An interactive processing method and device based on course management

Abstract

This invention discloses an interactive processing method and apparatus based on course management, relating to the field of intelligent education technology. The method includes: collecting student information and preprocessing it; generating an initial learning path based on the preprocessed information; monitoring student learning behavior data during the learning process according to the initial learning path; generating a personalized learning path based on the student's learning behavior data; real-time monitoring of student's emotional changes, cognitive load, and attention level data within the personalized learning path; constructing a learning state assessment model to evaluate the student's current learning state; and dynamically generating an immersive learning environment based on the student's current learning state. This invention improves the relevance and efficiency of educational resources, enhancing learning motivation and effectiveness.

Description

Technical Field

[0001] This invention relates to the field of intelligent education technology, and in particular to an interactive processing method and apparatus based on course management. Background Technology

[0002] With the development of information technology and the popularization of the internet, the education sector has also ushered in unprecedented changes. Traditional course management models mainly rely on fixed textbooks and syllabi, making it difficult to meet the differentiated needs of individual learners. In recent years, online education platforms based on advanced technologies such as big data analytics and artificial intelligence have gradually emerged, providing a certain degree of personalized services based on students' basic information and historical learning data. For example, by analyzing students' learning progress and test scores to adjust course content, a certain degree of adaptive learning path planning has been achieved. However, these methods are mostly limited to the application of surface data, lacking the ability to accurately assess and dynamically adjust students' deeper learning states. This not only limits the effectiveness of personalized learning programs but also fails to fully stimulate students' learning interest and potential.

[0003] While current technologies have improved the efficiency of educational resource utilization and teaching effectiveness to some extent, they still have significant shortcomings. First, existing technologies often overlook factors such as learners' emotional changes, cognitive load, and attention levels during the learning process, resulting in learning paths that do not fully match learners' actual needs. Second, existing adaptive learning path adjustment mechanisms are relatively simple, typically adjusting only based on academic performance or simple learning behaviors (such as answer accuracy or study time), without considering learners' emotional fluctuations and changes in psychological state. These problems make the optimization of learning paths less precise, thus affecting the improvement of learning outcomes. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an interactive processing method based on course management to solve the problem that the optimization of learning paths is not accurate enough in the prior art because it ignores factors such as students' emotional changes, cognitive load and attention level.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides an interactive processing method based on course management, comprising: collecting student information and preprocessing the student information; generating an initial learning path based on the preprocessed student information; monitoring student learning behavior data during the learning process according to the initial learning path; generating a personalized learning path based on the student learning behavior data; monitoring student emotional changes, cognitive load data, and attention level data in the personalized learning path in real time; constructing a learning status assessment model to assess the student's current learning status; and dynamically generating an immersive learning context based on the student's current learning status.

[0008] As a preferred embodiment of the interactive processing method based on course management described in this invention, the student information includes name, age, education level, available learning time, learning objectives, and existing knowledge reserves.

[0009] The preprocessing includes data cleaning, data standardization, and outlier detection.

[0010] As a preferred embodiment of the interactive processing method based on course management described in this invention, the specific steps for generating an initial learning path based on preprocessed student information are as follows:

[0011] Based on the preprocessed student information, the test scope is defined by the knowledge point screening method, and the probability of students mastering the knowledge points P(k) is initially predicted by combining the Bayesian knowledge tracing method.

[0012] Based on the probability of mastering knowledge points, the difficulty of questions is dynamically adjusted using Adaptive Testing to generate a preliminary intelligent test.

[0013] During the test, the time and accuracy of each student's answers to each knowledge point are recorded in real time.

[0014] Based on the students' answering time and accuracy for each knowledge point, courses are matched to students in the course database using a collaborative filtering algorithm.

[0015] Based on the courses matched to students and their available learning time, the courses are divided into multiple learning units using hierarchical clustering, and an initial learning path is generated using MDP.

[0016] As a preferred embodiment of the interactive processing method based on course management described in this invention, the step of monitoring students' learning behavior data during the learning process according to the initial learning path, and generating a personalized learning path based on the students' learning behavior data, includes the following specific steps.

[0017] Students begin their course learning based on an initial learning path, and their learning behavior data is monitored throughout the learning process;

[0018] The learning behavior data includes learning progress, test answer accuracy, answer time, learning frequency, and eye focus.

[0019] Input the students' emotional changes, cognitive load, and attention levels into the learning status assessment model to predict the students' current learning status index P'(K);

[0020] Based on historical learning data and mastery distribution, a low mastery threshold P1 and a high mastery threshold P2 are defined.

