Demonstration system and method for mathematics teaching

Abstract

The invention provides a demonstration system and method for mathematics teaching, and the method comprises the steps: extracting the behavior response characteristics of students in different interaction dimensions in each demonstration link, and determining the logic correlation characteristics between teaching animation segments in adjacent demonstration links; and performing multi-dimensional feature fusion on the logic association features of the teaching animation and all behavior response features to obtain fusion feature vectors of the teaching animation in different interaction dimensions, and performing multi-modal fusion evaluation on a demonstration effect in a demonstration process through the fusion feature vectors of the teaching animation. The effect difference characteristics of all the demonstration links are obtained; and screening out teaching animation segments with substandard demonstration effects from the demonstration process of mathematical demonstration teaching by using each effect difference feature, and further performing segmentation adjustment on the mathematical demonstration teaching based on each teaching animation segment with substandard demonstration effects. Based on the scheme, multi-dimensional segmented adjustment of multi-dimensional learning behaviors of students in mathematics demonstration teaching can be realized.

Description

Technical Field

[0001] The present application relates to the field of teaching demonstration technology, and more specifically, to a demonstration system and method for mathematics teaching. Background Art

[0002] With the continuous deepening of educational informatization, the abstract nature of mathematical knowledge makes students face great difficulties in the learning process. The demonstration teaching method presents abstract mathematical concepts, formulas and principles to students in an intuitive and vivid way, helping students to gain perceptual knowledge and thus better understand and master knowledge. The continuous updating of modern teaching technology and media, such as: real objects, models, pictures, slides, projections and geometric drawing boards, provides powerful tools for mathematical demonstration teaching.

[0003] The teaching effectiveness evaluation in existing mathematics teaching technology relies too much on teachers' subjective experience and judgment, resulting in a lag in problem discovery. Teachers usually need to use indirect methods such as classroom observation, homework correction and test results to discover students' knowledge weaknesses after class. The passive feedback mechanism makes teaching adjustments often miss the best time for intervention. In addition, existing mathematics teaching technology uses single-dimensional behavioral indicators (for example: the accuracy rate of answering questions), which is difficult to fully reflect the true cognitive state. The lagging single-dimensional evaluation leads to a lack of targeted remedial measures, which in turn affects the accuracy of the teaching effectiveness evaluation in the next stage. Therefore, how to achieve multi-dimensional segmented regulation of students' multi-dimensional learning behaviors in mathematics demonstration teaching has become a difficult problem faced by the industry. Summary of the Invention

[0004] The present application provides a demonstration system and method for mathematics teaching, which can realize multi-dimensional segmented adjustment of students' multi-dimensional learning behaviors in mathematics demonstration teaching.

[0005] In a first aspect, the present application provides a teaching demonstration adjustment method for adjusting a mathematics teaching demonstration in a mathematics teaching demonstration system, the method comprising: Collect students' response behavior information in different interaction dimensions when using intelligent demonstration terminals for mathematics demonstration teaching; Acquire multiple demonstration links of the intelligent demonstration terminal in mathematics demonstration teaching, extract the student's behavioral response characteristics in different interaction dimensions of each demonstration link from the reaction behavior information, and determine the logical association characteristics between teaching animation segments in adjacent demonstration links; Based on the attention mechanism, multi-dimensional feature fusion is performed on the logical association features and all behavioral response features to obtain fusion feature vectors of teaching animations in different interaction dimensions. The demonstration effect of mathematics demonstration teaching is evaluated by multimodal fusion through the fusion feature vectors, thereby obtaining the effect difference characteristics of each demonstration link; The different effect features are used to screen out teaching animation clips with substandard demonstration effects from the demonstration process of mathematics demonstration teaching, and then the mathematics demonstration teaching is segmented and adjusted based on the teaching animation clips with substandard demonstration effects.

[0006] In some embodiments, extracting the student's behavioral response characteristics in different interaction dimensions of each demonstration link from the reaction behavior information specifically includes: For each demonstration session, filtering out response behaviors of various interaction dimensions in the demonstration session from the response behavior information; The behavioral response characteristics of students in different interactive dimensions of the demonstration link are determined through all response behaviors, and then the behavioral response characteristics of students in different interactive dimensions of each demonstration link are obtained.

[0007] In some embodiments, determining the logical association features between teaching animation segments in adjacent demonstration sessions specifically includes: The logical similarity between adjacent demonstration links is calculated using the dynamic time warping algorithm; Based on finite state machine, we examine the state transfer characteristics of the previous link behavior to the subsequent link behavior in mathematics demonstration teaching; A cross-correlation analysis is performed on the behavioral associations between adjacent demonstration links in mathematics demonstration teaching according to the logical similarity and the state transition characteristics, so as to obtain the logical association characteristics between the teaching animation clips in adjacent demonstration links.

[0008] In some embodiments, multi-dimensional feature fusion is performed on the logical association features and all behavioral response features based on the attention mechanism to obtain fusion feature vectors of teaching animations in different interaction dimensions, specifically including: Based on the importance and difficulty of knowledge points in mathematics demonstration teaching, the feature-level attention and link-level attention of the intelligent demonstration terminal in different interaction dimensions are determined; The logical association features and all behavioral response features are integrated into behavioral feature values in each interaction dimension through the feature-level attention and the link-level attention; The fusion feature vectors of teaching animations in different interaction dimensions are determined based on all behavioral feature values.

