, which is used as the first usage period average of the first annotation information. For example, assuming that the first historical usage record includes three historical usage periods of 27 days, 28 days, and 32 days, the calculated first usage period average is 29 days.

[0053] The system determines whether the average usage period of the first resource aligns with the educational cycle. If not, the first resource is removed; if it does, the first resource and its corresponding first fit are added to the target candidate resource set. In other words, resources meeting the criteria are selected based on fit and educational cycle. If a resource meets both the fit requirements of the educational goals and the time requirements of the educational cycle, it is added to the target candidate resource set as one of the recommended resources. By comparing the resource's competency development goals with the student's educational goals, the system ensures that the recommended resources best meet the student's learning needs. By comparing the average usage period of the resource with the student's educational cycle, the system ensures that the resource does not exceed the student's learning time limit, thus avoiding delays in learning progress. Through resource screening, unqualified resources are quickly identified and removed, and more qualified resources are recommended to students, enhancing the personalized learning experience.

[0054] S400: The resource evaluation mechanism is invoked to perform fitness evaluation analysis on the target candidate resource set and generate a target candidate resource list.

[0055] Further details are attached. Figure 2 As shown, S400 of this application includes:

[0056] S410: Obtain any candidate resource from the target candidate resource set; S420: Extract the predetermined evaluation features embedded in the resource evaluation mechanism, and evaluate and analyze the arbitrary candidate resource based on the predetermined evaluation features to obtain an arbitrary fitness index; S430: Sort the target candidate resource set in descending order based on the arbitrary fitness index to obtain the target candidate resource list.

[0057] The predetermined evaluation characteristics include at least sustainability, professional requirements, and usage costs.

[0058] Specifically, the process involves acquiring any candidate resource from the target candidate resource set and then extracting each candidate resource for fitness evaluation and analysis. The resource evaluation mechanism is an embedded evaluation tool used for comprehensive resource evaluation and analysis. Evaluation dimensions include, but are not limited to, resource sustainability, professional requirements, and usage costs. The built-in resource evaluation mechanism module is invoked to obtain three main dimensions related to the resource: sustainability (whether the resource supports long-term use, can be continuously updated, and has a good technical maintenance background), professional requirements (whether professional knowledge or skills are required for effective use), and usage costs (evaluating the time, money, equipment, and other inputs required by students in using the resource). Sustainability values ​​range from 0 to 1, with higher values ​​indicating greater resource stability. Professional requirements values ​​also range from 0 to 1, with values ​​closer to 0 indicating good versatility and suitability for basic education, and values ​​closer to 1 indicating a high professional threshold. Usage costs are normalized to the 0-1 range for consistent measurement.

[0059] Based on predetermined evaluation characteristics, any candidate resource is evaluated and analyzed to obtain an arbitrary fitness index. Specifically, based on the extracted predetermined evaluation characteristics, each candidate resource is evaluated and analyzed one by one to obtain a sustainability index, a professionalism index, and a usage cost index. The arbitrary fitness index is calculated using the following formula: Fitness Index = a * Sustainability Index + b * Professionalism Index + c * Usage Cost Index, where a, b, and c are weighting coefficients, reflecting the influence weight of different indicators on the system, such as 0.4, 0.3, and 0.3. The arbitrary fitness index is a comprehensive score obtained after evaluating any candidate resource, representing the overall fitness of the resource relative to the predetermined evaluation characteristics.

[0060] The same evaluation is performed on other candidate resources in the target candidate resource set, resulting in multiple fitness indices. Based on the fitness index of each resource, they are sorted in descending order, with resources having higher fitness indices ranked higher, generating a list of target candidate resources. For example, resource R1 has a sustainability index of 0.9, a professionalism index of 0.3, and a usage cost of 100 yuan (normalized to 0.5), resulting in a calculated fitness index of 0.6; resource R2 has a sustainability index of 0.7, a professionalism index of 0.2, and a usage cost of 200 yuan (normalized to 0), resulting in a calculated fitness index of 0.34; resource R3 has a sustainability index of 0.6, a professionalism index of 0.6, and a usage cost of 50 yuan (normalized to 0.75), resulting in a calculated fitness index of 0.795. The ranking results are R3, R1, and R2. Using multi-dimensional indicators (such as sustainability, professionalism, and cost) transforms resource evaluation from subjective to objective and data-driven. By using resource assessment mechanisms and fitness index calculations, the ability of each resource to meet educational needs can be objectively evaluated, thereby improving the accuracy and effectiveness of educational resource allocation.