[0021] When P'(K)≤P1, the students' mastery is poor, so the basic courses in the initial learning path are increased and targeted practice is recommended;

[0022] When P1 < P'(K) ≤ P2, the students' mastery is average. Based on the initial learning path, we recommend consolidation exercises that combine new knowledge points.

[0023] When P'(K) > P2, the student has a good grasp of the material, and advanced courses are recommended based on the initial learning path.

[0024] Based on the comparison results of P'(K) with P1 and P2, a personalized learning path is generated.

[0025] As a preferred embodiment of the interactive processing method based on course management described in this invention, the emotion change data includes frowning frequency, tone fluctuation, speech rate change, and heart rate fluctuation.

[0026] The cognitive load data includes eye movement trajectory, number of repeated learning sessions, information dwell time, and distribution of eye movement dwell points;

[0027] The attention level data includes eye focus duration, blink frequency, head posture changes, and body forward tilt angle.

[0028] As a preferred embodiment of the interactive processing method based on course management described in this invention, the step of constructing a learning status assessment model involves inputting students' emotional change data, cognitive load data, and attention level data into the learning status assessment model to assess the students' current learning status. The specific steps are as follows:

[0029] A data input layer is constructed through a time synchronization mechanism to integrate trainees' emotional changes, cognitive load, and attention levels.

[0030] A feature extraction layer is constructed based on CNN and statistical analysis to obtain multi-level learning state feature vectors;

[0031] By employing time series alignment, feature normalization, and multimodal attention mechanisms, a data fusion layer is constructed to learn state features.

[0032] A learning state recognition layer is constructed based on Bayesian knowledge tracing and Hidden Markov Model (HMM) to identify the current learning state index of learners.

[0033] A learning state evaluation model is constructed based on a data input layer, a feature extraction layer, a data fusion layer, and a learning state recognition layer.

[0034] Input the students' emotional changes, cognitive load, and attention levels into the learning status assessment model to predict the students' current learning status index.

[0035] According to the learner's learning status index S t The learning status of trainees is categorized as focused, confused, and fatigued.

[0036] As a preferred embodiment of the interactive processing method based on course management described in this invention, the specific steps for dynamically generating an immersive learning scenario according to the student's current learning status are as follows:

[0037] An exploratory learning environment is generated when learners are focused.

[0038] When learners are confused, a guided learning scenario is generated.

[0039] A relaxing learning environment is created when learners are fatigued.

[0040] Secondly, the present invention provides an interactive processing device based on course management, comprising an initial path generation module, a personalized path generation module, a learning status assessment module, and a context generation module; the initial path generation module is used to collect student information, preprocess the student information, and generate an initial learning path based on the preprocessed student information; the personalized path generation module is used to monitor the student's learning behavior data during the learning process according to the initial learning path, and generate a personalized learning path based on the student's learning behavior data; the learning status assessment module is used to monitor the student's emotional change data, cognitive load data, and attention level data in the personalized learning path in real time, construct a learning status assessment model, and assess the student's current learning status; the context generation module is used to dynamically generate an immersive learning context based on the student's current learning status.

[0041] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the interactive processing method based on course management as described in the first aspect of the present invention.

[0042] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the interactive processing method based on course management as described in the first aspect of the present invention.

[0043] The beneficial effects of this invention are as follows: By collecting and preprocessing student information to generate personalized initial learning paths, and combining factors such as students' learning goals, existing knowledge, and available time, learning resources are accurately matched, effectively improving the relevance and efficiency of educational resources, and enhancing learning motivation and effectiveness. Simultaneously, based on real-time monitoring of students' emotional changes, cognitive load, and attention levels, teaching strategies are dynamically adjusted and immersive learning environments are generated. This not only focuses on improving academic performance but also pays special attention to students' emotional experience and psychological state, thereby effectively enhancing students' learning motivation and interest, reducing learning fatigue, and achieving the goal of comprehensively improving the learning experience and effectiveness. Attached Figure Description

[0044] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is a flowchart of the interactive processing method based on course management in Example 1.

[0046] Figure 2 This is a schematic diagram of the interactive processing device based on course management in Example 1. Detailed Implementation

[0047] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0048] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0049] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0050] Example 1, referring to Figure 1 and Figure 2 This is the first embodiment of the present invention, which provides an interactive processing method based on course management, including the following steps:

[0051] S1. Collect student information and preprocess it to generate an initial learning path based on the preprocessed student information.

[0052] Student information includes name, age, education level, available study time, learning goals, and existing knowledge.

