Smart vocational education method and control system thereof

Abstract

The invention relates to the technical field of education. The invention discloses a smart vocational education method and a control system thereof, and the system comprises a data collection module which is used for collecting learning data of students, a data processing and analysis module which carries out the comprehensive scoring of the learning data of the students through a scoring formula, and a personalized recommendation module which is used for initializing the learning paths of the students, collecting the learning feedback of the students, and carrying out the comprehensive scoring of the learning data of the students. The system comprises a data acquisition module used for acquiring data, a data processing and analysis module used for adjusting a learning path until a preset learning target is reached, a teaching evaluation module used for comprehensively evaluating various indexes of students, and a system integration module used for integrating the data acquisition module, the data processing and analysis module, the personalized recommendation module and the teaching evaluation module. The intelligent vocational education control system provides highly personalized learning experience for students through student feature recognition, adaptive learning path adjustment and comprehensive teaching effect evaluation. The system can dynamically adjust the learning resources and activities according to the learning characteristics and demands of each student, thereby improving the learning efficiency and motivation.

Description

Technical Field

[0001] The present invention relates to the field of educational technology, and in particular to a smart vocational education method and a control system thereof. Background Art

[0002] At present, most smart education systems have realized the intelligence of education to a certain extent, and can provide personalized learning resources and suggestions based on learning data. However, there is no smart system specifically for vocational education, and they still have some significant limitations, which are exactly what the present invention aims to overcome.

[0003] Although existing systems can collect and process a large amount of learning data, they often lack the depth of vocational skills-related data analysis due to the lack of data collection and evaluation of students' professional abilities. Most of them use simple statistical analysis methods, which makes it difficult to deeply explore the hidden information and correlations behind the data. This results in the personalized recommendations and learning paths generated by the system often having a low degree of correlation with the improvement of vocational skills, and cannot meet the needs of students to truly improve their vocational skills. Although existing smart education systems have introduced AI technology to varying degrees, in actual applications, the degree of AI intelligence is still limited. The system can often only perform simple reasoning and judgment based on preset rules and models. The lack of training and accumulation of vocational education-related models limits the flexibility and adaptability of the system in the process of vocational skills teaching, and cannot give full play to the potential of AI technology to empower vocational education. Existing systems often rely on teachers' subjective judgments or simple quantitative indicators in the evaluation of teaching effects, and the evaluation is mainly the effect of knowledge transfer. This evaluation method is easily affected by human factors and is difficult to fully reflect students' learning situation and the degree of improvement in vocational skills. Therefore, a more objective evaluation method related to vocational skills is needed to accurately measure the teaching effect. Due to the above defects, the existing smart education system has a low match with students' vocational skills improvement and often performs poorly in terms of user experience. Students may not be able to obtain vocational skills learning resources and improvement suggestions that truly meet their needs, resulting in a decline in learning interest and motivation. At the same time, teachers may find it difficult to adjust their teaching strategies due to the lack of accurate teaching feedback, affecting the quality of teaching. Existing systems are often composed of multiple independent modules, the data exchange and collaborative work between modules are not smooth enough, and there is often a lack of modules associated with vocational skills application scenarios. This leads to low overall system efficiency and difficulty in achieving effective vocational education intelligence. Therefore, a more integrated and collaborative system architecture is needed to improve the overall performance and intelligence level of the vocational education system.

[0004] In summary, the existing smart education system has significant defects in data analysis depth, intelligence level, teaching effect evaluation, user experience and system integration, etc. These defects limit the application effect and promotion value of the system in the teaching process. Summary of the invention

[0005] In order to address the deficiencies of the prior art, the present application discloses a more comprehensive, intelligent and efficient smart vocational education method and control system.

[0006] The present invention discloses a smart vocational education method and a control system thereof, which include a data acquisition module for collecting students' learning data, a data processing and analysis module for defining variables for the collected learning data, comprehensively scoring the students' learning data through a scoring formula, comparing the comprehensive score with standard features to judge the students' learning portrait, a personalized recommendation module for initializing the students' learning path based on the results of the data processing and analysis module, collecting the students' learning feedback on the initialized learning path, and adjusting the learning path until the preset learning goal is achieved, a teaching evaluation module for collecting the learning feedback of the personalized recommendation module and comprehensively evaluating the students' various indicators, and a system integration module for integrating the data acquisition module, the data processing and analysis module, the personalized recommendation module and the teaching evaluation module.

[0007] Furthermore, the data processing and analysis module includes defining variables : Including classroom interaction records, homework completion, online test scores, learning time, learning preferences, professional skills and job capabilities, weight variables : The weight of each variable, weight variable The weight is determined by the teacher's score. There are k teachers, and each teacher has a weight for each variable x i Ratings ij , i represents the index of the variable number, j represents the index of the teacher number, where i=1, 2, ..., n, j=1, 2, ..., k, the average score of each variable, , the average score of each variable is normalized to obtain the weight, , the scoring formula calculates the comprehensive score, , the comprehensive score S and the standard feature T 特征 Compare and judge the students’ learning profile, S>T 特征 :Students are active learners, S<T 特征 :Students are passive learners, T 特征 The scoring formula is used to calculate the comprehensive score of each subject based on the median value of the comprehensive scores of active learners and passive learners in historical data. , S in various disciplines 学科 For comparison, select S学科 The lowest area as the learning difficulty, choose S 学科 The highest areas serve as points of interest.

[0008] Furthermore, the personalized recommendation module includes a preliminary adjustment submodule based on instant feedback and an optimal learning path search submodule based on the leapfrog algorithm.

[0009] Furthermore, the preliminary adjustment module based on immediate feedback in the personalized recommendation module includes initializing the student's learning path P based on the results of the data processing and analysis module. 0 , according to the current learning path P t Recommended Learning ResourcesA t , initially P t The actual P 0 , obtain students' recommended learning resources A t Feedback F t , F t :Students recommend resources A t Feedback, including satisfaction rating and completion status, F t When it is positive feedback, the adjusted learning path P t+1 for, , if F t The learning path after negative feedback adjustment is, ,α: adjustment parameter, used to control the speed of learning path adjustment, set according to previous educational experience, according to the adjusted learning path P t+1 Replace the original learning path P t , repeat the above process until the preset learning target threshold G or end value T is reached, and the current optimal learning path is found.