[0061] S500: Based on the target candidate resource list, perform resource collaborative configuration for the student user group.

[0062] Furthermore, this application S500 includes:

[0063] S510: Collect the behavior tracking records of the student user group; S520: Analyze the first behavior in the behavior tracking records to obtain a first learning effect score, wherein the first behavior corresponds to a first occurrence time; S530: Combine the correspondence between the first occurrence time and the first learning effect score to form a score time series; S540: Perform weighted calculation on the time domain feature parameters of the score time series to obtain a comprehensive effect index; wherein the comprehensive effect index is used to quantitatively evaluate the effect of the student user group using the target candidate resource list.

[0064] Specifically, resources are collaboratively allocated to the student user group based on the target resource candidate list. That is, resource allocation rules are formulated based on the characteristics of the target candidate resources and the needs of the student user group. Resources from the target resource candidate list are then allocated to each student in the student user group according to a specific strategy. This involves more than just distributing resources to individual students; it involves multi-dimensional matching of resources with user characteristics, including intelligent configuration behaviors such as cross-resource and cross-time period scheduling and sharing. The sorted resources are rationally allocated to the student user group to ensure that each student receives educational resources most suitable for their current learning stage and development goals. Different resource combinations are assigned to each student based on their individual characteristics (such as age, gender, interests) and learning progress.

[0065] Collect behavioral tracking records from student users, which includes various behavioral data generated by students while using educational resources, such as learning time, completed tasks, answer accuracy, and interaction frequency. Select the first behavior from the behavioral tracking records for analysis to evaluate the learning effect of that behavior. For example, if a student completes a self-test after learning a certain module and scores 80 points, then 80 points will be used as the learning effect score for that behavior.

[0066] The first occurrence time is the actual time when the first behavior occurs. Based on the occurrence time of each behavior and its corresponding learning outcome score, a score time sequence is formed. The score time sequence is a series of learning outcome scores recorded in chronological order of student behaviors, with each score associated with a corresponding behavior time, forming a time-sorted sequence of learning outcome data. For example, the learning outcome score for June 1st is 75, for June 8th it is 80, for June 15th it is 85, and for June 22nd it is 90.

[0067] Obtain the temporal characteristic parameters of the score time series, that is, the time-related characteristic parameters in the score time series data, such as mean, variance, trend, etc., which are used to describe the temporal characteristics of the learning effect. Extract key temporal features from the score time series, such as average score, trend indicators, volatility (standard deviation), etc.

[0068] These features are combined using a weighted method to obtain a comprehensive effectiveness index. For example: Comprehensive Effectiveness Index = x × Average Score + y × Trend Indicator + z × Volatility, where x, y, and z are determined based on expert experience and represent the weights of the average score, trend indicator, and volatility, respectively. For instance, if the average score is 82.5, the trend indicator is 5 points / week, and the volatility is 6.45, with weights of x = 0.5, y = 0.4, and z = 0.1, then the comprehensive effectiveness index is 43.9. The comprehensive effectiveness index is a numerical index obtained by comprehensively calculating all student learning behaviors and their corresponding learning outcomes, used to evaluate the overall learning effectiveness of students after using the target candidate resource list.

[0069] Furthermore, this application also includes the following steps:

[0070] S541: Obtain any user and collect any feature vector of the user; S542: Extract any student user from the student user group and obtain any user feature vector of the student user; S543: Compare the arbitrary feature vector with the arbitrary user feature vector to obtain a feature distance value; S544: With the minimum feature distance value as the objective, iterate through and optimize to obtain the target student user; S545: Analyze the target records of the target student user in the behavior tracking records to obtain a target index; S546: Use the target index as the effect prediction result of the arbitrary user using the target candidate resource list.

[0071] The arbitrary feature vector and the arbitrary user feature vector are sequentially processed into curves to obtain an arbitrary curve and an arbitrary user curve, respectively. The minimum value of the maximum distance between the arbitrary curve and the arbitrary user curve is obtained by comparison and is used as the feature distance value.

