Intelligent teaching analysis system and method based on big data

Abstract

The invention provides an intelligent teaching analysis system and method based on big data, and the method comprises the steps: obtaining the real-time physiological monitoring data and behavior monitoring data of a student in a classroom progress process; real-time pattern recognition is carried out on the real-time physiological monitoring data and the behavior monitoring data, and a physiological state index sequence and an action feature vector set are generated; performing multi-modal time sequence correlation analysis on the physiological state index sequence and the action feature vector set to generate a dynamic learning state map; based on the dynamic learning state map, an intelligent teaching analysis result is generated, and the intelligent teaching analysis result comprises a content adjustment instruction and an interaction guide strategy for the current teaching stage. According to the invention, the dynamic adaptability of the classroom teaching strategy and the student participation degree are improved.

Description

Technical Field

[0001] This application relates to the field of big data technology, and in particular, to an intelligent teaching analysis system and method based on big data. Background Art

[0002] With the development of smart education, classroom teaching needs to monitor the attention state and emotional changes of students in real time, and dynamically adjust teaching strategies to improve learning effects. The traditional method relying on subjective observation by teachers is difficult to accurately grasp the real-time state of all students. There is an urgent need to use multi-modal data fusion analysis technology to achieve automatic and refined evaluation of students' learning states and generate personalized teaching intervention plans.

[0003] Currently, there is a solution that uses video analysis technology to collect students' facial expressions and body movement data through classroom cameras, uses a pre-trained behavior recognition model to judge the concentration level of students, and outputs simple classroom interaction suggestions in combination with a preset teaching strategy library. This solution realizes the basic attention monitoring function through the analysis of single-modal behavior data.

[0004] This solution only relies on visual behavior data for analysis, ignoring the reflection of physiological indicators such as heart rate and blood oxygen on emotions and cognitive states, resulting in a single dimension of state evaluation; at the same time, due to the lack of temporal correlation analysis of multi-modal data, it is difficult to capture the internal correlation between attention fluctuations and emotional changes, and the output teaching suggestions deviate from the real learning state. Summary of the Invention

[0005] This application provides an intelligent teaching analysis system and method based on big data to solve the problems of poor dynamic adaptability of classroom teaching strategies and low student participation in the prior art.

[0006] In a first aspect, this application provides an intelligent teaching analysis method based on big data, including: During the progress of the class, obtain the real-time physiological monitoring data and behavior monitoring data of students; Perform instant pattern recognition on the real-time physiological monitoring data and the behavior monitoring data respectively to generate a physiological state index sequence and an action feature vector set; Perform multi-modal temporal correlation analysis on the physiological state index sequence and the action feature vector set to generate a dynamic learning state map; Based on the dynamic learning state map, generate an intelligent teaching analysis result, where the intelligent teaching analysis result includes content adjustment instructions and interaction guidance strategies for the current teaching stage.

[0007] Optionally, the performing multi-modal temporal correlation analysis on the physiological state index sequence and the action feature vector set to generate a dynamic learning state map includes: Establish a cross-modal mapping relationship between the physiological state index sequence and the action feature vector set in the time dimension; Generate a dynamic learning state map according to the correlation strength between the changes in physiological state indicators and the changes in action feature vectors in the cross-modal mapping relationship.

[0008] Optionally, the generating a dynamic learning state map according to the correlation strength between the changes in physiological state indicators and the changes in action feature vectors in the cross-modal mapping relationship includes: Perform a correlation strength calculation on the amplitude change rate for reflecting the change in the physiological state indicator and the direction dimension change amount for reflecting the change in the action feature vector in the direction dimension to generate a correlation strength value; Mark the first physiological signal segment with the correlation strength value exceeding the set correlation strength threshold as an attention fluctuation node; Mark the second physiological signal segment with the change amount in the amplitude dimension exceeding the preset fluctuation range as an emotion change node, and the change amount in the amplitude dimension is used to reflect the change degree of the action feature vector in the amplitude dimension; Construct a dynamic learning state map according to the attention fluctuation node and the emotion change node.

[0009] Optionally, the constructing a dynamic learning state map according to the attention fluctuation node and the emotion change node includes: Perform an overlapping region detection on the start time of the attention fluctuation node and the duration of the emotion change node to generate a spatio-temporal interaction coefficient between the attention fluctuation node and the emotion change node; Generate a map connection line reflecting the correlation between attention fluctuation and emotion change according to the node pairs with the spatio-temporal interaction coefficient exceeding the preset coupling threshold; Count the number of interactions of the map connection line within a preset time window to generate a topological structure; Integrate the attention fluctuation node, the emotion change node, the map connection line, and the topological structure to construct a dynamic learning state map.