[0010] Furthermore, the optimal learning path search submodule of the personalized recommendation module based on the frog leaping algorithm includes defining variables, frog group L={l1, l2, ..., lN}: represents all frogs, where N is the number of frogs, subgroup B={b1, b2, ..., bM}: represents all subgroups, where M is the number of subgroups, local search step size δ and η: used to control the step size of local search, global search step size γ: used to control the step size of global search, iteration number U: the iteration number of the frog leaping algorithm, and the learning path P after preliminary adjustment t As a frog in a frog group, the frogs are divided into several subgroups. The frogs in each subgroup have similar learning paths. In each subgroup, the frogs perform local search based on their own and their neighbors' experience and update their own learning paths. , P t : Current frog learning path, P best : The optimal learning path in the current subgroup, P neighbor: The neighbor path of the current frog, δ and η: random numbers used to control the local search step size, the learning path P of the current frog t Will go to the optimal learning path P in the current subgroup best and neighbor path P neighbor The best learning path P in each subgroup is best Exchange information with the best paths of other subgroups, perform a global search among all subgroups, and find the optimal learning path in the entire frog group. , P t : Current frog learning path, P global_best : The global optimal learning path, γ: A random number used to control the global search step size. When the predetermined number of iterations U is reached or the termination condition is met, the optimal learning path P is output. optimal , the optimal learning path P found by the frog leaping algorithm optimal Update the current learning path P t .

[0011] Furthermore, the teaching evaluation module collects learning feedback from the personalized recommendation module and conducts a comprehensive evaluation of students’ multiple indicators. The evaluation indicators are: , including scores, rankings, learning progress, ability development and learning attitude, weight vector: The weight of each evaluation indicator is assigned according to teachers and historical experience, and the calculation of the comprehensive score E is ,E: Comprehensive score of students’ learning effect, result analysis, analysis of comprehensive score E, evaluate students’ learning outcomes, feedback the evaluation results to students and teachers, and adjust teaching strategies and learning paths according to the evaluation results P t+1 .

[0012] Furthermore, the data processing and analysis module also includes a data analysis submodule, which analyzes the student's variables through a machine learning algorithm and automatically updates the weight vector W.

[0013] Furthermore, the personalized recommendation module includes a real-time monitoring submodule, which collects students' learning behavior data in real time and dynamically adjusts the learning path P through a reinforcement learning algorithm. t+1 .

[0014] Furthermore, the smart vocational education control system includes a recommendation sub-module, which recommends personalized learning resources and activities based on students' historical learning data and current learning status. The data collection module collects students' learning data through educational terminals. The learning data includes classroom interaction records, homework completion status, online test scores, learning time, learning preferences, vocational skills and job capabilities.

[0015] Furthermore, H1: student feature identification, h11: collection of multiple variables of students, h12: determination of the weight of each variable by the teacher, and normalization to obtain the weight vector, h13: calculation of the comprehensive score of the students, h14: setting a personalized initial learning path P based on the comprehensive score and the learning characteristics of the students 0 , H2: Adaptive learning path adjustment, h21: According to the initial learning path P 0 Recommend learning resources and activities, h22: Collect student feedback on recommended learning resources or activities, h23: Dynamically adjust the learning path based on student feedback, h24: Update the current learning path P t ,H3: Teaching effectiveness evaluation,h31: Multiple evaluation indicators are used to evaluate students’ learning outcomes,h32: The weight of each evaluation indicator is determined based on the teacher and historical experience,h33: The comprehensive score of the students is calculated based on the weight and the value of each evaluation indicator,h34: Analyze the comprehensive score, compare the scores at different time points, identify the strengths and weaknesses of students, and feedback the evaluation results to students and teachers,h35: Adjust the teaching strategy and learning path based on the evaluation results.

[0016] Beneficial effects of the present invention: The intelligent vocational education control system provides students with a highly personalized learning experience through student feature recognition, adaptive learning path adjustment, and comprehensive teaching effect evaluation. The system can dynamically adjust learning resources and activities according to each student's learning characteristics and needs, thereby improving learning efficiency and motivation. At the same time, the multi-dimensional comprehensive evaluation system ensures the objective quantification of teaching effects, provides timely feedback and improvement suggestions for teachers and students, and further optimizes teaching strategies and learning paths. This system not only improves students' learning outcomes, but also enhances the overall quality and adaptability of vocational education.

[0017] A smart vocational education control system of the present invention provides students with a highly personalized learning experience through student feature recognition, adaptive learning path adjustment and comprehensive teaching effect evaluation. The smart vocational education control system can dynamically adjust learning resources and activities according to the learning characteristics and needs of each student, thereby improving learning efficiency and motivation. At the same time, the multi-dimensional comprehensive evaluation system ensures the objective quantification of teaching effects, provides timely feedback and improvement suggestions for teachers and students, and further optimizes teaching strategies and learning paths. The smart vocational education control system not only improves students' learning outcomes, but also enhances the overall quality and adaptability of vocational education.

[0018] The preliminary adjustment buns module based on instant feedback in the personalized recommendation module of this application, the system fine-tunes the learning path according to the feedback of students such as satisfaction, completion status, etc. This preliminary adjustment is based on instant feedback, and may not have reached the global optimum, but it can quickly respond to the needs of students. Based on the optimal learning path search submodule of the frog leaping algorithm, the frog leaping algorithm is used to further optimize the learning path and find the global optimal learning path. The frog leaping algorithm gradually optimizes the position of each frog, that is, each learning path, by exploring and comparing multiple learning paths, and finally finds the global optimal learning path. The preliminary adjustment buns module part based on instant feedback provides a preliminary optimized learning path, which serves as the initial position of the frog leaping algorithm. The frog leaping algorithm conducts a wider search on this basis to find the global optimal learning path. The preliminary adjustment buns module part based on instant feedback and the optimal learning path search submodule part based on the frog leaping algorithm together constitute a continuous optimization process. The preliminary adjustment buns module part based on instant feedback is responsible for short-term local adjustment, while the optimal learning path search submodule part based on the frog leaping algorithm is responsible for long-term global optimization. The combination of the two can more effectively find the optimal learning path.

[0019] The teaching effectiveness evaluation module of this application ensures that the optimal learning path of the personalized recommendation module is not only optimal in theory, but also effective in practical applications, and can be further adjusted and optimized according to the actual performance of students. Although the leapfrog algorithm can find the theoretically optimal learning path, the actual teaching effect still needs to be verified by students' performance. Through the evaluation of the teaching effectiveness evaluation module, students' feedback on learning resources or activities can be collected. This feedback can help further adjust the learning path to make it more in line with students' needs.