[0053] It should be noted that the above content has been agreed to by the user and is used for legitimate purposes.

[0054] Preprocessing includes data cleaning, data standardization, and outlier detection.

[0055] It should be noted that removing missing values ​​from student information through data cleaning involves identifying whether there are blank or unfilled fields in the student information. For missing values ​​of key attributes such as name and age, appropriate filling strategies are adopted, such as using default values ​​or estimating based on other relevant information; while for non-key attributes, records containing missing values ​​may be directly deleted.

[0056] Data standardization normalizes the numerical data in student information. First, it determines the range of numerical data to be processed, such as available learning time and existing knowledge reserves. Then, it uses specific mathematical formulas to map these values ​​to a common interval, usually between [0,1], to eliminate differences between different orders of magnitude and ensure that all features have the same scale.

[0057] Outlier detection is used to identify and correct outlier data in student information. First, outliers are defined, typically based on statistical principles; for example, data points deviating from the mean by more than three standard deviations are considered outliers. After identifying these outliers, different corrective measures are taken depending on the specific circumstances. For instance, if the outlier is due to a data entry error, it can be adjusted within a known reasonable range.

[0058] Based on the preprocessed student information, the test scope is defined using a knowledge point screening method, and the probability of students mastering the knowledge points, P(k), is initially predicted using the Bayesian knowledge tracing method. The expression is as follows:

[0059]

[0060] Wherein, P(k) represents the student's initial grasp of knowledge point k, P0(k) represents the estimated value of the student's grasp of knowledge point k, λ1 is the weight coefficient of the student's existing knowledge reserves, λ2 is the weight coefficient of the student's available learning time, B represents the student's existing knowledge reserves, T represents the student's available learning time, β represents the test difficulty decay coefficient, and G is the test difficulty.

[0061] It should be noted that, firstly, the scope of the assessment is determined based on the student's existing knowledge base and available learning time, ensuring that the test content is both relevant to the student's current knowledge level and effectively promotes their further learning and development. Next, based on the student's historical learning data and a specific algorithm, the student's initial mastery of each knowledge point is calculated. This process fully considers the student's background information and learning potential, thereby tailoring an initial learning path most suitable for each student's individual circumstances.

[0062] It should also be noted that P0(k) is determined based on the learner's historical learning data. Specifically, this estimate is made by analyzing the learner's past performance on relevant knowledge points, including test scores, homework completion, and other assessment indicators, to judge their initial mastery of a particular knowledge point.

[0063] The student's existing knowledge reserve B is obtained through a comprehensive analysis of the student's past learning experience, completed courses, grades, and performance in relevant tests. The quantitative assessment is based on the proportion of the analysis results in the current course and the courses not yet studied.

[0064] The difficulty level G is determined by analyzing the average accuracy of knowledge points in previous tests, the average time required to complete related questions, and the level of concepts involved in the questions.

[0065] Based on the probability of mastering knowledge points, the difficulty of questions is dynamically adjusted using Adaptive Testing to generate a preliminary intelligent test.

[0066] It should be noted that the initial difficulty of the test questions is determined based on the student's preliminary grasp of each knowledge point. During the test, the student's actual mastery of the current knowledge point is assessed in real time based on the accuracy and speed of their answers. If the student can answer questions quickly and accurately, the difficulty of subsequent questions is automatically increased; conversely, the difficulty is reduced to ensure the effectiveness and adaptability of the test. In this way, the understanding of the student's knowledge level is gradually refined, ultimately generating an initial intelligent test that accurately reflects the student's true ability and can specifically address their weaknesses. For example, suppose a student in a math course, based on an assessment of their probability of mastering basic algebra knowledge points, is initially given a medium-difficulty algebra problem. The student solves this problem quickly and accurately, indicating a good grasp of the knowledge point. Next, based on the student's answers, the Adaptive Testing mechanism automatically provides higher-difficulty questions, such as application problems involving complex systems of equations or inequalities. However, if the student makes mistakes on these high-difficulty questions and spends a long time on them, the difficulty of subsequent questions will be reduced accordingly, providing questions between basic and medium difficulty to help consolidate related concepts. By dynamically adjusting the difficulty of questions based on the student's performance, a preliminary intelligent test is generated for the student that reflects their true ability level and can specifically improve their weak areas.

[0067] During the test, the time and accuracy of each student's answers to each knowledge point are recorded in real time.

[0068] Based on the students' answering time and accuracy for each knowledge point, courses are matched to students in the course database using a collaborative filtering algorithm.