[0010] Optionally, the generating intelligent teaching analysis results based on the dynamic learning state map includes: Determine a mode evolution segment corresponding to the current teaching stage according to the dynamic learning state map; Dynamically match the mode evolution segment with the teaching scenario templates in the preset teaching strategy library; In the case of multiple teaching scenario templates in the matching result, generate a content adjustment instruction and an interaction guidance strategy according to the priority weights of different teaching scenario templates in the matching result; Combine the content adjustment instruction and the interactive guidance strategy to generate an intelligent teaching analysis result.

[0011] Optionally, generating the content adjustment instruction and the interactive guidance strategy according to the priority weights of different teaching scenario templates in the matching result includes: Assign a first priority weight to the teaching scenario template in the matching result where the attention fluctuation frequency exceeds the preset attention fluctuation threshold range and the emotional change amplitude is within the preset emotional fluctuation tolerance range; Assign a second priority weight to the teaching scenario template in the matching result where the emotional change amplitude exceeds the preset emotional fluctuation tolerance range and the attention fluctuation frequency is within the preset attention fluctuation threshold range; Assign a third priority weight to the teaching scenario template in the matching result where the emotional change amplitude exceeds the preset emotional fluctuation tolerance range and the attention fluctuation frequency exceeds the preset attention fluctuation threshold range; Generate the content adjustment instruction and the interactive guidance strategy according to the sorting result of the first priority weight, the second priority weight, and the third priority weight.

[0012] Optionally, respectively performing instant mode recognition on the real-time physiological monitoring data and the behavior monitoring data to generate a physiological state index sequence and an action feature vector set includes: Divide the real-time physiological monitoring data into multiple physiological signal segments according to a preset time window; Identify the physiological state indicators associated with attention fluctuations from each physiological signal segment; Arrange all the physiological state indicators in chronological order to form a physiological state index sequence; Split the limb movements captured in the behavior monitoring data into multiple action units; Identify the action feature vectors associated with emotional changes from each action unit; Combine all the action feature vectors to form an action feature vector set.

[0013] In a second aspect, the present application provides an intelligent teaching analysis system based on big data, including: An acquisition module, configured to acquire the real-time physiological monitoring data and behavior monitoring data of students during the classroom process; An identification module, configured to respectively perform instant mode recognition on the real-time physiological monitoring data and the behavior monitoring data to generate a physiological state index sequence and an action feature vector set; An analysis module, configured to perform multi-modal time series correlation analysis on the physiological state index sequence and the action feature vector set to generate a dynamic learning state map; A generation module, configured to generate an intelligent teaching analysis result based on the dynamic learning state map, where the intelligent teaching analysis result includes a content adjustment instruction and an interaction guidance strategy for the current teaching stage.

[0014] In a third aspect, the present application provides a computing device, including a processor and a memory, where a computer program is stored in the memory, and the processor is configured to run the computer program to execute any one of the methods of an intelligent teaching analysis method based on big data in the first aspect.

[0015] In a fourth aspect, the present application provides a computer storage medium, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, an intelligent teaching analysis method based on big data as described in any one of the first aspect is implemented.

[0016] In the present application, an intelligent teaching analysis method based on big data is provided. The method includes: during the classroom process, acquiring real-time physiological monitoring data and behavior monitoring data of students; respectively performing instant pattern recognition on the real-time physiological monitoring data and the behavior monitoring data to generate a physiological state index sequence and an action feature vector set; performing multimodal time series correlation analysis on the physiological state index sequence and the action feature vector set to generate a dynamic learning state map; and generating an intelligent teaching analysis result based on the dynamic learning state map, where the intelligent teaching analysis result includes a content adjustment instruction and an interaction guidance strategy for the current teaching stage.

[0017] The technical solution provided by the present application has the following beneficial effects: The present application realizes multi-dimensional data collection of students' learning states, providing a comprehensive and real-time data basis for subsequent analysis. Transforming the original data into structured features, extracting key indicators related to attention and emotion, which is convenient for subsequent correlation analysis. Through cross-modal data fusion, a dynamic model of the learning state is established to accurately reflect the attention fluctuation and emotion change trend of students. Outputting executable strategies for the current teaching stage to realize the dynamic adjustment of teaching content and the precise guidance of teacher-student interaction.