[0020] The teaching effect evaluation module of the present invention can comprehensively quantify students' learning outcomes through a multi-dimensional, multi-indicator comprehensive evaluation system. This evaluation method not only takes into account the degree of knowledge mastery, but also covers multiple aspects such as the level of ability development, learning attitude and learning progress, ensuring the comprehensiveness and objectivity of the evaluation. By regularly calculating the comprehensive score E of the students, the system can promptly discover the students' performance in various aspects, identify areas of strength and areas for improvement. This provides teachers with specific feedback and improvement suggestions to help them adjust their teaching strategies and better support students' learning. At the same time, students can also clearly understand their learning progress and shortcomings, so as to improve themselves in a more targeted manner. This comprehensive teaching effect evaluation helps to continuously optimize the teaching process, improve the overall quality of education, and ensure that every student can achieve the best learning effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1A flow chart of a smart vocational education method and its control system for implementing the present application. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the specific implementation modes of the present invention will be clearly and completely described below.

[0023] The present invention discloses a smart vocational education method and a control system thereof, which include a data acquisition module for collecting students' learning data, a data processing and analysis module for defining variables for the collected learning data, comprehensively scoring the students' learning data through a scoring formula, comparing the comprehensive score with standard features to judge the students' learning portrait, a personalized recommendation module for initializing the students' learning path based on the results of the data processing and analysis module, collecting the students' learning feedback on the initialized learning path, and adjusting the learning path until the preset learning goal is achieved, a teaching evaluation module for collecting the learning feedback of the personalized recommendation module and comprehensively evaluating the students' various indicators, and a system integration module for integrating the data acquisition module, the data processing and analysis module, the personalized recommendation module and the teaching evaluation module.

[0024] A smart vocational education control system of the present invention provides students with a highly personalized learning experience through student feature recognition, adaptive learning path adjustment and comprehensive teaching effect evaluation. The system can dynamically adjust learning resources and activities according to the learning characteristics and needs of each student, thereby improving learning efficiency and motivation. At the same time, the multi-dimensional comprehensive evaluation system ensures the objective quantification of teaching effects, provides timely feedback and improvement suggestions for teachers and students, and further optimizes teaching strategies and learning paths. The smart vocational education control system not only improves students' learning outcomes, but also enhances the overall quality and adaptability of vocational education.

[0025] As an implementation method, the data processing and analysis module includes: Defining variables : Including classroom interaction records, homework completion, online test scores, learning time, learning preferences, professional skills and job capabilities; Weight variables : The weight of each variable; Weight variables The weight is determined by the teacher's score: There are k teachers, each teacher has a i Ratings ij , i represents the index of the variable number, j represents the index of the teacher number, where i=1, 2, ..., n, j=1, 2, ..., k; The average score for each variable: ; The average score of each variable is normalized to obtain the weight: ; Scoring formula to calculate comprehensive score: ; Compare the comprehensive score S with the standard feature T to determine the student's learning profile: S>T 特征 : Students are active learners; S<T 特征 : Students are passive learners; T 特征 : The median value of the comprehensive scores of active learners and passive learners in historical data; The scoring formula calculates the overall score for each subject: ; S in various disciplines 学科 For comparison, select S 学科 The lowest area as the learning difficulty, choose S 学科 The highest areas serve as points of interest. Example

[0026] For a certain student, the variables and weights are as follows: x1: number of classroom interactions, x2: homework completion rate, x3: online test scores, x4: learning time.

[0027] The weight vector W={w1, w2, w3, w4} is calculated through teacher scoring.

[0028] The thresholds are, 特征 =60, T 难点 =40, T 兴趣 =80.

[0029] The values ​​of the variables for this student are: x1=70, x2=50, x3=60, x4=80.

[0030] Comprehensive score S=0.3×70+0.2×50+0.3×60+0.2×80=65; Because S>T 特征 , so the student is an active learner.

[0031] Calculate the overall score for each area: S(mathematics)=0.3×60+0.2×50+0.3×70+0.2×80=65; S(programming)=0.3×40+0.2×50+0.3×50+0.2×60=50; S(art)=0.3×70+0.2×60+0.3×80+0.2×90=76; S (science) = 0.3 × 80 + 0.2 × 70 + 0.3 × 90 + 0.2 × 85 = 82; Learning difficulty score S(dj): Select the lowest area as the learning difficulty. S(Programming) = 50 is the lowest, so programming is the learning difficulty.

[0032] Point of interest score S(ik): Select the highest field as the point of interest. S(Science) = 82 is the highest, so science is the point of interest.

[0033] The student feature recognition module of this application can accurately identify each student's learning characteristics, difficulties and interests by collecting and analyzing various variables of students, such as the number of classroom interactions, homework completion rate, online test scores, learning time, etc. This personalized recognition method enables the system to set the most suitable initial learning path for each student. 0 , thereby providing customized learning resources and activities. In this way, students can use their time and resources more effectively, focus on the areas where they need to improve the most, and get more support and development opportunities in areas of interest. This not only improves learning efficiency, but also enhances students' learning motivation and satisfaction.

[0034] As an implementation method, the initial adjustment module based on instant feedback in the personalized recommendation module includes: initializing the student's learning path P based on the results of the data processing and analysis module 0 , the initial learning path P 0 It is set based on the results of student feature identification. If the student is an active learner, then P 0 Contains more independent learning resources. Students have specific learning difficulties, then P 0 There will be more intensive exercises for this difficult point. If students have specific interests, then P 0 More resources related to this point of interest will be included. 0 It can be represented as a vector where each element represents a type of learning resource and its weight: ; p1: weight of self-learning resources; p2: the weight of programming intensive exercises; p3: weight of science-related resources; p4: weight of other resources; Active learner: Increase the weight of p1.

[0035] Programming difficulty: Increase the weight of p2.

[0036] Point of scientific interest: Increase the weight of p3.

[0037] According to the current learning path P t Recommended Learning ResourcesA t , initially P t The actual P 0 ; Get students' recommended learning resources A t Feedback F t ; Learning Resources A t This can include a variety of content and activities, such as online courses, hands-on projects, interactive discussions, gamified learning, video tutorials, reading materials, etc.