[0069] It should be noted that the course database originates from long-term accumulated teaching resources and student learning data, containing a wealth of detailed records regarding course content, knowledge point distribution, and student learning behavior. First, student performance in tests is analyzed, including metrics such as answering time and accuracy, to assess their mastery of different knowledge points. Next, a collaborative filtering algorithm is used to search the course database for historical student data with similar learning patterns and needs, and based on this data, the most suitable course content for the current student is recommended.

[0070] Based on the courses matched to students and their available learning time, the courses are divided into multiple learning units using hierarchical clustering, and an initial learning path is generated using MDP.

[0071] It should be noted that, firstly, the matched course content is rationally planned based on the learner's available learning time. Using hierarchical clustering, the course is divided into several learning units suitable for gradual learning based on the relevance and difficulty level of the knowledge points, ensuring that each unit is neither too complex nor too simple, and matches the learner's learning pace. Subsequently, Markov Decision Process (MDP) is applied to optimize the order and combination of these learning units, generating a personalized initial learning path with the goal of maximizing learning efficiency and effectiveness.

[0072] S2. During the learning process of students according to the initial learning path, monitor students' learning behavior data, and generate personalized learning paths based on students' learning behavior data.

[0073] Students begin their course learning based on an initial learning path, and their learning behavior data is monitored throughout the learning process;

[0074] Learning behavior data includes learning progress, test answer accuracy, answer time, learning frequency, and eye focus;

[0075] Based on students' learning behavior data, the probability of students mastering knowledge points in the initial learning path is updated using Bayesian knowledge tracing combined with exponential decay and nonlinear activation functions. The expression is as follows:

[0076]

[0077] Where P'(K) represents the updated student's mastery of knowledge point k, γ1 represents the weighting coefficient of the correct answer rate, δ represents the time deviation decay coefficient, η represents the learning frequency decay coefficient, Q represents the student's correct answer rate, T2 represents the standard answer time, F represents the student's learning frequency, E represents the student's eye focus during the answering process, γ2 represents the weighting coefficient of the answering time, and T1 represents the student's actual answering time.

[0078] It should be noted that the standard answering time refers to the average time required to complete a specific question, which is pre-set based on the course content and the difficulty of the knowledge points. It is calculated by analyzing the answering time data of a large number of students on the same question, and taking into account factors such as the difficulty of the question and the complexity of the knowledge points.

[0079] It should also be noted that, firstly, based on the deviations between the student's correct answer Q, actual answer time T1, and standard answer time T2 during the learning process, the influence of the time deviation attenuation coefficient δ is calculated. Next, using the student's learning frequency F and eye focus E, adjustments are made using the learning frequency attenuation coefficient η and the corresponding nonlinear activation function, respectively, to ensure that the contribution of these factors to the degree of knowledge mastery is accurately reflected. Here, P(k) is the student's preliminary mastery estimate of knowledge point k obtained based on the Bayesian knowledge tracing method. By combining the weighting coefficients γ1 (representing the correct answer rate) and γ2 (representing the answer time), the student's knowledge point mastery probability P'(K) is precisely adjusted.

[0080] Based on historical learning data and mastery distribution, a low mastery threshold P1 and a high mastery threshold P2 are defined.

[0081] When P'(K)≤P1, the students' mastery is poor, so the basic courses in the initial learning path are increased and targeted practice is recommended;

[0082] When P1 < P'(K) ≤ P2, the students' mastery is average. Based on the initial learning path, we recommend consolidation exercises that combine new knowledge points.

[0083] When P'(K) > P2, the student has a good grasp of the material, and advanced courses are recommended based on the initial learning path.

[0084] It should be noted that when P'(K)≤P1, for example, a student's mastery probability of knowledge points in the elementary mathematics course is only 0.3, which is lower than the low mastery threshold P1, indicating that their mastery of the current knowledge points is poor. Therefore, basic courses in the initial learning path are added, and exercises specifically targeting these weak knowledge points are recommended to strengthen understanding.

[0085] When P1 < P'(K) ≤ P2, assuming a student has a knowledge mastery probability of 0.6 in the intermediate programming course, which is between P1 and P2, it indicates that their mastery is average. In this case, reinforcement exercises are added to the original learning path, and an appropriate amount of new knowledge points are introduced to help the student smoothly transition to more complex topics.

[0086] When P'(K) > P2, if a student's mastery probability of knowledge points in the advanced physics course reaches 0.9, exceeding the high mastery threshold P2, it indicates that the student has a good grasp of the relevant knowledge points. Therefore, based on the initial learning path, a more challenging advanced course is recommended to further deepen and expand the student's knowledge system.