[0018] Furthermore, the present application also realizes in-depth coupled analysis of multimodal data by establishing a cross-modal mapping relationship between the physiological state index sequence and the action feature vector set in the time dimension, and generating a dynamic learning state map based on the correlation strength of the changes of the two.

[0019] Moreover, breaking through the limitations of single-modal data analysis, revealing the internal relationship between physiological indicators and behavioral characteristics through cross-modal time series correlation, making the generated learning state map more accurate and dynamic, and providing a reliable basis for optimizing teaching strategies.

[0020] These aspects or other aspects of the present application will be more clearly understood in the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0022] Figure 1 A flowchart of an intelligent teaching analysis method based on big data provided for an embodiment of the present application; Figure 2 A schematic structural diagram of an intelligent teaching analysis system based on big data provided for an embodiment of the present application; Figure 3 A schematic structural diagram of a computing device provided for an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] In order to enable those skilled in the art to better understand the solutions of the present application, the following will clearly and completely describe the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application.

[0024] In some processes described in the specification, claims and the above-mentioned drawings of the present application, a plurality of operations appear in a specific order. However, it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first" and "second" in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, nor do they limit that "first" and "second" are of different types.

[0025] The existing teaching status monitoring solutions mainly rely on the analysis of single visual behavior data, and only identify students' attention through facial expressions and body movements, which has limitations. On the one hand, it ignores the direct reflection of physiological indicators such as heart rate and blood oxygen on cognitive load and emotional state, resulting in the lack of physiological dimension support for state assessment. On the other hand, due to the lack of a dynamic association model between behavioral characteristics and physiological indicators, it is difficult to capture the co-evolution law of attention fluctuations and emotional changes, and the matching degree between the output teaching suggestions and the real learning state is insufficient, which restricts the effectiveness of personalized teaching intervention. The essence of this problem lies in the isolated processing method of multi-modal data in the existing technology, which cannot meet the technical requirements of cross-modal dynamic association analysis for accurate teaching decisions.

[0026] To address the above deficiencies, this application proposes an intelligent teaching analysis method based on big data. The core innovation is to synchronously collect physiological monitoring data and behavioral monitoring data. After generating a physiological state index sequence and an action feature vector set through instant pattern recognition, a cross-modal time-series association model is established between the two. Specifically, in the stage of constructing a dynamic learning state map, based on the spatio-temporal coupling analysis of the change rate of physiological indicators and the change amount of behavioral characteristics, the association strength between attention fluctuation nodes and emotional change nodes is quantified, and finally an intelligent decision including teaching rhythm adjustment and interaction strategy optimization is generated. This method breaks through the limitations of single-modal analysis. Through the deep spatio-temporal alignment and collaborative calculation of physiological-behavioral data, it not only solves the problem of state misjudgment caused by missing data dimensions in the existing technology, but also overcomes the decision-making lag caused by the isolated processing of multi-modal data, realizes the dynamic and accurate matching of teaching strategies and students' real learning states, and improves the timeliness of classroom interaction and the scientificity of teaching adjustment.

[0027] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application.

[0028] Figure 1 The flowchart of an intelligent teaching analysis method based on big data provided by the embodiments of the present application is as Figure 1 shown, and the method includes: Step 101: During the class, obtain the real-time physiological monitoring data and behavioral monitoring data of students.

[0029] In step 101, the physiological monitoring data represents the time-series signals reflecting the physiological state such as students' heart rate and blood oxygen saturation collected by wearable devices. The behavioral monitoring data represents the time-series data reflecting the classroom behavior characteristics such as sitting posture changes and body movement frequencies collected by video.

[0030] In the embodiments of the present application, after the class starts, the smart bracelet continuously collects the heart rate and blood oxygen data of students, forming a physiological signal stream in seconds; synchronously, the classroom camera captures the video stream of students' limb movements at a fixed frame rate. The two types of data are uploaded to the edge computing node in real time through the wireless transmission module, completing the standardization of data formats and the alignment of timestamps to ensure the temporal consistency of subsequent analysis.

[0031] For example, in a math class, the smart bracelet worn by student A collects heart rate data every 2 seconds (such as the heart rate value rising from 78 beats per minute to 92 beats per minute), and at the same time, the camera records his actions such as leaning forward and turning his head at 15 frames per second. Both types of data are transmitted to the server after being marked with a unified timestamp.

[0032] Step 102: Perform immediate pattern recognition on the real-time physiological monitoring data and the behavior monitoring data respectively to generate a physiological state index sequence and an action feature vector set.