[0038] According to the current learning path P t Recommended Learning ResourcesA t : Initial setting, initialization, P 0 The initial learning path is set according to the student characteristics identification module. 0 It covers students' basic learning needs and interests.

[0039] Data collection, collecting students’ current learning status data, including completed learning resources, learning progress, test scores, interaction records, etc.

[0040] Learning path update: dynamically adjust the learning path based on students’ learning behavior and feedback t For example, if a student performs well in a certain area, the system will recommend more challenging content; if a student encounters difficulties, the system will provide more supportive resources.

[0041] Content-based recommendation, based on the student's current learning path P t Recommend similar or related learning resources based on the content in the article. Recommend resources that other similar students have successfully used based on their learning paths and feedback.

[0042] Specific recommended process: Matching learning paths: According to P t Choose the most appropriate resource type based on the specific content and students' learning status.

[0043] Weight calculation: Calculate the recommendation weight of each resource based on the importance of the resource, students' interests and learning needs.

[0044] Sorting and selection: Sort resources according to the recommendation weight and select the top-ranked resources for recommendation.

[0045] Example: Assume that the current learning path P tIt indicates that the student has difficulty in programming and shows interest in interactive discussion. Then, the system can recommend the following learning resources A t : Programming tasks provide some basic programming exercises to help students consolidate basic knowledge. Recommend some simple programming projects and gradually increase the difficulty.

[0046] Interactive discussion: Post programming-related questions on the forum and encourage students to participate in the discussion. Arrange an online programming seminar and invite industry experts to explain and answer questions.

[0047] Educational games, recommend a programming educational game that allows students to learn programming concepts during the game.

[0048] Provide some programming challenges and you can get rewards by completing the tasks.

[0049] Video tutorials, recommend some introductory micro-class videos on programming to help students understand basic concepts.

[0050] Provide some practical programming demonstration videos to show the specific operation steps.

[0051] Personalized tutoring: arrange one-on-one programming tutoring to help students solve specific problems. Use the intelligent tutoring system to provide personalized programming learning suggestions. In this way, the system can t Dynamically recommend the most suitable learning resources for students t , thereby improving learning outcomes and satisfaction.

[0052] F t :Students recommend resources A t feedback from the project manager, including satisfaction ratings and completion status; F t When it is positive feedback, the adjusted learning path P t+1 for: ; If F t The learning path adjusted for negative feedback is: ; α: Adjustment parameter used to control the speed of learning path adjustment. It is set according to previous educational experience. Usually, the value range of α is between 0 and 1, and the specific value can be adjusted according to the actual situation. The larger α is, the faster the system responds to feedback and the faster the learning path is adjusted; conversely, the smaller α is, the slower the system responds to feedback and the slower the learning path is adjusted. Set an initial value based on previous educational experience and actual application. For example, if historical data shows that students' learning needs change rapidly, a larger α can be set; if students' learning needs are relatively stable, a smaller α can be set. During the operation of the system, α can be dynamically adjusted according to students' feedback and learning effects. For example, if it is found that the system adjusts too quickly and causes students to be confused, α can be appropriately reduced; if it is found that the system adjusts too slowly and causes students to lose interest, α can be appropriately increased.

[0053] According to the adjusted learning path P t+1 Replace the original learning path P t , repeat the above process until the preset learning target threshold G or end value T is reached.

[0054] Learning status S t : Indicates the student status at time t, such as knowledge mastery, skill level, etc.

[0055] Action A t : Represents the learning resources or activities recommended by the system at time t.

[0056] Feedback t : Indicates students’ feedback on recommended learning resources or activities, such as satisfaction, completion status, etc.

[0057] Tuning parameter α: used to adjust the speed of the learning path.

[0058] Initial Learning Path P 0 :The initial learning path is set based on the comprehensive score of each subject calculated according to the scoring formula in the data processing and analysis module.

[0059] Current Learning Path t : The learning path at time t.

[0060] Learning Objective G: Preset learning objectives.

[0061] Update learning path: According to the adjusted P t+1 Update the learning path, i.e. P t Replace with P t+1 .

[0062] Repeat the above steps until the preset learning goal G is reached. Example

[0063] For a certain student, the variables and weights are as follows: x1: number of classroom interactions, x2: homework completion rate, x3: online test scores, x4: study time.

[0064] The weight vector W={w1, w2, w3, w4} obtained by teacher scoring. , assuming the threshold is: T 特征 =60, T 难点 =40, T 兴趣 =80.

[0065] For this student, the value of the variable is: x1=70, x2=50, x3=60, x4=80; Comprehensive score S=0.3×70+0.2×50+0.3×60+0.2×80=65; Because S>T 特征 , so the student is an active learner.

[0066] Calculate the overall score for each area: S(mathematics)=0.3×60+0.2×50+0.3×70+0.2×80=65; S(programming)=0.3×40+0.2×50+0.3×50+0.2×60=50; S(art)=0.3×70+0.2×60+0.3×80+0.2×90=76; S (science) = 0.3 × 80 + 0.2 × 70 + 0.3 × 90 + 0.2 × 85 = 82; Learning difficulty score S(dj): Select the lowest field as the learning difficulty. S(programming) = 50 is the lowest, so programming is the learning difficulty.

[0067] Point of interest score S(ik): Select the highest field as the point of interest. S(Science) = 82 is the highest, so science is the point of interest.

[0068] Initial Learning Path P 0 Settings; There are the following types of resources: independent learning resources R1, programming intensive exercises R2, science-related resources R3.

[0069] Initial Learning Path P 0 Represented as a vector where each element represents the weight of a resource: P0=[p1, p2, p3], active learners: increase the weight of p1, programming difficulty: increase the weight of p2, scientific interest point: increase the weight of p3.

[0070] Specific initial path settings: Active learner: P 0 =[0.5, 0.3, 0.2], Programming difficulty: P 0 =[0.3, 0.5, 0.2], scientific interest point: P 0 =[0.3, 0.2, 0.5].

[0071] Learning objective G: Comprehensive score threshold G, G=80.

[0072] When the student's overall score S t When ≥G, the learning goal is considered to have been achieved.

[0073] End condition T End: When students do not achieve the preset learning goal S t <G. Update the new learning path. After a certain number of learning cycles, such as 100 iterations, if the comprehensive score does not improve significantly, for example, the score change is less than 1% for 10 consecutive iterations, the learning process is terminated.