[0009] In some embodiments, the multimodal fusion evaluation of the demonstration effect of mathematics demonstration teaching is performed by using the fusion feature vector, and the effect difference characteristics of each demonstration link are obtained, which specifically include: For each demonstration link in mathematics demonstration teaching, extracting the link behavior features of each interactive dimension in the demonstration link from the fusion feature vector; Determine the benchmark vector of the demonstration link based on the overall performance information of the demonstration link in mathematics demonstration teaching; All the behavioral features of each link are multimodally fused to obtain the feature vector to be evaluated for the demonstration link; The feature vector to be evaluated is abnormally matched with the reference vector to obtain effect difference features of the demonstration link, and then the effect difference features of each demonstration link are obtained.

[0010] In some embodiments, using various effect difference features to screen out teaching animation clips with substandard demonstration effects in the mathematics demonstration teaching process specifically includes: For each demonstration link, the causal anomaly features of each teaching animation clip are extracted from the effect difference features of the demonstration link; Map all causal anomaly features to the knowledge graph of the demonstration link, and then use the knowledge graph of the demonstration link to mark the demonstration anomaly nodes of the demonstration link, and then obtain the demonstration anomaly nodes of each demonstration link; Through all the abnormal demonstration nodes, the teaching animation clips in which the demonstration effect does not meet the standards in mathematics demonstration teaching are determined.

[0011] In some embodiments, the intelligent demonstration terminal includes an infrared eye tracking module, a capacitive touch screen, and a microphone array.

[0012] In a second aspect, the present application provides a mathematics teaching demonstration system, comprising a teaching demonstration adjustment unit, wherein the teaching demonstration adjustment unit comprises: The acquisition module is used to collect students' response behavior information in different interaction dimensions when using the intelligent demonstration terminal for mathematics demonstration teaching; a processing module for obtaining multiple demonstration links of the intelligent demonstration terminal in mathematics demonstration teaching, extracting the behavioral response characteristics of students in different interaction dimensions of each demonstration link from the reaction behavior information, and determining the logical association characteristics between teaching animation segments in adjacent demonstration links; The processing module is further configured to perform multi-dimensional feature fusion on the logical association features and all behavioral response features based on an attention mechanism to obtain fusion feature vectors of teaching animations in different interaction dimensions, and to perform multimodal fusion evaluation on the demonstration effect of mathematics demonstration teaching through the fusion feature vectors to obtain effect difference features of each demonstration link; The execution module is used to use various effect difference features to filter out teaching animation clips with substandard demonstration effects in the mathematics demonstration teaching from the demonstration process of the mathematics demonstration teaching, and then adjust the mathematics demonstration teaching in segments based on the teaching animation clips with substandard demonstration effects.

[0013] In a third aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the computer device executes the above-mentioned teaching demonstration adjustment method.

[0014] In a fourth aspect, the present application provides a computer-readable storage medium, in which instructions or codes are stored. When the instructions or codes are run on a computer, the computer implements the above-mentioned teaching demonstration adjustment method when executing.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: In a demonstration system and method for mathematics teaching provided by the present application, when using an intelligent demonstration terminal to conduct mathematics demonstration teaching, students' reaction behavior information in different interaction dimensions is collected; multiple demonstration links of the intelligent demonstration terminal in the mathematics demonstration teaching are obtained, and then the students' behavior response characteristics in different interaction dimensions of each demonstration link are extracted from the reaction behavior information, and the logical association characteristics between the teaching animation segments in adjacent demonstration links are determined; based on the attention mechanism, multi-dimensional feature fusion is performed on the logical association features and all behavior response features to obtain a fusion feature vector of the teaching animation in different interaction dimensions, and the demonstration effect of the mathematics demonstration teaching is evaluated by multimodal fusion through the fusion feature vector to obtain the effect difference characteristics of each demonstration link; each effect difference feature is used to screen out teaching animation segments with substandard demonstration effects in the mathematics demonstration teaching from the demonstration process of the mathematics demonstration teaching, and then the mathematics demonstration teaching is segmented and adjusted based on the teaching animation segments with substandard demonstration effects.