[0033] In step 102, the physiological state index sequence represents a sequence arranged in time of attention-related indexes (such as heart rate variability) calculated by segmenting the heart rate data. The action feature vector set represents a set of vectors obtained by encoding the limb movement parameters (such as the head deflection angle) extracted from the video frames.

[0034] In the embodiments of the present application, the heart rate data is segmented by a sliding window, and the ratio of the difference between the peak and valley of each window to the duration is calculated to generate a physiological index sequence reflecting the attention concentration; for the video data, a pose estimation algorithm is used to extract joint coordinates, and the angles and amplitudes of the joint displacements between adjacent frames are calculated to form a feature vector set describing the action direction and strength. Both types of data are stored in alignment according to the time window.

[0035] For example, the heart rate data of student A within 10 minutes is segmented by a 30-second window, and the heart rate change rate of each window is calculated (such as the change rate of window 1 is 0.6 and that of window 2 is 0.9) to generate a sequence [0.6, 0.9,...]; synchronously, his head-turning action is decomposed into the change amount of the direction angle (such as turning left by 15 degrees) and the amplitude (lasting for 2 seconds), and encoded into vectors [(15, 2),...].

[0036] Step 103: Perform multi-modal temporal correlation analysis on the physiological state index sequence and the action feature vector set to generate a dynamic learning state map.

[0037] In step 103, the dynamic learning state map represents a topological network with time as the horizontal axis and nodes as state markers (such as attention fluctuation nodes), and the edges represent state transition relationships.

[0038] In the embodiments of the present application, the heart rate change rate in the physiological index sequence is aligned with the direction change amount of the action feature vector according to a time window, and the product of the two is calculated as the correlation strength; when the correlation strength exceeds the threshold, attention fluctuation nodes are marked at the corresponding time points; time periods with sudden increases in action amplitude and opposite to the heart rate change are marked as emotion change nodes; finally, weighted connection edges are generated according to the temporal relationship between the nodes to construct a graph.

[0039] For example, within the time period of 20 - 25 minutes for student A, the correlation strength between the heart rate change rate of 0.8 and the head rotation angle change amount of 20 degrees is 16 (0.8×20), which exceeds the threshold of 12, so it is marked as an attention fluctuation node; during the same period, his sudden arm-crossing action (abrupt increase in amplitude) is accompanied by a decrease in heart rate, which is marked as an emotion change node, and finally a graph containing two types of nodes and connection relationships is generated.

[0040] Step 104: Generate intelligent teaching analysis results based on the dynamic learning state graph, where the intelligent teaching analysis results include content adjustment instructions and interactive guidance strategies for the current teaching stage.

[0041] In step 104, the intelligent teaching analysis results refer to a set of executable teaching instructions generated based on the analysis conclusions of the dynamic learning state graph, specifically including two types of strategies: the content adjustment instructions are used to dynamically optimize the teaching rhythm and knowledge presentation method according to the change in students' attention concentration (such as adjusting the explanation speed, increasing the repetition times, or inserting example questions); the interactive guidance strategy designs personalized interactive solutions according to the characteristics of students' emotional fluctuations (such as reducing the difficulty of questions, changing the interactive form, or adjusting the teacher-student distance), and the two work together to achieve the precise matching of teaching strategies and students' real-time states. The current teaching stage refers to the teaching progress or time period corresponding in real time during the classroom process.

[0042] In the embodiments of the present application, dense areas of attention nodes in three consecutive time windows are extracted from the graph, and the "key reinforcement" template in the teaching strategy library is matched to generate instructions to slow down the speech speed and increase example questions; for isolated emotion nodes, the "emotion counseling" template is matched to trigger strategies such as approaching the student and reducing the difficulty of questions.

[0043] For example, for the three consecutive attention nodes of student A that appear between 20 - 30 minutes, the system prompts the teacher to "slow down the progress"; for the emotion node at 25 minutes, the "encouraging question" strategy (such as "Can you talk about the problem-solving idea?") is pushed.

[0044] This method realizes the refined perception of students' attention and emotional states through the synchronous acquisition of physiological and behavioral data and cross-modal dynamic correlation modeling, and outputs teaching strategies adapted to the classroom progress, effectively improving the accuracy and timeliness of teaching intervention, and solving the problems of misjudgment and lag caused by single-modal analysis.

[0045] In order to solve the problem of inaccurate evaluation caused by the isolated analysis of multi-modal data in the existing teaching status monitoring, in some embodiments, step 103: performing multi-modal time-series correlation analysis on the physiological state index sequence and the action feature vector set to generate a dynamic learning state map, including: Step 201: Establish a cross-modal mapping relationship between the physiological state index sequence and the action feature vector set in the time dimension.