[0074] The personalized recommendation module of this application can dynamically optimize students' learning paths by continuously collecting students' feedback on recommended learning resources or activities, such as satisfaction and completion status, through the adaptive learning path adjustment module. This real-time feedback mechanism enables the system to flexibly adjust the learning content and progress according to the students' actual performance and needs. If a student performs well in a certain field, the system will recommend more challenging content to further enhance their abilities; if a student encounters difficulties in a certain field, the system will provide more support and auxiliary resources. This personalized adjustment not only improves learning efficiency, but also enhances students' learning experience and participation. By continuously optimizing the learning path, the system can ensure that students are always in the most suitable learning state for themselves, so as to achieve their learning goals more effectively.

[0075] As an implementation method, the optimal learning path search submodule of the personalized recommendation module based on the frog leaping algorithm includes: Define variables: Frog population L = {l1, l2, ..., lN}: represents all frogs, where N is the number of frogs; Subgroup B = {b1, b2, ..., bM}: represents all subgroups, where M is the number of subgroups; Local search step size δ and η: used to control the step size of local search; Global search step size γ: used to control the step size of global search; Iteration number U: the number of iterations of the frog leaping algorithm; The learning path P after preliminary adjustment t As a frog among frogs, The frogs are divided into several subgroups. The frogs in each subgroup have similar learning paths. In each subgroup, the frogs perform local searches based on their own and their neighbors' experiences and update their own learning paths. ; P t : Current frog learning path; P best : The optimal learning path in the current subgroup; P neighbor : The neighbor path of the current frog; δ and η: random numbers used to control the local search step size; Current Frog Learning Path P t Will go to the optimal learning path P in the current subgroup best and neighbor path P neighbor Move in the direction of to find a better learning path; The best learning path P in each subgroup best Exchange information with the best paths of other subgroups, perform a global search among all subgroups, and find the optimal learning path in the entire frog group: ;

[0076] P t : Current frog learning path; P global_best : The globally optimal learning path; γ: random number used to control the global search step size; When the predetermined number of iterations U is reached or the termination condition is met, the optimal learning path P is output optimal ; The optimal learning path P found by the frog leaping algorithm optimal Update the current learning path P t .

[0077] The optimal learning path search submodule of the personalized recommendation module based on the frog leaping algorithm includes: 1. Initialize the frog group Definition: Transform the student's learning path P t As a frog in a group of frogs. Each frog represents a possible learning path.

[0078] For example: suppose we have 50 different learning paths, then there are 50 frogs.

[0079] 2. Grouping Definition: Divide the frogs into several subgroups, and the frogs in each subgroup have similar learning paths.

[0080] For example: If there are 50 frogs, they can be divided into 5 subgroups, each with 10 frogs.

[0081] 3. Local Search Goal: Within each subgroup, frogs perform local search based on their own experience and that of their neighbors, updating their learning paths.

[0082] formula: ; P t : The current learning path of the frog.

[0083] P best : The optimal learning path in the current subgroup.

[0084] P neighbor : The neighbor path of the current frog. Randomly select a frog from the subgroup as a neighbor.

[0085] δ and η: random numbers used to control the local search step size.

[0086] The optimal learning path P in the current subgroup best : Each frog, i.e., learning path, has an initial position P. t The position P of each frog t It corresponds to a specific solution, i.e., a learning path, and each solution has a fitness value, which indicates how good the solution is. All frogs are divided into several subgroups. The frogs in each subgroup have similar learning paths. In each subgroup, the fitness value of each frog is calculated. The fitness value is usually calculated through an objective function, which reflects the quality of the learning path. For example, in the field of education, the fitness value can be a comprehensive score of indicators such as students' academic performance, speed of completing tasks, and satisfaction. In each subgroup, find the frog with the highest fitness value, and its corresponding learning path is P. best .

[0087] δ and η are random numbers: δ control to P best The step length of the directional jump: The larger δ is, the more likely the frog will move toward the optimal path P in the current subgroup. best The larger the jump step size, the smaller the jump step size.

[0088] η control to P neighbor The step length of the directional jump: The larger η is, the more likely the frog is to move toward the neighbor path P. neighbor The larger the jump step size, the smaller the jump step size.

[0089] A frog stands at the current position P t , it will move to the best position P in the current subgroupbest Jump a distance .

[0090] At the same time, it also refers to the neighbor's position P neighbor , and jump a distance .

[0091] In this way, the frog can find a new position P t+1 , this new location may be better than the original one.

[0092] 4. Information Sharing Goal: The best learning path P in each subgroup best Exchange information with the best paths of other subgroups to improve the global search capability.

[0093] The best path in each subgroup will be shared with other subgroups, so that all subgroups can benefit from the best paths of other subgroups, thereby improving the overall search efficiency.

[0094] 5. Global Search Goal: Perform a global search among all subgroups to find the optimal learning path for the entire frog population.

[0095] formula: ;

[0096] P t : The current learning path of the frog.

[0097] P global_best : The globally optimal learning path.

[0098] γ: A random number used to control the global search step size.

[0099] Popular explanation: A frog stands at the current position P t , it will move to the global optimal position P global_best Jump a distance In this way, the frog can find a new position P t+1 , this new location may be better than the original one.

[0100] 6. Termination Conditions Definition: When the predetermined number of iterations U is reached or other termination conditions are met, the search process is stopped and the optimal learning path P is output. optimal .

[0101] If the number of iterations U is set to 100, then when 100 iterations are reached, the search is stopped and the optimal learning path is output.

[0102] 7. Update learning paths Definition: The optimal learning path P found by the leapfrog algorithm optimal Update the current learning path P t .

[0103] If a new optimal learning path is found through the leapfrog algorithm, the current learning path is replaced by this new path.

[0104] As an implementation method, the personalized recommendation module includes a preliminary adjustment submodule based on instant feedback and an optimal learning path search submodule based on the leapfrog algorithm.

[0105] The main purpose of the frog leaping algorithm is to optimize the learning path P through local search and global search. t , to find the optimal learning path P optimal . And P t+1 It refers to the learning path that is updated in each iteration.

