Teaching style optimization method based on teaching quality feedback

Abstract

The invention discloses a teaching style optimization method based on teaching quality feedback. The method comprises the following steps: capturing classroom environment data in real time based on sensing equipment; performing time-frequency domain feature decoupling on the unstructured teaching behavior data based on an adaptive wavelet packet decomposition algorithm to obtain a multi-dimensional feature vector; constructing a teaching effect evaluation matrix, and obtaining a nonlinear mapping relationship between the teaching behavior characteristics and a preset teaching target; generating an adversarial network, obtaining a teaching behavior adjustment scheme, and verifying the effectiveness of the scheme based on a historical teaching case library; based on an adjustment scheme dynamic compensation mechanism, optimizing scheme parameters according to secondary feedback data after the scheme is executed; and constructing a teaching style evolution graph, and forming a traceable teaching style optimization track for the iterative path of the teaching scheme. The method has the advantages that teaching strategies can be accurately adjusted, classroom interaction and teaching effects are improved, and a dynamic optimization mechanism and a traceable evolutionary graph help teachers to realize continuous improvement of personalized teaching styles.

Description

Technical Field

[0001] The present invention relates to parameter optimization technology, and in particular to a teaching style optimization method based on teaching quality feedback. Background Art

[0002] With the diversification and personalization of educational needs, the traditional "one-size-fits-all" teaching model has been unable to meet the learning needs of different students. The core goal of teaching style optimization is to design teaching plans that meet their personalized needs based on students' cognitive characteristics, learning methods and interest preferences. In recent years, with the advancement of information technology, especially the application of artificial intelligence, big data and learning analysis technology, educators can more accurately understand students' learning behavior, performance and learning path, thus providing a scientific basis for the optimization of teaching style.

[0003] The current teaching style optimization methods on the market are mainly concentrated in the fields of interactive teaching and blended teaching. Interactive teaching focuses on the interaction between teachers and students, and between students. Through question guidance, discussion and group cooperation, it stimulates students' thinking and creativity, and enhances students' sense of participation and learning motivation. Blended teaching combines the advantages of online and offline learning, combining traditional teaching with modern technology. It not only ensures the interactivity of face-to-face teaching, but also utilizes the convenience and rich resources of the network platform to improve learning effects. However, the traditional teaching style optimization method lacks real-time feedback and flexible adjustment, and it is difficult to meet the personalized needs of students. In contrast, the optimization method based on teaching quality feedback can achieve precise adjustment through data-driven and improve teaching effects. Summary of the invention

[0004] In order to improve the existing teaching style optimization methods, a teaching style optimization method based on teaching quality feedback is provided. This method can accurately adjust teaching strategies and improve classroom interaction and teaching effectiveness by capturing multi-dimensional teaching behavior data in real time and analyzing it using advanced algorithms.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is: A teaching style optimization method based on teaching quality feedback, comprising: Based on sensor equipment, real-time capture of teaching behavior data in the classroom environment, including speech intonation spectrum characteristics, blackboard writing trajectory pressure value, and teacher-student interaction response delay; Based on the adaptive wavelet packet decomposition algorithm, the time-frequency domain features of the unstructured teaching behavior data are decoupled to obtain multi-dimensional feature vectors; Based on the obtained multi-dimensional feature vector, a teaching effect evaluation matrix is ​​constructed to obtain the nonlinear mapping relationship between teaching behavior characteristics and preset teaching objectives; Generate an adversarial network based on the nonlinear mapping relationship between teaching behavior characteristics and preset teaching goals, obtain teaching behavior adjustment plans, and verify the effectiveness of the plans based on the historical teaching case library; Based on the dynamic compensation mechanism of the adjustment scheme, the scheme parameters are optimized according to the secondary feedback data after the scheme is executed; Construct a teaching style evolution map to form a traceable teaching style optimization trajectory for the iterative path of the teaching plan.