[0087] Based on the comparison results of P'(K) with P1 and P2, a personalized learning path is generated.

[0088] S3. Real-time monitoring of learners' emotional changes, cognitive load, and attention levels in their personalized learning paths; construction of a learning status assessment model to evaluate learners' current learning status.

[0089] The data on mood changes included the frequency of frowning, fluctuations in tone of voice, changes in speech rate, and fluctuations in heart rate.

[0090] Cognitive load data includes eye movement trajectory, number of repetitions, information dwell time, and distribution of eye movement dwell points.

[0091] Attention level data include eye focus duration, blink frequency, head posture changes, and body tilt angle.

[0092] A data input layer is constructed through a time synchronization mechanism to integrate trainees' emotional changes, cognitive load, and attention levels.

[0093] It should be noted that, in order to build a data input layer through a time synchronization mechanism and integrate trainees’ emotional change data, cognitive load data and attention level data, it is first necessary to ensure that all collected data (such as frowning frequency, tone fluctuations, eye movement trajectories, etc.) can be accurately aligned to the same point in time.

[0094] A feature extraction layer is constructed based on CNN and statistical analysis to obtain multi-level learning state feature vectors;

[0095] It should be noted that convolutional neural networks (CNNs) are used to automatically identify complex patterns, such as extracting specific patterns representing high cognitive load from eye-tracking data. CNNs process the input data through multiple convolutional and pooling layers, capturing local dependencies and high-level abstract features. Simultaneously, statistical analysis methods are employed to further process the data, such as calculating the mean and variance of eye-tracking focus duration or analyzing the changing trends of blink frequency. These statistical indicators provide important information about the learner's attention level. Next, the high-level features extracted by the deep learning model are combined with the features obtained from statistical analysis to form a comprehensive feature vector set. This set contains multi-dimensional information reflecting the learner's current learning state. For example, when assessing the learner's cognitive load, not only are the eye-tracking features identified by the CNN considered, but also the statistical analysis results of blink frequency are combined to ensure the comprehensiveness and accuracy of the features. Through this comprehensive approach, a deep-level mining and precise expression of the learner's learning state features are achieved, constructing a feature extraction layer.

[0096] By employing time series alignment, feature normalization, and multimodal attention mechanisms, a data fusion layer is constructed to learn state features.

[0097] It should be noted that, firstly, for data from different sources (such as emotion change data, cognitive load data, and attention level data), precise time series alignment is performed based on their respective timestamps to ensure consistency of all information across the time dimension. For example, eye movement trajectory data is synchronized with corresponding moment-to-moment emotion change data (such as frowning frequency) to analyze the correlation of these data at the same point in time. Next, feature normalization techniques are used to adjust the scale of each feature, eliminating biases caused by differences in magnitude between different data types; for example, the movement speed of eye movements and blink frequency are standardized to the same numerical range. Finally, a multimodal attention mechanism is applied, dynamically adjusting weights according to the importance of each modality of data, allowing for greater focus on certain types of data in specific situations, such as considering eye focus rather than body lean angle when analyzing learners' concentration. Through this comprehensive processing method, efficient and accurate data fusion is achieved, constructing a data fusion layer.

[0098] A learning state recognition layer is constructed based on Bayesian knowledge tracing and Hidden Markov Model (HMM) to identify the current learning state index of learners.

[0099] It should be noted that Bayesian knowledge tracing is used to continuously track and update students' knowledge mastery. The mastery probability is adjusted based on students' performance on different knowledge points; for example, by analyzing students' accuracy and reaction time when answering specific questions, their mastery of a particular knowledge point is dynamically updated. Next, a Hidden Markov Model (HMM) is used, employing historical learning behavior data to predict students' learning state transition sequences, such as the probability of transitioning from a focused state to a confused state. The HMM establishes a series of hidden states (such as focused, confused, and fatigued) and their transition probabilities, and combines this with observed learning behavior characteristics (such as answering speed and eye movement patterns) to infer the most likely learning state path, thus constructing a learning state recognition layer.

[0100] A learning state evaluation model is constructed based on a data input layer, a feature extraction layer, a data fusion layer, and a learning state recognition layer.

[0101] It should be noted that by integrating the above layers, a complete chain from raw data collection to final learning status assessment is achieved. Each step focuses on solving specific problems, such as data synchronization, feature extraction, data fusion, and status recognition, working together to achieve a comprehensive understanding and accurate quantification of the learner's learning status.