[0106] Initial Adjustment Based on Instant Feedback Bun module makes initial adjustments to the learning path based on student feedback; formula: (positive feedback); (negative feedback); At this stage, the system fine-tunes the learning path based on student feedback such as satisfaction, completion, etc. This initial adjustment is based on immediate feedback and may not yet reach the global optimum, but it can quickly respond to student needs.

[0107] The optimal learning path search submodule based on the frog leaping algorithm uses the frog leaping algorithm to further optimize the learning path and find the global optimal learning path; formula: ; ; At this stage, the leapfrog algorithm gradually optimizes the position of each frog, that is, each learning path, by exploring and comparing multiple learning paths, and finally finds the globally optimal learning path.

[0108] Relationships and collaboration: Relationship: In the initial adjustment of the bun module based on immediate feedback, P t+1 is the preliminary adjustment result based on immediate feedback; in the optimal learning path search submodule based on the frog leaping algorithm, P t+1 It is the result of further optimization through the frog leaping algorithm. Both are aimed at improving the learning path, but the optimization of the optimal learning path search submodule based on the frog leaping algorithm is more comprehensive and in-depth.

[0109] Collaboration: The initial adjustment of the bun module based on immediate feedback provides a preliminary optimized learning path, which serves as the initial position of the frog leaping algorithm. The frog leaping algorithm conducts a more extensive search on this basis to find the global optimal learning path.

[0110] Real-time response: Preliminary adjustment of the P of the bun module based on immediate feedback t+1 Ability to quickly respond to students' immediate feedback to ensure that learning paths can be adjusted in a timely manner, improving student satisfaction and engagement.

[0111] Initializing the frog leaping algorithm: Preliminary adjustments based on immediate feedback The preliminary adjustments to the bun module provide a good starting point for the frog leaping algorithm. Without this preliminary adjustment, the frog leaping algorithm may require more time and iterations to converge to a better solution.

[0112] Continuous optimization: The initial adjustment module based on instant feedback and the optimal learning path search module based on the frog leaping algorithm together constitute a continuous optimization process. The initial adjustment module based on instant feedback is responsible for short-term local adjustments, while the optimal learning path search module based on the frog leaping algorithm is responsible for long-term global optimization. The combination of the two can more effectively find the optimal learning path.

[0113] As an implementation method, the teaching evaluation module includes: collecting learning feedback from the personalized recommendation module and conducting a comprehensive evaluation of multiple indicators of students; Evaluation indicators: , including scores, rankings, learning progress, ability development and learning attitude; Weight vector: , the weight of each evaluation indicator; assigned according to teachers and historical experience.

[0114] Calculation of comprehensive score E: ; E: Comprehensive evaluation of students’ learning outcomes.

[0115] y1, y2, …, ym represent multiple aspects that need to be evaluated, such as knowledge mastery, ability development level, learning attitude, learning progress and other related indicators.

[0116] Student feature identification: Through the analysis in the data processing and analysis module, the student's learning features, active learners or passive learners, learning difficulties and interests are obtained. These features will be used to set the initial learning path P 0 and adjust the subsequent learning path P t Through the adaptive learning path adjustment algorithm in the personalized recommendation module, we continuously adjust the learning path based on students’ feedback. tAdjust the learning path P t+1 , to optimize the recommendation of learning resources.

[0117] Specific steps for teaching effectiveness evaluation: Data collection, collects students' evaluation indicator data Y at different stages, including online test scores, homework completion rate, classroom interaction records, learning time, professional skills test scores, project completion status, course completion, etc.

[0118] Comprehensive score calculation, using weight vector V to calculate the comprehensive score E: ;

[0119] Result analysis, analyze the comprehensive score E, and evaluate students' learning outcomes. Compare the comprehensive scores at different time points to understand students' learning progress. Identify where students perform well and where they need improvement.

[0120] Feedback and adjustment: Feedback the evaluation results to students and teachers, and provide suggestions for improvement. Adjust teaching strategies and learning paths based on the evaluation results to further optimize the learning experience. Example

[0121] For a certain student, the variables and weights are as follows: Knowledge mastery level y1: Online test score 70 points, homework completion rate 80%.

[0122] Ability development level y2: Vocational skills test score 60 points, project completion 90%.

[0123] Learning attitude y3: Classroom interactions 50 times, learning time 10 hours / week.

[0124] Learning progress y4: Course completion 70%, learning plan execution 85%.

[0125] The weight vectors are: V1=0.3, V2=0.2, V3=0.2, V4=0.3.

[0126] The calculation of the comprehensive score E is as follows: E = 0.3 × 75 + 0.2 × 75 + 0.2 × 75 + 0.3 × 77.5 = 76.25. The calculated comprehensive score is compared with the preset initial comprehensive score to arrange the students' study. The initial comprehensive score can be the comprehensive score of the previous month, the previous quarter, and the previous semester.

[0127] The teaching effectiveness evaluation module of this application can comprehensively quantify students' learning outcomes through a multi-dimensional, multi-indicator comprehensive evaluation system. This evaluation method not only takes into account the degree of knowledge mastery, but also covers multiple aspects such as the level of ability development, learning attitude and learning progress, ensuring the comprehensiveness and objectivity of the evaluation. By regularly calculating the student's comprehensive score E, the system can promptly discover the student's performance in various aspects, identify areas of strength and areas for improvement. This provides teachers with specific feedback and improvement suggestions to help them adjust their teaching strategies to better support students' learning. At the same time, students can also clearly understand their learning progress and shortcomings, so as to improve themselves in a more targeted manner. This comprehensive teaching effectiveness evaluation helps to continuously optimize the teaching process, improve the overall quality of education, and ensure that every student can achieve the best learning results.

[0128] The teaching effectiveness evaluation module of this application ensures that the optimal learning path of the personalized recommendation module is not only optimal in theory, but also effective in practical applications, and can be further adjusted and optimized according to the actual performance of students. Although the leapfrog algorithm can find the theoretically optimal learning path, the actual teaching effect still needs to be verified by students' performance. Through the evaluation of the teaching effectiveness evaluation module, students' feedback on learning resources or activities can be collected. This feedback can help further adjust the learning path to make it more in line with students' needs.