[0006] Preferably, the step of performing time-frequency domain feature decoupling on the unstructured teaching behavior data based on the adaptive wavelet packet decomposition algorithm to obtain a multi-dimensional feature vector specifically includes: Based on the acquired teaching behavior data, a synchronized timing signal x(t) is acquired; Based on the signal characteristics, the wavelet basis is adaptively selected by minimizing the reconstruction error. The formula is: , in, is the candidate wavelet basis function, is the wavelet basis candidate set, , is the inner product of the signal and the wavelet basis, is the wavelet basis function with scale j and translation k, is the L2 norm; Based on the Shannon entropy criterion, the number of decomposition layers L is adaptively determined, and the sub-bands whose energy proportion exceeds the threshold are retained; The multidimensional feature vector specifically includes the emotion intensity index E, the knowledge transfer efficiency coefficient K, and the classroom attention concentration A; The energy and variance of the high-frequency sub-bands are used to obtain the intensity index E, and the formula is: , in, is the set of high frequency sub-bands, is the L2 norm square of the sub-band wavelet coefficients, is the variance, , is the weight coefficient; The knowledge transfer efficiency coefficient K is obtained based on the fundamental frequency stability and semantic information entropy of the teacher's voice. The formula is: , in, is the fundamental frequency of the teacher's voice, is the semantic sentiment score, is the total duration of the lesson, It is the inverse of semantic information entropy. The lower the entropy, the more concentrated the knowledge transfer. Based on the low-frequency energy proportion of the student's posture signal P(t) and the eye movement entropy, the classroom attention concentration A is obtained. The formula is: , in, Gesture signals for students The low-frequency wavelet packet coefficients of is the entropy of eye movement trajectory, is the total number of fixations, , A weight parameter for balancing low-frequency energy and eye movement entropy.

[0007] Preferably, constructing a teaching effect evaluation matrix based on the obtained multidimensional feature vector to obtain a nonlinear mapping relationship between teaching behavior characteristics and preset teaching objectives specifically includes: Based on the obtained multi-dimensional feature vector V=[E,K,A] and the preset teaching target vector T, the teaching effect evaluation matrix is ​​constructed: , in, is a Gaussian kernel function that measures the nonlinear similarity between features and targets. is the target weight vector, is the Gaussian kernel bandwidth, which controls the similarity decay speed; Based on the deep neural network architecture, obtain the mapping function from the learning feature vector V to the teaching target T .

[0008] Preferably, generating an adversarial network based on the nonlinear mapping relationship between teaching behavior characteristics and preset teaching objectives, obtaining a teaching behavior adjustment plan, and verifying the effectiveness of the plan based on a historical teaching case library specifically include: Based on the historical teaching quality data in the historical case library, it is input into the adversarial network for alternating training, with the generator minimizing the generator loss and the discriminator minimizing the discriminator loss; The generator part adopts a conditional variational autoencoder (CVAE), whose latent space embeds the syllabus constraints; The discriminator part includes a rule-based knowledge graph verification module for connecting the subject knowledge ontology library to verify the logical consistency of the teaching strategy, and a statistical feature comparison module for calculating the Wasserstein distance between the generated strategy and excellent teaching cases; Add a gradient penalty term to the discriminator loss to obtain stable training results; The multi-dimensional feature vector group under optimal adjustment is obtained through iterative optimization of the generator; Based on the generated adjustment plan, its similarity with the cases with higher feedback on history teaching quality is calculated to obtain the effectiveness of the generated adjustment plan.

[0009] Preferably, the generating adjustment plan, calculating the similarity between the generated adjustment plan and the case with high feedback on history teaching quality, and obtaining the effectiveness of the generated adjustment plan specifically includes: The similarity calculation formula is: , Among them, the threshold Indicates high similarity, the pass rate needs to be greater than 70%.

[0010] Preferably, the dynamic compensation mechanism based on the adjustment scheme and optimizing the scheme parameters according to the secondary feedback data after the scheme is executed specifically includes: Based on each generated adjustment plan, obtain real-time teaching quality feedback data and teaching behavior data, and calculate the error vector between it and the teaching target vector ; The compensation amount is calculated based on the PID control update formula, which is: , in, Control the size of the error, The impact of cumulative error, Control error change rate, is the integral term of the error, indicating the accumulation of past errors, is the rate of change of the error, indicating the rate of error; Based on the PID control update formula, the new parameter feature vector E, K, A is calculated and applied to the new teaching plan to obtain the secondary feedback data and calculate the new error vector between it and the teaching target vector. ; Based on the secondary feedback error vector Further adjust the teaching plan. If the error vector If it is small enough, stop adjusting; if it is still large, repeat the above process to further optimize the parameters.