[0102] Input the learners' emotional changes, cognitive load, and attention levels into the learning status assessment model to predict the learners' current learning status index, expressed as:

[0103]

[0104] Among them, St X is the learning status index of the student at time t. t The input data at time t is the input layer data, M is the total number of input data, and f is the input data. i (X t ) is the i-th learning state feature extracted by the feature extraction layer, α i Here, v1 is the attention weight for the i-th learning state feature, v1 is the weight coefficient for the updated probability of the learner mastering knowledge point k, and H is the attention weight for the ith learning state feature. t Here, N is the hidden variable of the learning state obtained from the HMM, N is the window length of the past time step, and v2 is the hidden variable of the learning state H. t The weighting coefficient, g j (X j ) represents the learning state features extracted by the feature extraction layer at past time j, and β j The learning state feature g of past time j j (X j The weighting coefficients of ψ(H) t H is the latent variable H of the learner's learning state. t Sensitivity coefficient to past learning experiences, o i μ represents the standard deviation of the feature of the i-th learning state. i Let P' represent the mean of the features of the i-th learning state, σ represent the activation function, and P' represent the mean of the features of the i-th learning state. min The minimum probability that a student has mastered all the knowledge points, P' max This represents the maximum probability that all students have mastered the knowledge point;

[0105] It should be noted that, firstly, the data X at time t is obtained from the input layer. t M learning state features f are calculated in the feature extraction layer. i (X t Each feature is assigned a different attention weight α based on its importance. i This forms a weighted product. Next, combining the updated student's mastery level P'(K) of knowledge point K with the corresponding weight coefficient v1, this part reflects the impact of the student's knowledge mastery on their current learning state. Simultaneously, the latent variable H of the learning state, obtained using a Hidden Markov Model (HMM), is used... t And its weight coefficient v2, taking into account the influence of past learning experience on the current state, through the features g of N time points in historical data. j X j and its weighting coefficient β j These features are also quantified based on the latent variables H of the learning state. t The sensitivity coefficient ψ(H) tThe weights are adjusted to reflect the impact of long-term learning trends on the current state. Finally, all these factors are integrated through a nonlinear transformation σ to obtain the learner's learning state index S at time t. t .

[0106] According to the learner's learning status index S t The learning status of trainees is categorized as focused, confused, and fatigued.

[0107] It should be noted that, firstly, based on the calculated S t The value is used to judge the student's current psychological and cognitive state. For example, if a student's learning state index S at a certain moment... t If the score falls within a preset attention range, such as between 0.8 and 1.0, the learner is marked as being in a focused state, indicating a high level of attention and low cognitive load, suitable for in-depth learning or exploring new knowledge. If S... t A score within the confused range, such as between 0.4 and 0.7, indicates that the learner may be experiencing some degree of comprehension difficulty or cognitive overload, requiring additional guidance and support to help them overcome the obstacles. Finally, when S... t Below a certain threshold, say between 0 and 0.3, learners are labeled as fatigued, indicating a significant decrease in attention and an increase in cognitive load. In this case, rest or light activities are recommended to restore energy. This is achieved based on the learning state index S. t The specific numerical range can accurately identify and mark the student's learning status, thereby enabling appropriate educational interventions to optimize learning outcomes.

[0108] S4. Dynamically generate immersive learning scenarios based on the learner's current learning status.

[0109] An exploratory learning environment is generated when learners are focused.

[0110] It should be noted that the initial approach relies on a precise assessment of the learner's current learning status index, based on previously collected and analyzed data on emotional changes, cognitive load, and attention levels. After confirming the learner's focused state using this data, the next step is to deeply analyze their learning path and mastery of knowledge points. By comparing the knowledge points already mastered by the learner with other untouched knowledge points in the course syllabus, content that is both challenging and aligned with the learner's interests and developmental direction is identified. For example, in a history course, if a learner is particularly interested in ancient civilizations and has already grasped basic background knowledge, articles and videos on the latest research findings or archaeological discoveries related to that specific civilization are recommended. Furthermore, algorithms are used to filter suitable resources from the course database, and interactive elements such as virtual reality experiences or online discussion groups are combined to further enhance the learning experience.

[0111] When learners are confused, a guided learning scenario is generated.

[0112] It should be noted that a detailed analysis of multi-dimensional data, including students' incorrect answers, answer times, and eye-tracking patterns, is conducted during the learning process to determine the specific knowledge points where students encounter difficulties. For example, in a physics course, if a student shows confusion in understanding the application of Newton's Third Law, relevant teaching materials will be selected based on this information. These include detailed explanatory videos, workshops demonstrating step-by-step how to apply the law to solve practical problems, and a set of quizzes containing similar questions with gradually increasing difficulty. All these resources are carefully selected and ordered to help students gradually build an understanding of the knowledge point. A feedback mechanism is also provided, allowing students to submit questions and receive targeted answers. This not only directly addresses students' confusion but also enhances their confidence.