[0016] It can be seen that in this application, various effect difference characteristics are used to screen out teaching animation clips with substandard demonstration effects in the mathematics demonstration teaching from the demonstration process of mathematics demonstration teaching, and then the mathematics demonstration teaching is segmented and adjusted based on the teaching animation clips with substandard demonstration effects; first, the logical correlation characteristics of the teaching animation are determined to obtain the cognitive state transmission rules and knowledge transfer paths between adjacent teaching links, thereby constructing a dynamic influence network between teaching links, and using the dynamic time warping algorithm to quantify the logical similarity between links, combined with the finite state machine modeling state transition probability, it is possible to accurately identify the potential impact of the previous link on subsequent learning. When there is a strong correlation between the attention distraction of the previous link and the subsequent link, the intelligent demonstration terminal can automatically establish an abnormal conduction model across links. This correlation analysis makes the teaching adjustment predictable, and can predict and prevent possible cognitive faults in the current link, thereby It can improve the effectiveness of the evaluation of mathematics teaching effects; then, by determining the effect difference characteristics, the precise positioning and severity classification of the teaching links that require priority intervention can be obtained, thereby realizing differentiated adjustment strategy allocation. This process uses a multimodal fusion evaluation model to compare the fusion feature vector of the teaching animation with the knowledge graph benchmark, not only calculating the overall deviation, but also analyzing the abnormal contribution ratio of each dimension. When the link abnormality reaches the abnormality threshold and is mainly caused by the confusion of the touch trajectory, the intelligent demonstration terminal will automatically trigger segmented adjustment, which can reduce the teaching animation clips that do not meet the demonstration effect. It can return to normal level after a single adjustment, while avoiding invalid interference with the normal teaching links, thereby further improving the effectiveness of the evaluation of mathematics teaching effects; in summary, based on the above scheme, multi-dimensional segmented adjustment of students' multi-dimensional learning behaviors in mathematics demonstration teaching can be realized, thereby improving the effectiveness of mathematics teaching effect evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0018] Figure 1 is an exemplary flow chart of a teaching demonstration adjustment method according to some embodiments of the present application; Figure 2 This is a schematic diagram of a process for determining a teaching animation clip whose demonstration effect does not meet the standards according to some embodiments of the present application; Figure 3 is a structural diagram of a teaching demonstration adjustment unit according to some embodiments of the present application; Figure 4It is a structural diagram of a computer device for implementing a teaching demonstration adjustment method according to some embodiments of the present application. DETAILED DESCRIPTION

[0019] In order to better understand the technical solution of the present application, the technical solution of the present application will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0020] refer to Figure 1 , which is an exemplary flow chart of a teaching demonstration adjustment method according to some embodiments of the present application, and the teaching demonstration adjustment method mainly includes the following steps: In step 101, when using an intelligent demonstration terminal to conduct mathematics demonstration teaching, students' reaction behavior information in different interaction dimensions is collected.

[0021] It should be noted that, in this application, the reaction behavior information represents the multimodal interaction data of students in mathematics demonstration teaching, and the reaction behavior information includes eye movement trajectory, touch operation and voice response; different interaction dimensions refer to eye movement dimension, tactile dimension and voice dimension, and the intelligent demonstration terminal includes an infrared eye tracking module, a capacitive touch screen and a microphone array; in specific implementation, when using the intelligent demonstration terminal for mathematics demonstration teaching, the infrared eye tracking module of the intelligent demonstration terminal is used to collect the eye movement data of the intelligent demonstration terminal in the mathematics demonstration teaching within a specified time period (the default is 30 days), the capacitive touch screen of the intelligent demonstration terminal is used to collect the touch data of the intelligent demonstration terminal in the mathematics demonstration teaching within the specified time period, and the microphone array of the intelligent demonstration terminal is used to collect the voice data of the intelligent demonstration terminal in the mathematics demonstration teaching within the specified time period, and the collection of eye movement data, touch data and voice data is used as the student's reaction behavior information in different interaction dimensions.

[0022] In step 102, multiple demonstration links of the intelligent demonstration terminal in mathematics demonstration teaching are obtained, and then the behavioral response characteristics of students in different interaction dimensions of each demonstration link are extracted from the reaction behavior information, and the logical correlation characteristics between the teaching animation clips in adjacent demonstration links are determined.

[0023] It should be noted that in this application, the demonstration link refers to an independent teaching unit in mathematics teaching divided by knowledge points or teaching logic, and the demonstration link has clear teaching objectives and time boundaries; in specific implementation, obtaining multiple demonstration links in the mathematics demonstration teaching of the intelligent demonstration terminal can be achieved in the following way, namely: obtaining multiple demonstration links in the mathematics demonstration teaching from the console of the intelligent demonstration terminal.

[0024] In some embodiments, extracting the behavioral response characteristics of students in different interaction dimensions of each demonstration link from the reaction behavior information can be achieved by using the following steps: For each demonstration session, filtering out response behaviors of various interaction dimensions in the demonstration session from the response behavior information; The behavioral response characteristics of students in different interactive dimensions of the demonstration link are determined through all response behaviors, and then the behavioral response characteristics of students in different interactive dimensions of each demonstration link are obtained.

[0025] In a specific implementation, first, for each demonstration link, the response behavior of each interactive dimension in the demonstration link is filtered out from the reaction behavior information. This can be achieved in the following manner: for each demonstration link, the eye movement behavior of the demonstration link in the eye movement dimension is obtained from the eye movement data of the reaction behavior information, the touch behavior of the demonstration link in the voice dimension is obtained from the touch data of the reaction behavior information, and the voice behavior of the demonstration link in the eye movement dimension is obtained from the voice data of the reaction behavior information, so as to obtain the response behavior of each interactive dimension in the demonstration link; then, the behavioral response characteristics of the students in different interactive dimensions of the demonstration link are determined through all the response behaviors, and then the behavioral response characteristics of the students in different interactive dimensions of each demonstration link are obtained. This can be achieved in the following manner: for each interactive dimension, the characteristics of the response behavior corresponding to the interactive dimension in the demonstration link are generated through multi-level feature engineering (for example, a polynomial feature generation algorithm) as the response sub-feature of the interactive dimension, so as to obtain the response sub-feature of each interactive dimension. The set of all response sub-features can be used as the behavioral response characteristics of the students in different interactive dimensions of the demonstration link. The behavioral response characteristics of the students in different interactive dimensions of each demonstration link can be obtained through the above method.

[0026] It should be noted that in this application, behavioral response features refer to the set of learning features exhibited by students in a specific teaching link; response behavior refers to the specific interactive actions generated by students during the teaching process; multi-level feature engineering.