[0129] As an implementation method, the data processing and analysis module also includes a data analysis submodule, which analyzes the student's variables through a machine learning algorithm and automatically updates the weight vector W. Through the machine learning algorithm, the system can dynamically adjust the weight vector W according to the student's learning data, so that the weight is more in line with the actual situation. This can better reflect the student's actual learning situation than static teacher evaluation and historical experience evaluation. The data analysis submodule can analyze the student's learning behavior and performance in real time, and automatically update the weight vector W, so as to more accurately identify the student's learning characteristics, difficulties and points of interest. This makes personalized recommendations more accurate and improves the matching degree of learning resources and activities. The data analysis submodule can analyze the student's learning behavior and performance in real time, and automatically update the weight vector W, so as to more accurately identify the student's learning characteristics, difficulties and points of interest. This makes personalized recommendations more accurate and improves the matching degree of learning resources and activities. The data analysis submodule automatically processes a large amount of student data through a machine learning algorithm, which reduces the workload of teachers manually analyzing and adjusting weights, allowing teachers to devote more energy to teaching design and student counseling.

[0130] As an implementation method, the personalized recommendation module includes a real-time monitoring submodule, which collects students' learning behavior data in real time and dynamically adjusts the learning path P through a reinforcement learning algorithm.t+1 The real-time monitoring submodule can continuously collect students’ learning behavior data, such as learning progress, completion status, interaction records, etc., and quickly adjust the learning path through reinforcement learning algorithms. t+1 This instant feedback mechanism enables the system to quickly respond to students' needs and changes. The reinforcement learning algorithm can find the optimal learning path through continuous trial and error and optimization, thereby maximizing students' learning efficiency. Students can achieve the best learning effect in the shortest time.

[0131] As an implementation method, the smart vocational education control system includes a recommendation submodule, which recommends personalized learning resources and activities based on the student's historical learning data and current learning status. The recommendation submodule can provide highly personalized learning resources and activities based on the student's historical learning data and current learning status. This precise recommendation enables each student to obtain the most suitable learning content for him or her, improving the pertinence and effectiveness of learning. Recommendations are not limited to traditional learning resources, but also include content such as games and entertainment, which increases the fun and diversity of learning and helps to improve students' interest and participation in learning. The recommendation submodule can provide suitable learning resources and activities based on the student's learning progress and performance to maintain the student's learning motivation. For example, when a student performs well in a certain field, the system can recommend more challenging content; conversely, if the student encounters difficulties, the system can provide more support and auxiliary resources. By introducing elements such as games and entertainment, the recommendation submodule can make the learning process more interesting and interactive, increase students' sense of participation and achievement, and thus enhance learning motivation.

[0132] As an implementation method, the data collection module collects students' learning data through educational terminals. The learning data includes classroom interaction records, homework completion status, online test scores, learning time, learning preferences, professional skills and job capabilities.

[0133] As an implementation method, the smart vocational education method includes: H1: Identification of student characteristics: h11: Collect various variables of students; h12: The teacher determines the weight of each variable and normalizes it to obtain the weight vector; h13: Calculate the comprehensive score of students; h14: Set a personalized initial learning path based on the comprehensive score and the student’s learning characteristics 0 ; H2: Adaptive learning path adjustment: h21: According to the initial learning path P 0 recommend learning resources and activities; h22: Collect student feedback on recommended learning resources or activities; h23: Dynamically adjust learning paths based on student feedback; h24: Update the current learning path P t ; H3: Teaching effectiveness evaluation: h31: Multiple evaluation indicators assess students’ learning outcomes; h32: Determine the weight of each evaluation indicator based on teachers and historical experience; h33: Calculate the comprehensive score of the student based on the weight and the value of each evaluation indicator; h34: Analyze composite scores, compare scores at different points in time, identify students’ strengths and weaknesses, and provide feedback to students and teachers; h35: Adjust teaching strategies and learning paths based on assessment results.

[0134] Student feature identification (H1), by collecting multiple variables of students and having the teacher determine the weight of each variable, the system can more accurately identify students’ learning characteristics, difficulties and interests. This makes the initial learning path P 0 It better meets the needs of each student and provides a personalized learning experience. The calculation of comprehensive scores helps the system quickly identify students' strengths and weaknesses, thereby setting the most appropriate learning path for students in the initial stage, avoiding ineffective learning content and improving learning efficiency. Personalized learning paths can better stimulate students' interest and motivation in learning, because students can focus on areas that they are interested in or need to improve.

[0135] Adaptive learning path adjustment (H2): By collecting students’ feedback on recommended learning resources or activities, the system can dynamically adjust the learning path to ensure that the learning content always matches the students’ needs. This real-time feedback mechanism makes the learning process more flexible and efficient. The current learning path is continuously updated based on student feedback. t , the system can gradually optimize the learning path and improve the learning effect. This continuous optimization process helps students make continuous progress in the learning process. Adaptive learning path adjustment enables students to obtain learning resources that are more suitable for them, thereby improving students' learning satisfaction and participation.

[0136] Teaching effectiveness evaluation (H3) evaluates students' learning outcomes through multiple evaluation indicators. The system can fully understand students' performance in terms of knowledge mastery, ability development level, learning attitude and learning progress, ensuring the comprehensiveness and objectivity of the evaluation. The weight of each evaluation indicator is determined based on teachers and historical experience, making the evaluation results more scientific and reliable. This method combines expert experience and data-driven methods to improve the accuracy of the evaluation. By analyzing the comprehensive scores and comparing the scores at different time points, the system can promptly identify students' strengths and weaknesses and feedback the evaluation results to students and teachers. This helps students and teachers to adjust their learning strategies in a timely manner and further optimize the learning path.

[0137] The smart vocational education approach provides a highly personalized learning experience and enhances learning efficiency and motivation through student feature identification, adaptive learning path adjustment, and teaching effect evaluation. The system's real-time feedback and dynamic adjustment mechanism ensure that the learning path always matches the needs of students, while the comprehensive teaching effect evaluation provides students and teachers with timely feedback and improvement suggestions. These features work together to improve not only students' learning outcomes, but also the overall quality of education.

[0138] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all these improvements and changes should fall within the scope of protection of the appended claims of the present invention.

Claims

1. A smart vocational education method and control system, characterized in that: include: Data collection module, used to collect students’ learning data; The data processing and analysis module defines variables for the collected learning data, comprehensively scores the students' learning data through the scoring formula, and compares the comprehensive scores with the standard features to determine the students' learning profile; The personalized recommendation module initializes the student's learning path based on the results of the data processing and analysis module, collects the student's learning feedback on the initialized learning path, and adjusts the learning path until the preset learning goal is achieved; The teaching evaluation module collects learning feedback from the personalized recommendation module and conducts a comprehensive evaluation of students' multiple indicators; The system integration module is used to integrate the data acquisition module, data processing and analysis module, personalized recommendation module and teaching evaluation module.