[0046] In step 201, the cross-modal mapping relationship refers to establishing a corresponding relationship between the heart rate change trend in the physiological state index sequence and the change in the limb movement direction in the action feature vector set within the same time segment, which is used to reflect the co-variation characteristics of physiological responses and behavioral performances.

[0047] In the embodiments of the present application, first, the physiological state index sequence is divided into time segments according to a fixed duration, and the rising or falling trend of the heart rate change is extracted within each segment; synchronously, the change trend of the deflection angle of the limb movement direction in the action feature vector is extracted within the same time segment; finally, a one-to-one mapping relationship table is established after aligning the two types of trend changes in time to form the basic framework of cross-modal data association.

[0048] Step 202: Generate a dynamic learning state map according to the correlation strength between the changes in the physiological state index and the changes in the action feature vector in the cross-modal mapping relationship.

[0049] In step 202, the correlation strength refers to the coupling coefficient obtained by quantifying the matching degree between the change amplitude of the physiological index and the change direction of the action feature, which is used to judge the change nodes of the student's attention or emotional state.

[0050] In the embodiments of the present application, based on the established cross-modal mapping relationship table, the product value of the heart rate change amplitude and the limb movement direction change angle within each time segment is calculated as the initial correlation strength; the correlation strength of consecutive time segments is smoothed, and the peak intervals exceeding the preset threshold are screened out; the attention fluctuation nodes are marked within these intervals, and at the same time, the time segments with a sudden increase in the limb movement amplitude and opposite to the heart rate change direction are marked as emotional change nodes; finally, all the nodes are connected in chronological order to construct a dynamic learning state map.

[0051] The following is a specific example: In a math class scenario, the smart bracelet worn by student A collects heart rate data every 2 seconds. The heart rate value rises from 78 beats per minute to 92 beats per minute. At the same time, the classroom camera records the student's forward body lean and head-turning movements at a frequency of 15 frames per second. After the two types of data are synchronously transmitted to the server through a unified timestamp, the system divides the heart rate data within 10 minutes into 30-second windows, and calculates the heart rate change rate within each window using the formula: heart rate change rate equals (the maximum heart rate value in the window minus the minimum heart rate value) divided by the time window length. For example, in the first window, the measured heart rate rises from 80 beats per minute to 92 beats per minute, and the change rate is (92 - 80) divided by 30, which equals 0.4. In the second window, the change rate is 0.6, generating a sequence of 0.4, 0.6. When synchronously processing the behavior data, the head-turning movement is decomposed into a direction angle change of 15 degrees to the left and a duration of 2 seconds, and encoded as a vector [15, 2]. When analyzing the time period from 20 to 25 minutes, the system calculates the correlation intensity value between the heart rate change rate of 0.8 and the head-turning angle change of 20 degrees in this period, which is 0.8 multiplied by 20, equal to 16. Since it exceeds the preset threshold of 12, it is marked as an attention fluctuation node. At the same time, it is detected that the student suddenly makes a crossed-arm movement and the movement amplitude reaches 1.5 times the preset threshold, accompanied by a 10-beat-per-minute decrease in heart rate, and the system marks it as an emotion change node. Finally, a dynamic learning state map containing these two types of nodes and their time correlation relationships is generated.

[0052] In the embodiment of the present application, by establishing a cross-modal dynamic association model for physiological and behavioral data, a three-dimensional evaluation of the student's learning state is realized, enabling the generated teaching strategy to consider both attention persistence and emotion change points, and improving the accuracy and timeliness of classroom intervention.

[0053] To further improve the accuracy of learning state analysis, in some embodiments, step 202: generating a dynamic learning state map according to the correlation intensity between the change of physiological state indicators and the change of action feature vectors in the cross-modal mapping relationship includes: Step 301: Calculate the correlation intensity between the amplitude change rate reflecting the change of the physiological state indicator and the direction dimension change amount reflecting the change of the action feature vector in the direction dimension to generate a correlation intensity value.

[0054] In step 301, the rate of amplitude change refers to calculating the difference between adjacent peaks and troughs of the heart rate data in the physiological signal segment, and dividing it by the time length of the signal segment (in seconds) to obtain the rate of change of the heart rate fluctuation amplitude per unit time. The calculation formula is: rate of amplitude change = (current peak heart rate value - previous trough heart rate value) / peak-trough time interval. The change amount in the direction dimension refers to comparing the difference in the spatial orientation angles of consecutive action units in the action feature vector set, taking the absolute value and dividing it by the action duration (in seconds) to obtain the change amount of the limb movement direction per unit time. The calculation formula is: change amount in the direction dimension = |current action unit orientation angle - previous action unit orientation angle| / action duration. The association strength value is the product of the rate of amplitude change and the change amount in the direction dimension, reflecting the degree of coordination between the physiological response and the behavior change.