[0027] In some embodiments, determining the logical association features between teaching animation segments in adjacent demonstration sessions may be achieved by using the following steps: The logical similarity between adjacent demonstration links is calculated using the dynamic time warping algorithm; Based on finite state machine, we examine the state transfer characteristics of the previous link behavior to the subsequent link behavior in mathematics demonstration teaching; A cross-correlation analysis is performed on the behavioral associations between adjacent demonstration links in mathematics demonstration teaching according to the logical similarity and the state transition characteristics, so as to obtain the logical association characteristics between the teaching animation clips in adjacent demonstration links.

[0028] In the specific implementation, first, the logical similarity between adjacent demonstration links is calculated by the dynamic time warping algorithm, which can be implemented in the following way, namely: for each group of adjacent demonstration links, the dynamic time warping algorithm is used to calculate the behavioral association value between the group of adjacent demonstration links, that is, the cosine similarity between the behavioral response characteristics of the two demonstration links in the group of adjacent demonstration links is used as the behavioral association value of the group of adjacent demonstration links. The behavioral association value between each group of adjacent demonstration links can be obtained by the above method, and the set of all behavioral association values can be used as the value range of logical similarity to obtain the logical similarity between adjacent demonstration links; then, based on the finite state machine, the state transfer characteristics of the behavior of the previous link to the behavior of the subsequent link in the mathematics demonstration teaching can be implemented in the following way, namely: the mathematics demonstration link is abstracted into a state node using a finite state machine (for example: "concept introduction S1→formula derivation S2→example application S3"), so as to associate each state node with a legal behavior set (for example: in the S2 state, it must be satisfied: the number of touch formula edits ≥ 3 times), and the historical data statistics are used to predict A transfer condition matrix is set up to detect in real time whether the state jump sequence meets the transfer condition. When an undefined transfer is detected (for example: S1→S3 skips S2), a state fracture feature is generated as the state transfer feature of the behavior of the previous link to the behavior of the subsequent link in mathematics demonstration teaching; finally, according to the logical similarity and the state transfer feature, a cross-correlation analysis is performed on the behavioral association between adjacent demonstration links in mathematics demonstration teaching, and the logical association feature between the teaching animation segments in adjacent demonstration links is obtained. This can be achieved in the following way, namely: initialize a correlation analysis model based on cross entropy loss, use logical similarity as the similarity measurement parameter in the correlation analysis model, use state transfer feature as the transfer probability in the correlation analysis model, use the correlation analysis model to cross-analyze the behavioral association between the teaching animation segments between adjacent demonstration links in mathematics demonstration teaching, and use the result of the cross-analysis as the logical association value between each teaching animation segment in the adjacent demonstration link, so that the set of all logical association values is used as the logical association feature between the teaching animation segments in the adjacent demonstration link.

[0029] It should be noted that in this application, the logical association feature represents the spatiotemporal correlation feature of the student interaction behavior between adjacent teaching links, and the logical association feature can be used to analyze the continuity influence between teaching links; the teaching animation clip represents the minimum demonstration unit in each demonstration link; the logical similarity represents the degree of matching of student behavior patterns in different teaching links; the state transition feature represents the change feature of the student's cognitive state from the previous teaching link to the subsequent link, and the state transition feature reflects the evolution law of the knowledge mastery process; the correlation analysis model is a probabilistic statistical analysis framework constructed based on cross-entropy loss. The correlation analysis model optimizes the teaching process evaluation by quantifying the behavioral correlation between teaching links. The correlation analysis model uses the logical similarity calculated by dynamic time warping as the similarity measurement parameter to reflect the matching degree of student behavior patterns between links; at the same time, the state transition feature of the finite state machine model is converted into a transition probability to characterize the evolution law of the cognitive state. By minimizing the cross-entropy loss function, the model effectively captures the differences in the probability distribution of behavioral data between adjacent links. The true distribution of the loss function is generated by state transition features, while the predicted distribution is calculated based on logical similarity. The final cross-analysis output is encoded as the logical association features of the teaching animation, which not only reflects the strength of causal relationships in time series but also reflects the efficiency of knowledge transfer across links. This model solves the problem of noise interference in teaching animation data through probabilistic processing, providing an interpretable quantitative basis for subsequent abnormal node detection.

[0030] In step 103, multi-dimensional feature fusion is performed on the logical association features and all behavioral response features based on the attention mechanism to obtain fusion feature vectors of teaching animations in different interaction dimensions. The demonstration effect of mathematics demonstration teaching is evaluated by multimodal fusion through the fusion feature vectors to obtain the effect difference characteristics of each demonstration link.

[0031] In some embodiments, multi-dimensional feature fusion is performed on the logical association features and all behavioral response features based on the attention mechanism to obtain fused feature vectors of teaching animations in different interaction dimensions. The following steps can be used to achieve this: Based on the importance and difficulty of knowledge points in mathematics demonstration teaching, the feature-level attention and link-level attention of the intelligent demonstration terminal in different interaction dimensions are determined; The logical association features and all behavioral response features are integrated into behavioral feature values in each interaction dimension through the feature-level attention and the link-level attention; The fusion feature vectors of teaching animations in different interaction dimensions are determined based on all behavioral feature values.