[0011] Preferably, the construction of the teaching style evolution map to form a traceable teaching style optimization trajectory for the iterative path of the teaching plan specifically includes: Based on each iteration of the teaching plan, data is recorded for each dimension; Use graph database to store nodes and edges. Nodes include teaching behaviors and student feedback, and edges include optimization paths and effect associations. If the current solution improves students’ higher-order thinking, the weight of the edge will be increased; otherwise, the teaching plan will be refined and the solution parameters will be optimized; Based on the iteration records, the timeline shows the style evolution, the heat map shows the effectiveness of the strategy, and the trajectory map of each style optimization is drawn; The decision basis for a specific iteration is located through SPARQL queries, and an early warning is triggered when the solution deviates from the historical optimal range.

[0012] Compared with the prior art, the advantages of the present invention are: By comprehensively utilizing a variety of sensor devices to capture teaching behavior data in the classroom environment in real time, the teacher's teaching activities and classroom effects can be fully and accurately reflected. The time-frequency domain features of these unstructured data are decoupled by the adaptive wavelet packet decomposition algorithm, and multi-dimensional feature vectors can be extracted to construct a teaching effect evaluation matrix. This makes it possible to clarify the nonlinear mapping relationship between teaching behavior and preset teaching objectives, and then use the generative adversarial network to design a teaching behavior adjustment plan, and verify the actual effect of the plan in combination with the historical teaching case library. This secondary feedback data optimization based on the dynamic compensation mechanism enables the plan to be continuously iterated and adjusted, improving the accuracy and flexibility of the teaching plan. Finally, by constructing a teaching style evolution map and forming a traceable optimization path, the continuous evolution and personalized adjustment of the teaching style are achieved. This method can refine the teaching effect, optimize the teaching strategy, and provide teachers with a more scientific and data-driven teaching improvement plan, which has significant advantages, especially in improving the quality of classroom interaction and personalized teaching. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 A schematic diagram of the method proposed by the present invention; Figure 2 A schematic diagram of obtaining a multi-dimensional feature vector of the method proposed in the present invention; Figure 3 A schematic diagram of the nonlinear mapping relationship of the method proposed in the present invention; Figure 4 A schematic diagram of adjusting the scheme of the method proposed in the present invention; Figure 5 This is a schematic diagram of the scheme parameter optimization of the method proposed in the present invention; Figure 6 This is a schematic diagram of the solution management of the method proposed in the present invention. DETAILED DESCRIPTION

[0014] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0015] See also Figure 1 As shown, a teaching style optimization method based on teaching quality feedback includes: Step 1: Based on the sensor equipment, real-time capture of teaching behavior data in the classroom environment, including speech intonation spectrum characteristics, blackboard writing trajectory pressure value, and teacher-student interaction response delay; Step 2: Decouple the time-frequency domain features of the unstructured teaching behavior data based on the adaptive wavelet packet decomposition algorithm to obtain a multi-dimensional feature vector; Step 3: Based on the obtained multi-dimensional feature vector, a teaching effect evaluation matrix is ​​constructed to obtain the nonlinear mapping relationship between the teaching behavior characteristics and the preset teaching objectives; Step 4: Generate an adversarial network based on the nonlinear mapping relationship between teaching behavior characteristics and preset teaching objectives, obtain a teaching behavior adjustment plan, and verify the effectiveness of the plan based on a historical teaching case library; Step 5: Based on the dynamic compensation mechanism of the adjustment scheme, optimize the scheme parameters according to the secondary feedback data after the scheme is executed; Step 6: Construct a teaching style evolution map to form a traceable teaching style optimization trajectory for the iterative path of the teaching plan.

[0016] See also Figure 2 As shown, based on the adaptive wavelet packet decomposition algorithm, the time-frequency domain feature decoupling of the unstructured teaching behavior data is performed to obtain the multi-dimensional feature vector, which specifically includes: Based on the acquired teaching behavior data, a synchronized timing signal x(t) is acquired; Based on the signal characteristics, the wavelet basis is adaptively selected by minimizing the reconstruction error. The formula is: , in, is the candidate wavelet basis function, is the wavelet basis candidate set, , is the inner product of the signal and the wavelet basis, is the wavelet basis function with scale j and translation k, is the L2 norm; Based on the Shannon entropy criterion, the number of decomposition layers L is adaptively determined, and the sub-bands whose energy proportion exceeds the threshold are retained; The multidimensional feature vector specifically includes the emotion intensity index E, the knowledge transfer efficiency coefficient K, and the classroom attention concentration A; The energy and variance of the high-frequency sub-bands are used to obtain the intensity index E, and the formula is: , in, is the set of high frequency sub-bands, is the L2 norm square of the sub-band wavelet coefficients, is the variance, , is the weight coefficient; The knowledge transfer efficiency coefficient K is obtained based on the fundamental frequency stability and semantic information entropy of the teacher's voice. The formula is: , in, is the fundamental frequency of the teacher's voice, is the semantic sentiment score, is the total duration of the lesson, It is the inverse of semantic information entropy. The lower the entropy, the more concentrated the knowledge transfer. Based on the low-frequency energy proportion of the student's posture signal P(t) and the eye movement entropy, the classroom attention concentration A is obtained. The formula is: , in, Gesture signals for students The low-frequency wavelet packet coefficients of is the entropy of eye movement trajectory, is the total number of fixations, , A weight parameter for balancing low-frequency energy and eye movement entropy.