[0072] Specifically, this involves acquiring any user, i.e., randomly selecting a user. Collecting arbitrary feature vectors from this arbitrary user, i.e., numerical vectors of various attributes of this arbitrary user, such as age, learning ability, historical grades, and learning interests. Similarly, randomly selecting any student user from the student user group, and collecting arbitrary user feature vectors from this student user.

[0073] Curveification is performed on arbitrary feature vectors and arbitrary user feature vectors, arranging the feature vectors sequentially by dimension to form a broken line or continuous curve, mapping it to a curve function form. The purpose of curveification is to transform discrete feature values ​​(such as numbers) into smoother curves, making the trend of feature changes clearer. Curveification typically includes interpolation or smoothing processes, transforming the discrete values ​​in the feature vector into a continuous trend. Numerical interpolation algorithms (such as spline interpolation) are used to convert discrete feature points into curves.

[0074] After curve transformation, the arbitrary feature vector and the arbitrary user feature vector respectively generate smooth arbitrary curves and arbitrary user curves. The arbitrary curve represents the trend of any user's features changing with a certain dimension (such as time, task completion, etc.); the arbitrary user curve represents the curve of the target student's features changing with time, which can better show the user's changing trend.

[0075] By comparing any curve with any user's curve, the maximum distance between them is calculated, serving as a metric to measure the difference between the two users in the feature space. The Fraser distance is used to measure the maximum distance between two curves, taking into account the curve's path and shape to reflect geometric differences. This maximum distance calculation is performed on all arbitrary users and the target student user, yielding the maximum distance between each pair of users. The minimum maximum distance is then selected—the pair with the smallest distance between the farthest points on the two curves among all possible comparisons. The feature distance value, derived from calculating the minimum maximum distance, is used to quantify the feature similarity between two users. A smaller feature distance value indicates greater similarity in features between the two users, and vice versa.

[0076] The process iterates through the student user group, calculating the feature distance value for each student. The student with the smallest feature distance is selected as the target student user, meaning the student most similar to any other user is found. Target records for the target student user are then retrieved from behavioral tracking data, and the same analysis is performed to obtain the target index. In other words, behavioral data of the target student when using certain resources is extracted, and the actual learning effectiveness of the target student is calculated based on this data. The target index is calculated based on the target student's behavioral data to quantify and predict the student's effectiveness when using the target resources.

[0077] Using the target index of a target student user as a prediction result for the effect of any user using the same resource list, we can infer the expected learning effect of any user using the target candidate resource list. By calculating feature distance values, we can accurately find the target student most similar to any user's characteristics, thus providing a basis for personalized recommendations. By calculating the target index through behavioral tracking records, we can dynamically evaluate the effect of students using resources and optimize the allocation of educational resources. By using the target index to predict effects, we can accurately predict the learning effect of students using specific resources, ensuring the optimal allocation of educational resources.

[0078] In summary, the method for collaborative allocation of on-campus and off-campus quality education resources based on big data analysis provided in this application has the following technical effects:

[0079] By constructing a resource library based on internal and external resource sets, wherein the resource library includes multiple resources with labeled information; collecting user group characteristic information of student users, wherein the user group characteristic information includes individual user characteristics and educational needs characteristics; filtering the multiple labeled resources based on the individual user characteristics and educational needs characteristics to obtain a target candidate resource set; invoking a resource evaluation mechanism to perform fitness evaluation analysis on the target candidate resource set and generating a target candidate resource list; and performing collaborative resource configuration for the student user group based on the target candidate resource list. In other words, by using big data analysis of mobile user group characteristic information, filtering from the resource library to obtain a target candidate resource set, using a resource evaluation mechanism to perform fitness evaluation analysis to obtain a target candidate resource list, and performing collaborative resource configuration, the accuracy and efficiency of resource configuration are improved.