[0113] A relaxing learning environment is created when learners are fatigued.

[0114] It should be noted that assessing the physical and mental fatigue levels of learners relies on analyzing data such as eye focus, blink frequency, and head posture changes. Once fatigue is confirmed, the next step is to design a series of activities to help restore energy. For example, after long periods of reading and technical learning, recommending a 5-minute meditation video of a natural landscape or playing some soft classical music as a short break. Additionally, providing light and engaging learning content, such as stories or anecdotes related to the learner's field but not overly stressful, like introducing interesting historical snippets of early computer development in a programming course. This content is designed to reduce learner stress, allowing for effective relaxation and recovery in a short time. In this way, learners can maintain learning continuity while avoiding excessive fatigue.

[0115] This embodiment also provides an interactive processing device based on course management, including: an initial path generation module, a personalized path generation module, a learning status assessment module, and a context generation module; the initial path generation module is used to collect student information, preprocess the student information, and generate an initial learning path based on the preprocessed student information; the personalized path generation module is used to monitor the student's learning behavior data during the learning process according to the initial learning path, and generate a personalized learning path based on the student's learning behavior data; the learning status assessment module is used to monitor the student's emotional change data, cognitive load data, and attention level data in the personalized learning path in real time, construct a learning status assessment model, and assess the student's current learning status; the context generation module is used to dynamically generate an immersive learning context based on the student's current learning status.

[0116] This embodiment also provides a computer device suitable for interactive processing methods based on course management, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the interactive processing method based on course management as proposed in the above embodiment.

[0117] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0118] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the interactive processing method for course management as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0119] In summary, this invention, by collecting and preprocessing student information to generate personalized initial learning paths, and combining factors such as students' learning goals, existing knowledge, and available time, accurately matches learning resources, effectively improving the relevance and efficiency of educational resources, and enhancing learning motivation and effectiveness. Simultaneously, based on real-time monitoring of students' emotional changes, cognitive load, and attention levels, it dynamically adjusts teaching strategies and generates immersive learning environments, focusing not only on improving academic performance but also on students' emotional experience and psychological state, thereby effectively enhancing students' learning motivation and interest, reducing learning fatigue, and achieving the goal of comprehensively improving the learning experience and effectiveness.

[0120] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An interactive process method based on course management, characterized by: Comprising, Collecting student information and pre-processing the student information, generating an initial learning path according to the pre-processed student information, the steps are as follows, According to the preprocessed student information, the test range is defined through a knowledge point screening method, and the knowledge point mastering probability of the student is preliminarily predicted in combination with a Bayesian knowledge tracking method ; According to the knowledge point mastery probability, the Adaptive Testing is used to dynamically adjust the difficulty of the questions, and finally a preliminary intelligent test is generated; Record the answering time and accuracy of each knowledge point of the student in the test process in real time; According to the answering time and accuracy of each knowledge point of the student, the course database is matched with the student through the collaborative filtering algorithm; Based on the courses matched for the student, combined with the available learning time of the student, the courses are split into multiple learning units through hierarchical clustering method, and the initial learning path is generated combined with MDP; In the process of learning according to the initial learning path of the student, the learning behavior data of the student is monitored, and the personalized learning path is generated according to the learning behavior data of the student; The process of generating a personalized learning path includes updating the knowledge point mastery probability of the student in the initial learning path based on the learning behavior data of the student through the Bayesian knowledge tracking method combined with exponential decay and nonlinear activation function, and the expression is: ； wherein, represents the updated mastery level of the student on the knowledge point, represents the weight coefficient of the correct answer rate, represents the time deviation decay coefficient, represents the learning frequency decay coefficient, represents the correct answer rate of the student, represents the standard answer time, represents the learning frequency of the student, represents the eye concentration of the student during the answering process, represents the weight coefficient of the answering time, actual answering time of the student;​ Real-time monitoring of the student's emotional change data, cognitive load data and attention level data in the personalized learning path, constructing a learning state evaluation model to evaluate the current learning state of the student, the specific steps are as follows, Through the time synchronization mechanism, the data input layer is constructed, and the emotional change data, cognitive load data and attention level data of the student are integrated; Based on CNN and statistical analysis method, a feature extraction layer is constructed to obtain multi-level learning state feature vectors; Through time series alignment, feature normalization and multi-modal attention mechanism, a data fusion layer is constructed to fuse learning state features; Based on Bayesian knowledge tracking and HMM, a learning state recognition layer is constructed to identify the current learning state index of the student; Based on the data input layer, the feature extraction layer, the data fusion layer and the learning state recognition layer, a learning state evaluation model is constructed; The emotional change data, cognitive load data and attention level data of the student are input into the learning state evaluation model to predict the current learning state index of the student, and the expression is: ； wherein, is the learning state index of the student at time is the time input data of the input layer, is the total number of input data, is the i-th learning state feature extracted by the feature extraction layer, is the attention weight of the i-th learning state feature, is the updated weight coefficient of the student's mastery probability of the knowledge point is the learning state hidden variable obtained by the HMM, is the window length of the past time, is the weight coefficient of the learning state hidden variable is the learning state feature extracted by the feature extraction layer at the past time is the weight coefficient of the learning state feature at the past time is the sensitivity coefficient of the student to the past learning experience in the learning state hidden variable represents the standard deviation of the i-th learning state feature, represents the mean of the i-th learning state feature, represents the activation function, the minimum value of the student's mastery probability of all knowledge points, represents the maximum value of the mastery probability of the knowledge points of all students;​​​​​​​​​​ According to the learning state index of the student The learning state of the student is marked as a focused state, a confused state, and a tired state. According to the current learning state of the student, an immersive learning situation is dynamically generated.