[0017] It is understandable that the adaptive determination of the decomposition level L depends on the Shannon entropy criterion, but different signal characteristics may cause large changes in the entropy value, which in turn affects the selection of the L value, resulting in insufficient or excessive feature decoupling. Multi-scale analysis can be introduced, combined with the energy distribution, variance and other characteristics of the signal, and appropriate constraints can be introduced to ensure that the selection of the L value is reasonable and avoid excessive or insufficient decomposition.

[0018] At the same time, the energy and variance of high-frequency sub-bands, the proportion of low-frequency energy and eye movement entropy may be easily affected by noise or environmental factors, resulting in unstable calculations. Especially in complex environments, multimodal data such as heart rate and expression recognition can be used to increase comprehensive monitoring of students' status and ensure more accurate assessment of attention concentration.

[0019] See also Figure 3 As shown, based on the obtained multi-dimensional feature vector, a teaching effect evaluation matrix is ​​constructed to obtain the nonlinear mapping relationship between the teaching behavior characteristics and the preset teaching objectives, which specifically includes: Based on the obtained multi-dimensional feature vector V=[E,K,A] and the preset teaching target vector T, the teaching effect evaluation matrix is ​​constructed: , in, is a Gaussian kernel function that measures the nonlinear similarity between features and targets. is the target weight vector, is the Gaussian kernel bandwidth, which controls the similarity decay speed; Based on the deep neural network architecture, obtain the mapping function from the learning feature vector V to the teaching target T .

[0020] Specifically, the deep neural network architecture includes an input layer, multiple hidden layers, and an output layer. The input layer receives the learning feature vector, and the output layer predicts the teaching target vector. Select a suitable loss function (such as mean square error or cross entropy) to measure the error between the predicted output and the actual teaching target. Select a suitable loss function (such as mean square error or cross entropy) to measure the error between the predicted output and the actual teaching target. Use the validation set to evaluate the trained neural network and check the generalization ability and prediction accuracy of the model. According to the training results, output the student's achievement in each teaching target dimension and generate the teaching effect evaluation results.

[0021] See also Figure 4 As shown in the figure, an adversarial network is generated based on the nonlinear mapping relationship between teaching behavior characteristics and preset teaching objectives to obtain a teaching behavior adjustment plan, and the effectiveness of the plan is verified based on a historical teaching case library. Specifically, it includes: Based on the historical teaching quality data in the historical case library, it is input into the adversarial network for alternating training, with the generator minimizing the generator loss and the discriminator minimizing the discriminator loss; The generator part adopts a conditional variational autoencoder (CVAE), whose latent space embeds the syllabus constraints; The discriminator part includes a rule-based knowledge graph verification module for connecting the subject knowledge ontology library to verify the logical consistency of the teaching strategy, and a statistical feature comparison module for calculating the Wasserstein distance between the generated strategy and excellent teaching cases; Add a gradient penalty term to the discriminator loss to obtain stable training results; The multi-dimensional feature vector group under optimal adjustment is obtained through iterative optimization of the generator; Based on the generated adjustment plan, its similarity with the cases with higher feedback on history teaching quality is calculated to obtain the effectiveness of the generated adjustment plan.

[0022] Based on the generated adjustment plan, the similarity between it and the case with high feedback on history teaching quality is calculated, and the effectiveness of the generated adjustment plan is obtained, which specifically includes: The similarity calculation formula is: , Among them, the threshold Indicates high similarity, the pass rate needs to be greater than 70%.