[0032] In the specific implementation, first, based on the importance and difficulty of knowledge points in mathematics demonstration teaching, the feature-level attention and link-level attention of the intelligent demonstration terminal in different interaction dimensions are determined, which can be implemented in the following way, namely: using the importance and difficulty of knowledge points in mathematics demonstration teaching to construct a mathematics knowledge point map, for each interaction dimension, using a cross-dimensional interaction attention mechanism to extract the feature-level attention and link-level attention of the interaction dimension in the mathematics knowledge point map, the feature-level attention and link-level attention of each interaction dimension in the mathematics knowledge point map can be obtained by the above method; then, the logical association features and all behavioral response features are fused into behavioral feature values in each interaction dimension through the feature-level attention and the link-level attention, which can be implemented in the following way, namely: for each interaction dimension, Initialize a weighted fusion normalization model, take the logical association features of the teaching animation as the input features in the normalization model, take each behavioral response feature as the weighting factor in the normalization model, take the feature-level attention as the feature weight in the normalization model, take the link-level attention as the link weight in the normalization model, use the normalization model to perform feature normalization fusion, take the normalized fusion features as the behavioral feature values in the interaction dimension, and the behavioral feature values in each interaction dimension can be obtained by the above method; finally, the fusion feature vectors of the teaching animation in different interaction dimensions are determined according to all the behavioral feature values, which can be achieved in the following way: use the feature splicing algorithm based on the attention mechanism in the intelligent demonstration terminal to splice all the behavioral feature values into the fusion feature vectors of the teaching animation in different interaction dimensions.

[0033] It should be noted that, in this application, the fusion feature vector of the teaching animation is a standardized vector used to comprehensively characterize the student behavior pattern; the feature-level attention represents a dynamic weight allocation mechanism for features of different interaction dimensions; the link-level attention represents a quantitative indicator of the degree of focus on key links in the teaching link sequence; the behavioral feature value is a representation value that quantifies the behavior in a single interaction dimension; the normalization model is a multimodal feature integration framework based on weighted fusion, which realizes the standardized representation of teaching animation features through a hierarchical attention mechanism. The normalization model uses the logical association features of the teaching animation as the basic input, and dynamically allocates weights through a dual attention mechanism. In the feature dimension, feature-level attention is used to calculate the contribution weight of each interaction dimension (eye movement / touch / voice); in the link dimension, link-level attention is used to determine the importance of different teaching links. The normalization model uses Min-Max normalization to unify the scale of the fusion features to ensure that behavioral response features of different dimensions (for example, millisecond-level delay and pixel-level coordinates) are comparable. The final output behavioral feature value not only retains the semantic information of the original data, but also eliminates the dimensional differences between modalities, providing standardized input for subsequent multimodal evaluation. The normalization model effectively solves the problem of feature heterogeneity fusion in teaching animation analysis through interpretable weight distribution.

[0034] In some embodiments, the multimodal fusion evaluation of the demonstration effect of mathematics demonstration teaching is performed by using the fusion feature vector, and then the effect difference characteristics of each demonstration link are obtained. The following steps can be used to achieve this: For each demonstration link in mathematics demonstration teaching, extracting the link behavior features of each interactive dimension in the demonstration link from the fusion feature vector; Determine the benchmark vector of the demonstration link based on the overall performance information of the demonstration link in mathematics demonstration teaching; All the behavioral features of each link are multimodally fused to obtain the feature vector to be evaluated for the demonstration link; The feature vector to be evaluated is abnormally matched with the reference vector to obtain effect difference features of the demonstration link, and then the effect difference features of each demonstration link are obtained.

[0035] In the specific implementation, first, for each demonstration link in mathematics demonstration teaching, extracting the link behavior characteristics of each interactive dimension in the demonstration link from the fusion feature vector can be implemented in the following way, namely: for each demonstration link in mathematics demonstration teaching, obtaining the behavior feature values in each interactive dimension in the demonstration link from the fusion feature vector of the teaching animation as the corresponding link behavior characteristics, and thus obtaining the link behavior characteristics of each interactive dimension in the demonstration link; secondly, determining the benchmark vector of the demonstration link based on the overall performance information of the demonstration link in mathematics demonstration teaching can be implemented in the following way, namely: collecting the link feature vectors of at least 100 groups of excellent teachers as positive samples, clustering the positive samples using the K-means clustering algorithm (K=5), selecting the center point of each cluster as a candidate benchmark, obtaining the historical test scores of the students in the current class as the average level, and updating the cluster center with the newly generated excellent samples after completing 5 teaching cycles, and using the average level to select the most matching cluster center from the candidate benchmarks as the benchmark vector; then, multimodally fusing all the link behavior characteristics to obtain The feature vector to be evaluated of the demonstration link can be implemented in the following manner, namely: use a feature fusion algorithm based on the attention mechanism (for example: attention-guided feature fusion algorithm) to perform feature fusion on all link behavior features, and use the fused vector as the feature vector to be evaluated of the demonstration link, for example: assign weights to the dimensional features of each demonstration link, visual dimension weight = 0.5, touch dimension weight = 0.3, voice dimension weight = 0.2 (adjustable according to the link type); finally, perform abnormal matching on the feature vector to be evaluated and the reference vector to obtain the effect difference feature of the demonstration link, and then obtain the effect difference feature of each demonstration link. It can be implemented in the following manner, namely: use cosine similarity to measure the matching degree between the vector to be evaluated and the reference vector, that is, use the cosine similarity between the feature vector to be evaluated and the reference vector as the vector matching degree, convert the vector matching degree into the effect difference feature, that is, the effect difference feature = 1-vector matching degree, and the effect difference feature of the demonstration link can be obtained. The effect difference feature of each demonstration link can be obtained in the above manner.