[0055] In the embodiment of the present application, first, the heart rate fluctuation amplitude of each time segment is extracted from the physiological state index sequence, and the ratio of it to the time window length is calculated to obtain the rate of amplitude change; at the same time, the difference in the movement direction angles of adjacent action units is extracted from the action feature vector set as the change amount in the direction dimension; finally, the rate of amplitude change and the change amount in the direction dimension within the same time segment are multiplied to generate the association strength value representing the degree of association between the two. Exemplarily, in a math class scenario, when it is detected that student A is in the time period from 10:00 to 10:05: the heart rate amplitude change rate is 0.8 beats per second (calculation process: (95 beats per minute - 79 beats per minute) ÷ 20 seconds = 0.8); the change amount in the head rotation direction is 15 degrees per second (calculation process: |45 degrees - 30 degrees| ÷ 1 second = 15); the association strength value = rate of amplitude change × change amount in the direction dimension × time window weight coefficient (0.5) = 0.8 × 15 × 0.5 = 6.0 (when the association strength value is greater than the preset threshold of 5, an attention fluctuation node is generated).

[0056] Step 302: Mark the first physiological signal segment whose association strength value exceeds the set association strength threshold as an attention fluctuation node.

[0057] In step 302, the first physiological signal segment refers to the data segment containing heart rate changes and blood oxygen saturation in the physiological state index sequence after dividing the real-time physiological monitoring data according to a preset time window, and after calculating the association strength, its association strength value exceeds the set association strength threshold. The attention fluctuation node is a special node in the graph that marks these time segments and is used to indicate abnormal changes in the student's attention state.

[0058] In an embodiment of the present application, a sliding window detection is performed on the calculated sequence of association strength values. When the association strength values of three consecutive time segments all exceed the threshold, this time period is marked as the first physiological signal segment, and an attention fluctuation node is created at the corresponding time position in the dynamic learning state map. The size of the node is proportional to the association strength value.

[0059] Step 303: Mark the second physiological signal segment with a change in the amplitude dimension exceeding the preset fluctuation range as an emotion change node. The change in the amplitude dimension is used to reflect the degree of change of the action feature vector in the amplitude dimension.

[0060] In step 303, the second physiological signal segment refers to a data segment in the physiological state index sequence that, after dividing the real-time physiological monitoring data according to a preset time window, contains data such as heart rate changes and blood oxygen saturation, and the change in the amplitude dimension of the corresponding action feature vector exceeds the preset fluctuation range. The emotion change node is a node in the map that marks these special time periods and is used to reflect the sudden change in the student's emotion.

[0061] In an embodiment of the present application, the change in the amplitude dimension of the action feature vector is monitored. When it is detected that the amplitude value suddenly increases and exceeds twice the average fluctuation range, combined with the characteristic of the heart rate decrease during this time period, the corresponding time segment is marked as the second physiological signal segment, and an emotion change node is created in the map. The depth of the node color reflects the intensity of the emotion fluctuation.

[0062] Step 304: Construct a dynamic learning state map according to the attention fluctuation node and the emotion change node.

[0063] In an embodiment of the present application, the marked attention fluctuation nodes and emotion change nodes are arranged in chronological order, the time interval and attribute similarity between adjacent nodes are calculated to generate weighted connection edges, and finally a dynamic learning state map containing a complete state evolution path is constructed.