[0023] See also Figure 5 As shown, based on the dynamic compensation mechanism of the adjustment scheme, the scheme parameters are optimized according to the secondary feedback data after the scheme is executed, including: Based on each generated adjustment plan, obtain real-time teaching quality feedback data and teaching behavior data, and calculate the error vector between it and the teaching target vector ; The compensation amount is calculated based on the PID control update formula, which is: , in, Control the size of the error, The impact of cumulative error, Control error change rate, is the integral term of the error, indicating the accumulation of past errors, is the rate of change of the error, indicating the rate of error; Based on the PID control update formula, the new parameter feature vector E, K, A is calculated and applied to the new teaching plan to obtain the secondary feedback data and calculate the new error vector between it and the teaching target vector. ; Based on the secondary feedback error vector Further adjust the teaching plan. If the error vector If it is small enough, stop adjusting; if it is still large, repeat the above process to further optimize the parameters.

[0024] See also Figure 6 As shown in the figure, the teaching style evolution graph is constructed to form a traceable teaching style optimization trajectory for the iterative path of the teaching plan. Specifically, it includes: Based on each iteration of the teaching plan, data is recorded for each dimension; Use graph database to store nodes and edges. Nodes include teaching behaviors and student feedback, and edges include optimization paths and effect associations. If the current solution improves students’ higher-order thinking, the weight of the edge will be increased; otherwise, the teaching plan will be refined and the solution parameters will be optimized; Based on the iteration records, the timeline shows the style evolution, the heat map shows the effectiveness of the strategy, and the trajectory map of each style optimization is drawn; The decision basis for a specific iteration is located through SPARQL queries, and an early warning is triggered when the solution deviates from the historical optimal range.

[0025] Specifically, in each iteration of the teaching plan, data of various dimensions are first recorded, including students' classroom performance, participation, emotional response, and academic performance. All data is stored in the graph database in the form of nodes, and each node represents a teaching behavior or student feedback. The edges represent the optimization paths and effect associations between different teaching plans. Whenever a plan effectively improves students' high-order thinking ability, the weight of the relevant edges in the graph database will increase, indicating the effectiveness of the plan in this field. Each refinement records the specific adjustment path and improvement effect through the update of nodes and edges in the graph database, providing data support and decision-making basis for subsequent optimization of teaching plans.

[0026] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The above is a description of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0027] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0028] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A teaching style optimization method based on teaching quality feedback, characterized in that: include: Based on sensor equipment, real-time capture of teaching behavior data in the classroom environment, including speech intonation spectrum characteristics, blackboard writing trajectory pressure value, and teacher-student interaction response delay; Based on the adaptive wavelet packet decomposition algorithm, the time-frequency domain features of the unstructured teaching behavior data are decoupled to obtain multi-dimensional feature vectors; Based on the obtained multi-dimensional feature vector, a teaching effect evaluation matrix is ​​constructed to obtain the nonlinear mapping relationship between teaching behavior characteristics and preset teaching objectives; Generate an adversarial network based on the nonlinear mapping relationship between teaching behavior characteristics and preset teaching goals, obtain teaching behavior adjustment plans, and verify the effectiveness of the plans based on the historical teaching case library; Based on the dynamic compensation mechanism of the adjustment scheme, the scheme parameters are optimized according to the secondary feedback data after the scheme is executed; Construct a teaching style evolution map to form a traceable teaching style optimization trajectory for the iterative path of the teaching plan.

2. The teaching style optimization method based on teaching quality feedback according to claim 1, characterized in that: The method of performing time-frequency domain feature decoupling on the unstructured teaching behavior data based on the adaptive wavelet packet decomposition algorithm to obtain a multi-dimensional feature vector specifically includes: Based on the acquired teaching behavior data, a synchronized timing signal x(t) is acquired; Based on the signal characteristics, the wavelet basis is adaptively selected by minimizing the reconstruction error. The formula is: ， in, is the candidate wavelet basis function, is the wavelet basis candidate set, , is the inner product of the signal and the wavelet basis, is the wavelet basis function with scale j and translation k, is the L2 norm; Based on the Shannon entropy criterion, the number of decomposition layers L is adaptively determined, and the sub-bands whose energy proportion exceeds the threshold are retained; The multidimensional feature vector specifically includes the emotion intensity index E, the knowledge transfer efficiency coefficient K, and the classroom attention concentration A; The energy and variance of the high-frequency sub-bands are used to obtain the intensity index E, and the formula is: ， in, is the set of high frequency sub-bands, is the L2 norm square of the sub-band wavelet coefficients, is the variance, , is the weight coefficient; The knowledge transfer efficiency coefficient K is obtained based on the fundamental frequency stability and semantic information entropy of the teacher's voice. The formula is: ， in, is the fundamental frequency of the teacher's voice, is the semantic sentiment score, is the total duration of the lesson, It is the inverse of semantic information entropy. The lower the entropy, the more concentrated the knowledge transfer. Based on the low-frequency energy proportion of the student's posture signal P(t) and the eye movement entropy, the classroom attention concentration A is obtained. The formula is: , in, Gesture signals for students The low-frequency wavelet packet coefficients of is the eye movement trajectory entropy, is the total number of fixations, , A weight parameter for balancing low-frequency energy and eye movement entropy.