2. According to the smart vocational education method and control system described in claim 1, it is characterized by: Data processing and analysis modules include: Defining variables : Including classroom interaction records, homework completion, online test scores, learning time, learning preferences, professional skills and job capabilities; Weight variables : The weight of each variable; Weight variables The weight is determined by the teacher's score: There are k teachers, each teacher has a i Ratings ij , i represents the index of the variable number, j represents the index of the teacher number, where i=1, 2, ..., n, j=1, 2, ..., k; The average score for each variable: ； The average score of each variable is normalized to obtain the weight: ； The scoring formula calculates the comprehensive score: ； Combine the comprehensive score S with the standard feature T 特征 Compare and judge the students’ learning portraits: S>T 特征 : Students are active learners; S<T 特征 : Students are passive learners; T 特征 : The median value of the comprehensive scores of active learners and passive learners in historical data; The scoring formula calculates the overall score for each subject: ； S in various disciplines 学科 For comparison, select S 学科 The lowest area as the learning difficulty, choose S 学科 The highest areas serve as points of interest.

3. According to the smart vocational education method and control system described in claim 2, it is characterized by: The personalized recommendation module includes a preliminary adjustment submodule based on instant feedback and an optimal learning path search submodule based on the leapfrog algorithm.

4. According to the smart vocational education method and control system described in claim 3, it is characterized by: The initial adjustment module based on immediate feedback in the personalized recommendation module includes: Initialize the student's learning path P0 based on the results of the data processing and analysis module; According to the current learning path P t Recommended Learning ResourcesA t , initially P t The actual value is P0; Get students' recommended learning resources A t Feedback F t ; F t :Students recommend resources A t feedback from the project manager, including satisfaction ratings and completion status; F t When it is positive feedback, the adjusted learning path P t+1 for: ； If F t The learning path adjusted for negative feedback is: ； α: adjustment parameter used to control the speed of learning path adjustment, which is set based on previous educational experience; According to the adjusted learning path P t+1 Replace the original learning path P t , repeat the above process until the preset learning target threshold G or end value T is reached, and the current optimal learning path is found.

5. According to the smart vocational education method and control system described in claim 4, it is characterized by: The optimal learning path search submodule of the personalized recommendation module based on the frog leaping algorithm includes: Define variables: Frog population L = {l1, l2, ..., lN}: represents all frogs, where N is the number of frogs; Subgroup B = {b1, b2, ..., bM}: represents all subgroups, where M is the number of subgroups; Local search step size δ and η: used to control the step size of local search; Global search step size γ: used to control the step size of global search; Iteration number U: the number of iterations of the frog leaping algorithm; The learning path P after preliminary adjustment t As a frog among frogs, The frogs are divided into several subgroups. The frogs in each subgroup have similar learning paths. In each subgroup, the frogs perform local searches based on their own and their neighbors' experiences and update their own learning paths. ； P t : Current frog learning path; P best : The optimal learning path in the current subgroup; P neighbor : The neighbor path of the current frog; δ and η: random numbers used to control the local search step size; Current Frog Learning Path P t Will go to the optimal learning path P in the current subgroup best and neighbor path P neighbor Move in the direction of finding a better learning path; The best learning path P in each subgroup best Exchange information with the best paths of other subgroups, perform a global search among all subgroups, and find the optimal learning path in the entire frog group: ； P t : Current frog learning path; P global_best : The globally optimal learning path; γ: random number used to control the global search step size; When the predetermined number of iterations U is reached or the termination condition is met, the optimal learning path P is output optimal ; The optimal learning path P found by the frog leaping algorithm optimal Update the current learning path P t .

6. According to the smart vocational education method and control system described in claim 5, it is characterized by: The teaching evaluation module collects learning feedback from the personalized recommendation module and conducts a comprehensive evaluation of students' multiple indicators; Evaluation indicators: , including scores, rankings, learning progress, ability development and learning attitude; Weight vector: ,The weight of each evaluation indicator is assigned according to teachers and historical experience; Calculation of comprehensive score E: ； E: Comprehensive evaluation of students’ learning effect; Result analysis, analysis of comprehensive scores E, evaluation of students' learning outcomes, feedback of evaluation results to students and teachers, and adjustment of teaching strategies and learning paths based on evaluation results P t+1 .

7. According to the smart vocational education method and control system described in claim 2, it is characterized by: The data processing and analysis module also includes a data analysis submodule, which analyzes the student's variables through a machine learning algorithm and automatically updates the weight vector W.

8. According to the smart vocational education method and control system described in claim 4, it is characterized by: The personalized recommendation module includes a real-time monitoring submodule, which collects students' learning behavior data in real time and dynamically adjusts the learning path P through a reinforcement learning algorithm. t+1 .

9. According to the smart vocational education method and control system described in claim 1, it is characterized by: The smart vocational education control system includes a recommendation submodule, which recommends personalized learning resources and activities based on students' historical learning data and current learning status; The data collection module collects students’ learning data through educational terminals; Learning data includes classroom interaction records, homework completion, online test scores, learning time, learning preferences, professional skills and job capabilities.

10. According to the smart vocational education method and control system described in claim 1, it is characterized by: H1: Identification of student characteristics: h11: Collect various variables of students; h12: The teacher determines the weight of each variable and normalizes it to obtain the weight vector; h13: Calculate the comprehensive score of students; h14: Set a personalized initial learning path P0 based on the comprehensive score and students’ learning characteristics; H2: Adaptive learning path adjustment: h21: Recommend learning resources and activities based on the initial learning path P0; h22: Collect student feedback on recommended learning resources or activities; h23: Dynamically adjust learning paths based on student feedback; h24: Update the current learning path P t ; H3: Teaching effectiveness evaluation: h31: Multiple evaluation indicators assess students’ learning outcomes; h32: Determine the weight of each evaluation indicator based on teachers and historical experience; h33: Calculate the comprehensive score of the student based on the weight and the value of each evaluation indicator; h34: Analyze composite scores, compare scores at different points in time, identify students’ strengths and weaknesses, and provide feedback to students and teachers; h35: Adjust teaching strategies and learning paths based on assessment results.

Images