[0064] The following is a specific example: In a math class scenario, student A's smart bracelet continuously monitors that their heart rate fluctuates from 82 beats per minute to 98 beats per minute during the 20 - 25 minute period. The system uses a 30 - second time window to calculate the heart rate change rate. The specific formula is that the heart rate change rate equals the maximum heart rate in the window minus the minimum heart rate divided by the window duration. The measured heart rate change rate during this period is (98 - 82) / 30 = 0.53. At the same time, the camera captures an increased head - turning frequency of the student during the same period, and through analysis and calculation, the average change in the direction dimension is 22 degrees every 30 seconds. The system multiplies the heart rate change rate of 0.53 by the change in the direction dimension of 22 to obtain an association strength value of 11.66. When this value exceeds the preset threshold of 10, the system automatically marks this period as an attention - fluctuation node. Subsequently, at 26 minutes, the system detects that the student suddenly makes a hugging - the - arms movement with an amplitude of 35 cm, which exceeds 1.75 times the baseline amplitude of 20 cm. At the same time, the heart rate drops suddenly from 95 beats per minute to 83 beats per minute, meeting the conditions that the change in the amplitude dimension exceeds 1.5 times the preset fluctuation range and the heart rate drops by more than 10 beats per minute. Therefore, it is marked as an emotion - change node. Finally, the system connects these nodes in chronological order to construct a dynamic learning - state map that shows the attention fluctuation preceding the emotion change. The distance between nodes reflects the state - transition speed, and the thickness of the connection lines represents the transition strength, providing an intuitive visual analysis basis for the teacher on the state evolution.

[0065] In the embodiment of the present application, by quantifying the dynamic association characteristics of physiological and behavioral data, the refined capture of the learning - state change is realized. The constructed map can intuitively display the complete process of the student's transition from attention fluctuation to emotion change, providing accurate state - evolution clues and a basis for judging the intervention timing for the teacher.

[0066] To further improve the accuracy and interpretability of the learning - state map, in some embodiments, step 304: constructing a dynamic learning - state map according to the attention - fluctuation node and the emotion - change node includes: Step 401: Detect the overlapping area between the start time of the attention - fluctuation node and the duration of the emotion - change node to generate a spatio - temporal interaction coefficient between the attention - fluctuation node and the emotion - change node.

[0067] In step 401, the starting time of the attention fluctuation node refers to the specific moment when the student's attention begins to fluctuate, which is identified through the calculation of the association strength. This moment marks the critical point at which the change rate of the physiological index and the change amount of the behavioral characteristics first reach the preset association threshold. The duration of the emotion change node refers to the total time length from the moment when the change amount of the amplitude dimension of the action feature vector first exceeds the preset fluctuation range to the moment when this change amount falls back to the normal range, reflecting the continuous span of the abnormal emotional state of the student. The spatio-temporal interaction coefficient refers to the weighted value of the overlapping part of the time interval of the attention fluctuation node and the time interval of the emotion change node, which is used to quantify the association degree between the two types of nodes in the spatio-temporal dimension. When calculating, two factors, namely the overlapping duration and the node strength, are considered.

[0068] In the embodiment of the present application, first, the starting time and duration of each attention fluctuation node, as well as the occurrence time and duration of the emotion change node, are obtained, and the overlapping duration of the two time intervals is calculated; then, the overlapping duration is multiplied by the strength values of the two nodes to obtain the interaction coefficient reflecting the tightness of the spatio-temporal association. The larger this coefficient is, the stronger the association between the two types of nodes.

[0069] Step 402: Generate a graph connection line reflecting the association between attention fluctuation and emotion change according to the node pairs whose spatio-temporal interaction coefficient exceeds the preset coupling threshold.

[0070] In step 402, the graph connection line is a directed edge connecting different types of nodes in the dynamic learning state graph. The direction of the edge represents the state evolution path from attention fluctuation to emotion change, and the thickness of the edge reflects the association strength between the two types of nodes.

[0071] In the embodiment of the present application, the node pairs whose spatio-temporal interaction coefficient exceeds the preset threshold are screened out, and the initial weight of the connection line is determined according to the coefficient size; then, in chronological order, a directed connection line is generated from the attention fluctuation node to the emotion change node. The thickness of the connection line is proportional to the spatio-temporal interaction coefficient.

[0072] Step 403: Count the number of interactions of the graph connection line within a preset time window to generate a topological structure.

[0073] In step 403, the topological structure refers to a network relationship graph composed of multiple nodes and connection lines. The dense connection area in the structure represents the critical period when the learning state fluctuates frequently, and the sparse area represents the stable state period.