2. The interactive process method based on lesson management as claimed in claim 1, wherein: The student information includes name, age, education, available learning time, learning goal and existing knowledge reserve; The pre-processing includes data cleaning, data standardization and outlier detection.

3. The interactive process method based on lesson management as claimed in claim 2, wherein: In the process of learning according to the initial learning path of the student, the learning behavior data of the student is monitored, and the personalized learning path is generated according to the learning behavior data of the student, the specific steps are as follows, The student starts course learning based on the initial learning path, and monitors the learning behavior data of the student in the learning process; The learning behavior data includes learning progress, test answering accuracy, answering time, learning frequency and eye concentration; Based on the learning behavior data of the students, the knowledge point mastering probability of the students in the initial learning path is updated by the Bayesian knowledge tracing method combined with exponential decay and a nonlinear activation function ; Based on the historical learning data and mastery degree distribution, the low mastery degree threshold P1 and the high mastery degree threshold P2 are defined; When When P1, the student's mastery is poor, increase the basic course in the preliminary learning path, recommend targeted practice; When P1 < When the level is ≤P2, the students' grasp of the material is generally average. Based on the initial learning path, we recommend reinforcement exercises that incorporate new knowledge points. When When P2, the student has a good grasp of the situation, and on the basis of the initial learning path to recommend advanced courses; According to A personalized learning path is generated based on the comparison results of P1 and P2.

4. The interactive process method based on lesson management as claimed in claim 3, wherein: The emotional change data includes frown frequency, tone fluctuation, speech speed change and heart rate fluctuation; The cognitive load data includes eye movement trajectory, repeated learning times, information dwell time and eye movement dwell point distribution; The attention level data includes eye focus duration, blink frequency, head posture change and body inclination angle.

5. The interactive process method based on lesson management as claimed in claim 4, wherein: The immersive learning situation is dynamically generated according to the current learning state of the student, and the specific steps are as follows, When the student is in a state of concentration, an exploratory learning situation is generated; When the student is in a state of confusion, a guided learning situation is generated; When the student is in a state of fatigue, a relaxing learning situation is generated.

6. An interactive processing apparatus based on lesson management, based on the interactive processing method based on lesson management according to any one of claims 1 to 5, characterized by: It comprises an initial path generation module, a personalized path generation module, a learning state evaluation module and a situation generation module. The initial path generation module is used for collecting student information and pre-processing the student information, and generating an initial learning path according to the pre-processed student information. The personalized path generation module is used for monitoring the learning behavior data of the student in the process of learning according to the initial learning path, and generating a personalized learning path according to the learning behavior data of the student. The learning state evaluation module is used for real-time monitoring of the emotional change data, cognitive load data and attention level data of the student in the personalized learning path, constructing a learning state evaluation model, and evaluating the current learning state of the student. The situation generation module is used for dynamically generating an immersive learning situation according to the current learning state of the student. 7.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor executes the computer program to realize the steps of the interactive processing method based on course management according to any one of claims 1-5.

8. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to realize the steps of the interactive processing method based on course management according to any one of claims 1-5.

Images