[0036] It should be noted that in this application, the effect difference feature represents the degree to which the student behavior characteristics in the current teaching link deviate from the baseline value, and the effect difference feature can be used to quantify the abnormal degree of teaching effect; the link behavior feature represents the characterization of user behavior in a specific teaching link; the benchmark vector represents the standard behavior feature set under the teaching demonstration, which serves as a reference benchmark for effect evaluation; the feature vector to be evaluated represents the behavior feature set to be evaluated in the current teaching link, and the feature vector to be evaluated can be used for comparison and analysis with the benchmark vector.

[0037] In step 104, various effect difference features are used to screen out teaching animation clips with substandard demonstration effects from the demonstration process of the mathematics demonstration teaching, and then the mathematics demonstration teaching is segmented and adjusted based on the teaching animation clips with substandard demonstration effects.

[0038] In some embodiments, the teaching animation clips with substandard demonstration effects in the mathematics demonstration teaching are screened out from the demonstration process of the mathematics demonstration teaching using the various effect difference features. Figure 2 As described above, the figure is a schematic diagram of a process for determining a teaching animation segment whose demonstration effect does not meet the standard in some embodiments of the present application. In this embodiment, determining a teaching animation segment whose demonstration effect does not meet the standard can be achieved by using the following steps: In step 1041, for each demonstration link, the causal anomaly features of each teaching animation segment are extracted from the effect difference features of the demonstration link; In step 1042, all causal anomaly features are mapped to the knowledge graph of the demonstration link, and then the demonstration anomaly nodes of the demonstration link are marked using the knowledge graph of the demonstration link, thereby obtaining the demonstration anomaly nodes of each demonstration link; In step 1043, the teaching animation segments in which the demonstration effect does not meet the standards in the mathematics demonstration teaching are determined through all the demonstration abnormal nodes.

[0039] In specific implementation, first, for each demonstration link, the causal anomaly features of each teaching animation segment are extracted from the effect difference features of the demonstration link. This can be achieved in the following way: for each demonstration link, the causal relationship between the effect difference features between the demonstration link and the adjacent demonstration link is analyzed using the Granger causality test, and node pairs with significant causal influence (i.e., causal relationship less than 0.05) are screened out. For node pairs with causal relationship, the abnormality transmission strength of the node pair is recorded (for example, for every 1 unit increase in the abnormality of node A, the abnormality of node B increases by 0.6 units). Based on the Bayesian network, all the abnormality transmission strengths are used to construct an abnormal propagation model for each teaching animation segment in the demonstration link. When each teaching animation segment is abnormal, the conditional probability of its directly associated teaching animation segment being abnormal in the subsequent link is calculated using the abnormal propagation model. The strong correlation relationship with a conditional probability greater than a preset strong correlation probability threshold (which can be preset through historical experience) is retained as the causal anomaly feature. A triple feature vector is generated for each teaching animation segment as a strong correlation relationship, i.e., the sum of the upstream abnormality input, the sum of the downstream abnormality output, and the cross-link influence range. The causal anomaly features of each teaching animation segment in the demonstration link can be obtained.

[0040] Then, in the specific implementation, all causal anomaly features are mapped to the knowledge graph of the demonstration link, and then the demonstration anomaly nodes of the demonstration link are marked using the knowledge graph of the demonstration link, and then the demonstration anomaly nodes of each demonstration link can be obtained. The following method can be used to implement it, namely: construct a knowledge point graph according to the mathematical subject system, the nodes are the knowledge points in the intelligent demonstration terminal, the edges are the logical dependencies in the intelligent demonstration terminal, and each abnormal causal anomaly feature is aligned with the knowledge point graph. If a certain teaching animation segment covers the knowledge point K, the abnormal feature of the teaching animation segment is accumulated to the abnormal load attribute of K, and the abnormal influence is propagated along the edge of the graph. The parent node The abnormal load is transferred to the child nodes according to the weight coefficient (default 0.6), so as to obtain the abnormal load value of each teaching animation segment, and the teaching animation segment with an abnormal load value greater than the preset load threshold is used as the demonstration abnormal node of the demonstration link. The demonstration abnormal node of each demonstration link can be obtained in the above way; finally, the teaching animation segment with substandard demonstration effect in mathematics demonstration teaching can be determined through all demonstration abnormal nodes. The following method can be used, that is, the set of all demonstration abnormal nodes can be used as the node range of the teaching animation segment with substandard demonstration effect, and the teaching animation segment with substandard demonstration effect in mathematics demonstration teaching can be obtained.

[0041] It should be noted that, in this application, the teaching animation clips with substandard demonstration effects represent steps in the teaching process where the knowledge transfer effect is significantly lower than expected due to abnormal student interaction behaviors; the causal anomaly feature represents the transmission relationship of abnormal behaviors between adjacent teaching links, and the causal anomaly feature can be used to locate the root cause of the problem; the demonstration anomaly node represents a teaching weakness with abnormal interactions in the knowledge graph.