[0080] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0081] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for collaborative allocation of on-campus and off-campus quality education resources based on big data analysis, characterized in that: include: A resource library is constructed based on an internal resource set and an external resource set, wherein the resource library includes multiple resources with annotation information; Collect user group characteristic information of student user groups, wherein the user group characteristic information includes individual user characteristics and educational needs characteristics; By combining the individual user characteristics and the educational needs characteristics, the multiple resources with labeled information are filtered to obtain a target candidate resource set; The resource evaluation mechanism is invoked to perform fitness evaluation analysis on the target candidate resource set and generate a target candidate resource list; Based on the target candidate resource list, resources for the student user group are collaboratively configured from three perspectives: sustainability, professional requirements, and usage cost. The method of coordinating the user's individual characteristics and the educational needs characteristics to filter the multiple resources with labeled information yields a target candidate resource set, including: Using the individual user characteristics as filtering constraints, the multiple resources with annotation information are filtered to obtain an initial candidate resource set; Using the educational demand characteristics as screening constraints, the initial candidate resource set is screened to obtain the target candidate resource set; The characteristics of the educational needs include educational goals and educational cycles; The process of filtering the initial candidate resource set using the educational demand characteristics as a screening constraint to obtain the target candidate resource set includes: Extract the first resource from the initial candidate resource set, and the first resource corresponds to the first annotation information; Extract the first quality cultivation objective from the first annotation information, and compare it with the first quality cultivation objective to obtain the first fit degree between the first quality cultivation objective and the educational objective; Extract the average first usage period from the first annotation information, and determine whether the average first usage period conforms to the education cycle; If the conditions are met, the first resource is added to the target candidate resource set; The step of extracting the first quality cultivation objective from the first annotation information and comparing it with the first quality cultivation objective to obtain the first fit degree of the educational objective includes: Extract multiple sub-goals from the stated educational objectives; By comparing the multiple sub-goals with the first quality development goal, the coverage of the first goal is obtained; Extract the first sub-goal from the plurality of sub-goals, and iterate through the first sub-goal in the first quality cultivation goal to obtain the first preset coefficient; The first target matching degree is obtained by performing a variation-weighted calculation on the first preset coefficient; The average of the first target coverage and the first target matching degree is taken as the first fit degree.

2. The method for collaborative allocation of on-campus and off-campus quality education resources based on big data analysis as described in claim 1, characterized in that, Collect the first historical usage records of the first resource, and take the average of multiple historical usage periods in the first historical usage records as the average of the first usage period.

3. The method for collaborative allocation of on-campus and off-campus quality education resources based on big data analysis as described in claim 1, characterized in that, The resource evaluation mechanism is invoked to perform fitness evaluation analysis on the target candidate resource set and generate a list of target candidate resources, including: Obtain any candidate resource from the target candidate resource set; Extract the predetermined evaluation features embedded in the resource evaluation mechanism, and evaluate and analyze any candidate resource based on the predetermined evaluation features to obtain any fitness index; The target candidate resource set is sorted in descending order based on the arbitrary fitness index to obtain the target candidate resource list.

4. The method for collaborative allocation of on-campus and off-campus quality education resources based on big data analysis as described in claim 1, characterized in that, After configuring resource collaboration for the student user group based on the target candidate resource list, the process also includes: Collect behavioral tracking records of the aforementioned student user group; The first behavior in the behavior tracking record is analyzed to obtain a first learning effect score, wherein the first behavior corresponds to a first occurrence time; A score time series is formed by combining the correspondence between the first occurrence time and the first learning effect score; The time-domain characteristic parameters of the score time series are weighted and calculated to obtain the comprehensive effect index; The comprehensive effectiveness index is used to quantitatively evaluate the effectiveness of the student user group's use of the target candidate resource list.

5. The method for collaborative allocation of on-campus and off-campus quality education resources based on big data analysis as described in claim 4, characterized in that, Also includes: Obtain any user and collect any feature vector of the user; Extract any student user from the student user group and obtain any user feature vector of the student user. The feature distance value is obtained by comparing the arbitrary feature vector with the arbitrary user feature vector; The target student user is obtained by iterating through the search to find the minimum feature distance value. Analyze the target records of the target student users in the behavior tracking records to obtain the target index; The target index is used as the predicted effect of any user using the target candidate resource list.

6. The method for collaborative allocation of on-campus and off-campus quality education resources based on big data analysis as described in claim 5, characterized in that, The feature distance value is obtained by comparing the arbitrary feature vector with the arbitrary user feature vector, including: The arbitrary feature vector and the arbitrary user feature vector are sequentially subjected to curve conversion processing to obtain an arbitrary curve and an arbitrary user curve, respectively. The minimum value of the maximum distance between the arbitrary curve and the arbitrary user curve is obtained by comparison and is used as the feature distance value.

Images