3. The teaching style optimization method based on teaching quality feedback according to claim 1, characterized in that: The method of constructing a teaching effect evaluation matrix based on the obtained multidimensional feature vector and obtaining a nonlinear mapping relationship between teaching behavior characteristics and preset teaching objectives specifically includes: Based on the obtained multi-dimensional feature vector V=[E,K,A] and the preset teaching target vector T, the teaching effect evaluation matrix is ​​constructed: ， in, is a Gaussian kernel function that measures the nonlinear similarity between features and targets. is the target weight vector, is the Gaussian kernel bandwidth, which controls the similarity decay speed; Based on the deep neural network architecture, obtain the mapping function from the learning feature vector V to the teaching target T .

4. The teaching style optimization method based on teaching quality feedback according to claim 1, characterized in that: The method of generating an adversarial network based on the nonlinear mapping relationship between teaching behavior characteristics and preset teaching objectives, obtaining a teaching behavior adjustment plan, and verifying the effectiveness of the plan based on a historical teaching case library specifically includes: Based on the historical teaching quality data in the historical case library, it is input into the adversarial network for alternating training, with the generator minimizing the generator loss and the discriminator minimizing the discriminator loss; The generator part adopts a conditional variational autoencoder (CVAE), whose latent space embeds the syllabus constraints; The discriminator part includes a rule-based knowledge graph verification module for connecting the subject knowledge ontology library to verify the logical consistency of the teaching strategy, and a statistical feature comparison module for calculating the Wasserstein distance between the generated strategy and excellent teaching cases; Add a gradient penalty term to the discriminator loss to obtain stable training results; The multi-dimensional feature vector group under optimal adjustment is obtained through iterative optimization of the generator; Based on the generated adjustment plan, its similarity with the cases with higher feedback on history teaching quality is calculated to obtain the effectiveness of the generated adjustment plan.

5. The teaching style optimization method based on teaching quality feedback according to claim 1, characterized in that: The step of calculating the similarity between the generated adjustment plan and the case with higher feedback on the quality of history teaching and obtaining the effectiveness of the generated adjustment plan specifically includes: The similarity calculation formula is: , Among them, the threshold Indicates high similarity, the pass rate needs to be greater than 70%.

6. The teaching style optimization method based on teaching quality feedback according to claim 1, characterized in that: The dynamic compensation mechanism based on the adjustment scheme and optimizing the scheme parameters according to the secondary feedback data after the scheme is executed specifically include: Based on each generated adjustment plan, obtain real-time teaching quality feedback data and teaching behavior data, and calculate the error vector between it and the teaching target vector ; The compensation amount is calculated based on the PID control update formula, which is: , in, Control the size of the error, The impact of cumulative error, Control error change rate, is the integral term of the error, indicating the accumulation of past errors, is the rate of change of the error, indicating the rate of error; Based on the PID control update formula, the new parameter feature vector E, K, A is calculated and applied to the new teaching plan to obtain the secondary feedback data and calculate the new error vector between it and the teaching target vector. ; Based on the secondary feedback error vector Further adjust the teaching plan. If the error vector If it is small enough, stop adjusting; if it is still large, repeat the above process to further optimize the parameters.

7. The teaching style optimization method based on teaching quality feedback according to claim 1 is characterized in that: The construction of the teaching style evolution map to form a traceable teaching style optimization trajectory for the iterative path of the teaching plan specifically includes: Based on each iteration of the teaching plan, data is recorded for each dimension; Use graph database to store nodes and edges. Nodes include teaching behaviors and student feedback, and edges include optimization paths and effect associations. If the current solution improves students’ higher-order thinking, the weight of the edge will be increased; otherwise, the teaching plan will be refined and the solution parameters will be optimized; Based on the iteration records, the timeline shows the style evolution, the heat map shows the effectiveness of the strategy, and the trajectory map of each style optimization is drawn; The decision basis for a specific iteration is located through SPARQL queries, and an early warning is triggered when the solution deviates from the historical optimal range.

Images