[0042] In some embodiments, segmented adjustment of mathematics demonstration teaching based on each teaching animation segment that does not meet the demonstration effect can be achieved in the following manner, namely: for each teaching animation segment that does not meet the demonstration effect, the abnormal load value of the teaching animation segment that does not meet the demonstration effect is used as the teaching animation segment that does not meet the demonstration effect; if the abnormality is less than 1.5 times the preset load threshold, an immediate pop-up window is triggered to prompt the teacher to strengthen the explanation; if the abnormality is greater than or equal to 1.5 times the preset load threshold, a remedial teaching link is inserted: the teaching time of the node is extended by 50%, and an alternative demonstration method is enabled (for example: converting abstract algebraic derivation into a graphical dynamic demonstration).

[0043] In addition, in another aspect of the present application, in some embodiments, the present application provides a mathematics teaching demonstration system, the mathematics teaching demonstration system includes a teaching demonstration adjustment unit, reference Figure 3 , which is a schematic diagram of the structure of a teaching demonstration adjustment unit according to some embodiments of the present application. The teaching demonstration adjustment unit includes: an acquisition module 201, a processing module 202 and an execution module 203, which are described as follows: The acquisition module 201 in this application is mainly used to collect students' reaction behavior information in different interaction dimensions when using the intelligent demonstration terminal for mathematics demonstration teaching; Processing module 202, in this application, is used to obtain multiple demonstration links of the intelligent demonstration terminal in mathematics demonstration teaching, and then extract the behavioral response characteristics of students in different interaction dimensions of each demonstration link from the reaction behavior information, and determine the logical association characteristics between teaching animation segments in adjacent demonstration links; It should be noted that the processing module 202 is further configured to perform multi-dimensional feature fusion on the logical association features and all behavioral response features based on the attention mechanism, thereby obtaining fusion feature vectors of the teaching animations in different interaction dimensions, and perform multimodal fusion evaluation on the demonstration effect of the mathematics demonstration teaching through the fusion feature vectors, thereby obtaining the effect difference characteristics of each demonstration link; Execution module 203. In this application, execution module 203 is mainly used to use various effect difference characteristics to filter out teaching animation clips with substandard demonstration effects in the mathematics demonstration teaching from the demonstration process of the mathematics demonstration teaching, and then perform segmented adjustment on the mathematics demonstration teaching based on the teaching animation clips with substandard demonstration effects.

[0044] The examples of the demonstration system and method for mathematics teaching provided by the embodiments of the present application are described in detail above. It is understandable that the corresponding device includes a hardware structure and / or software module for performing each function in order to realize the above functions. Those skilled in the art should easily appreciate that, in conjunction with the units and algorithm steps of each example described in the embodiments disclosed herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a certain function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to realize the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0045] In some embodiments, the present application also provides a computer device, which includes a memory and a processor, the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the computer device executes the above-mentioned teaching demonstration adjustment method.

[0046] In some embodiments, reference Figure 4 The dotted line in the figure indicates that the unit or module is optional. The figure is a structural diagram of a computer device for implementing a teaching demonstration adjustment method according to an embodiment of the present application. The teaching demonstration adjustment method described in the above embodiment can be Figure 4The computer device shown in the figure is implemented, and the computer device includes at least one processor 301, a memory 302 and at least one communication unit 305. The computer device can be a terminal device, a server or a chip.

[0047] The processor 301 may be a general-purpose processor or a dedicated processor. For example, the processor 301 may be a central processing unit (CPU), which may be used to control the computer device, execute software programs, and process data from the software programs. The computer device may also include a communication unit 305 for inputting (receiving) and outputting (transmitting) signals.

[0048] For example, the computer device may be a chip, the communication unit 305 may be an input and / or output circuit of the chip, or the communication unit 305 may be a communication interface of the chip, and the chip may be a component of a terminal device, a network device, or other device.

[0049] For another example, the computer device may be a terminal device or a server, and the communication unit 305 may be a transceiver of the terminal device or the server, or the communication unit 305 may be a transceiver circuit of the terminal device or the server.

[0050] The computer device may include one or more memories 302, on which a program 304 is stored. The program 304 can be executed by the processor 301 to generate instructions 303, so that the processor 301 executes the method described in the above method embodiment according to the instructions 303. Optionally, data (such as a target audit model) can also be stored in the memory 302. Optionally, the processor 301 can also read data stored in the memory 302. The data can be stored at the same storage address as the program 304, or at a different storage address from the program 304.

[0051] The processor 301 and the memory 302 may be provided separately or integrated together, for example, integrated on a system on chip (SOC) of a terminal device.

[0052] It should be understood that each step of the above method embodiment can be completed by a hardware-based logic circuit or software-based instructions in the processor 301. The processor 301 can be a CPU, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, such as discrete gates, transistor logic devices, or discrete hardware components.

[0053] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0054] For example, in some embodiments, the present application also provides a computer-readable storage medium, in which instructions or codes are stored. When the instructions or codes are run on a computer, the computer implements the above-mentioned teaching demonstration adjustment method when executing.

[0055] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0056] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A teaching demonstration adjustment method for adjusting mathematics demonstration teaching in a mathematics teaching demonstration system, characterized in that: The method comprises the following steps: Collect students' response behavior information in different interaction dimensions when using intelligent demonstration terminals for mathematics demonstration teaching; Acquire multiple demonstration links of the intelligent demonstration terminal in mathematics demonstration teaching, extract the student's behavioral response characteristics in different interaction dimensions of each demonstration link from the reaction behavior information, and determine the logical association characteristics between teaching animation segments in adjacent demonstration links; Based on the attention mechanism, multi-dimensional feature fusion is performed on the logical association features and all behavioral response features to obtain fusion feature vectors of teaching animations in different interaction dimensions. The demonstration effect of mathematics demonstration teaching is evaluated by multimodal fusion through the fusion feature vectors, thereby obtaining the effect difference characteristics of each demonstration link; The different effect features are used to screen out teaching animation clips with substandard demonstration effects from the demonstration process of mathematics demonstration teaching, and then the mathematics demonstration teaching is segmented and adjusted based on the teaching animation clips with substandard demonstration effects.

2. The method according to claim 1, wherein The behavioral response characteristics of students in different interaction dimensions of each demonstration link are extracted from the reaction behavior information, specifically including: For each demonstration session, filtering out response behaviors of various interaction dimensions in the demonstration session from the response behavior information; The behavioral response characteristics of students in different interactive dimensions of the demonstration link are determined through all response behaviors, and then the behavioral response characteristics of students in different interactive dimensions of each demonstration link are obtained.

3. The method according to claim 1, wherein Determining the logical association features between teaching animation segments in adjacent demonstration links specifically includes: The logical similarity between adjacent demonstration links is calculated using the dynamic time warping algorithm; Based on finite state machine, we examine the state transfer characteristics of the previous link behavior to the subsequent link behavior in mathematics demonstration teaching; A cross-correlation analysis is performed on the behavioral associations between adjacent demonstration links in mathematics demonstration teaching according to the logical similarity and the state transition characteristics, so as to obtain the logical association characteristics between the teaching animation clips in adjacent demonstration links.

4. The method according to claim 1, wherein Based on the attention mechanism, the logical association features and all behavioral response features are fused in multiple dimensions to obtain the fusion feature vectors of the teaching animation in different interaction dimensions. Specifically, the following are included: Based on the importance and difficulty of knowledge points in mathematics demonstration teaching, the feature-level attention and link-level attention of the intelligent demonstration terminal in different interaction dimensions are determined; The logical association features and all behavioral response features are integrated into behavioral feature values in each interaction dimension through the feature-level attention and the link-level attention; The fusion feature vectors of teaching animations in different interaction dimensions are determined based on all behavioral feature values.

5. The method according to claim 1, wherein The demonstration effect of mathematics demonstration teaching is evaluated by multimodal fusion through the fusion feature vector, and the effect difference characteristics of each demonstration link are obtained, which specifically include: For each demonstration link in mathematics demonstration teaching, extracting the link behavior features of each interactive dimension in the demonstration link from the fusion feature vector; Determine the benchmark vector of the demonstration link based on the overall performance information of the demonstration link in mathematics demonstration teaching; All the behavioral features of each link are multimodally fused to obtain the feature vector to be evaluated for the demonstration link; The feature vector to be evaluated is abnormally matched with the reference vector to obtain effect difference features of the demonstration link, and then the effect difference features of each demonstration link are obtained.

6. The method according to claim 1, wherein Using the various effect difference features, we can screen out the teaching animation clips that do not meet the demonstration effect standards in mathematics demonstration teaching. Specifically, they include: For each demonstration link, the causal anomaly features of each teaching animation clip are extracted from the effect difference features of the demonstration link; Map all causal anomaly features to the knowledge graph of the demonstration link, and then use the knowledge graph of the demonstration link to mark the demonstration anomaly nodes of the demonstration link, and then obtain the demonstration anomaly nodes of each demonstration link; Through all the abnormal demonstration nodes, the teaching animation clips in which the demonstration effect does not meet the standards in mathematics demonstration teaching are determined.

7. The method according to claim 1, wherein The intelligent demonstration terminal includes an infrared eye tracking module, a capacitive touch screen and a microphone array.

8. A mathematics teaching demonstration system, comprising a teaching demonstration adjustment unit, characterized in that: The teaching demonstration adjustment unit includes: The acquisition module is used to collect students' response behavior information in different interaction dimensions when using the intelligent demonstration terminal for mathematics demonstration teaching; a processing module for obtaining multiple demonstration links of the intelligent demonstration terminal in mathematics demonstration teaching, extracting the behavioral response characteristics of students in different interaction dimensions of each demonstration link from the reaction behavior information, and determining the logical association characteristics between teaching animation segments in adjacent demonstration links; The processing module is further configured to perform multi-dimensional feature fusion on the logical association features and all behavioral response features based on an attention mechanism to obtain fusion feature vectors of teaching animations in different interaction dimensions, and to perform multimodal fusion evaluation on the demonstration effect of mathematics demonstration teaching through the fusion feature vectors to obtain effect difference features of each demonstration link; The execution module is used to use various effect difference features to filter out teaching animation clips with substandard demonstration effects in the mathematics demonstration teaching from the demonstration process of the mathematics demonstration teaching, and then adjust the mathematics demonstration teaching in segments based on the teaching animation clips with substandard demonstration effects.

9. A computer device, characterized in that: The computer device includes a memory and a processor, the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the computer device executes the teaching demonstration adjustment method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions or codes, and when the instructions or codes are run on a computer, the computer implements the teaching demonstration adjustment method according to any one of claims 1 to 7.

Images