[0074] In the embodiments of the present application, the interaction times of all connection lines within a preset time window are counted, and clustering analysis is performed on the areas with high interaction times to identify hotspots of state transitions; according to the distribution characteristics of the hotspots and the connection line density, a topological structure framework reflecting the overall state evolution law is generated. Exemplarily, in a 45-minute classroom scenario, a 5-minute analysis time window is set. When it is detected that within the 10-15 minute window for a certain student, there are 3 overlaps between the attention fluctuation nodes and the emotion change nodes (the calculation method for the number of overlaps is: when there are both attention fluctuation nodes and emotion change nodes within the same time segment, it is counted as 1 overlap), and the spatio-temporal interaction coefficient for each overlap exceeds a preset coupling threshold of 0.7 (this threshold is obtained through statistical analysis of historical classroom data, and the calculation formula is: coupling threshold = effective interaction times / total interaction times × 0.9, where effective interaction times refer to node interaction events that actually affect teaching strategies), then a graph connection line with a connection strength of 3 is generated within this time window; when the connection strength for 3 consecutive time windows (i.e., 15 minutes) is greater than 2, an evolution path from the "focused state" to the "emotion fluctuation state" is generated, and the weight value of this path is 2.3 (weight value = Σ window connection strength / number of windows × 0.8, where 0.8 is the classroom scenario correction coefficient). Finally, the topological structure containing this evolution path is integrated into the dynamic learning state graph.

[0075] Step 404: Integrate the attention fluctuation nodes, the emotion change nodes, the graph connection lines, and the topological structure to construct a dynamic learning state graph.

[0076] In the embodiments of the present application, the marked attention fluctuation nodes and emotion change nodes are arranged on the time axis in chronological order, and the vertical coordinate position is determined according to the node strength; then the connection lines between the nodes are drawn according to the connection relationship of the topological structure; finally, all elements are integrated to generate a complete dynamic learning state graph, where the size of the nodes represents the state fluctuation strength, and the depth of the color of the connection lines reflects the transition speed.

[0077] The following is a specific example: In a math class scenario, student A was identified by the system as having an attention fluctuation node during the period from 20 minutes and 30 seconds to 23 minutes and 45 seconds. The start time of this node was 20 minutes and 30 seconds, and the duration was 3 minutes and 15 seconds. Subsequently, an emotion change node was detected during the period from 24 minutes and 10 seconds to 26 minutes and 20 seconds, with a duration of 2 minutes and 10 seconds. The system calculated the overlapping period of the two nodes as 1 minute and 35 seconds from 24 minutes and 10 seconds to 23 minutes and 45 seconds. The formula for the spatio-temporal interaction coefficient is the overlapping duration multiplied by the average node intensity. Taking the average of the attention node intensity value of 12 and the emotion node intensity value of 8, which is 10, the interaction coefficient is 1.58 minutes multiplied by 10, equal to 15.8. Since this value exceeds the preset coupling threshold of 12, the system generates a graph connection line from the attention node to the emotion node, and the line width is set to 15.8 divided by the reference value of 5, equal to 3.16 units. During a 30-minute teaching period, a total of 4 interaction times of similar node pairs were counted, forming a dense topological structure area centered around 25 minutes. During the final integration, the attention node was marked with a blue circle at the 20 minutes and 30 seconds position, the emotion node was marked with a red triangle at the 24 minutes and 10 seconds position, the two nodes were connected by an orange connection line with a width of 3.16 units, and a topological heat map reflecting the four state transitions was generated on the right side of the graph, fully demonstrating the dynamic evolution process of the student from attention dispersion to emotional depression. The heat value of the topological structure is proportional to the number of interaction times, providing a clear reminder for the teacher to "focus on observation around 25 minutes".

[0078] In the embodiments of the present application, the relevance of nodes is accurately quantified through the spatio-temporal interaction coefficient, and the law of state evolution is revealed by using the topological structure. The constructed dynamic learning state graph can not only intuitively display the details of the state changes of individual students, but also reflect the typical state transition patterns in classroom teaching, providing a reliable basis for teachers to implement precise teaching interventions.

[0079] To further improve the accuracy and practicality of teaching strategy generation, in some embodiments, step 104: generating an intelligent teaching analysis result based on the dynamic learning state graph, including: Step 501: Determine the mode evolution segment corresponding to the current teaching stage according to the dynamic learning state graph.

[0080] In step 501, the mode evolution segment refers to a typical state transition period extracted from the dynamic learning state graph, which contains continuous attention fluctuation nodes and emotion change nodes, and is used to reflect the learning state change law of students in a specific teaching stage.

[0081] In the embodiments of the present application, first, analyze the temporal distribution characteristics of the nodes in the graph spectrum to identify consecutive time periods when the nodes appear densely; then extract all the nodes and their connection relationships within these time periods to form a pattern fragment reflecting the complete state evolution process; finally, according to the current teaching progress, select the pattern evolution fragment closest in time as the analysis basis.

[0082] Step 502: Dynamically match the pattern evolution fragment with the teaching scenario templates in the preset teaching strategy library.

[0083] In step 502, the teaching scenario template is a description of a typical teaching situation stored in the preset strategy database, including the corresponding relationships between attention fluctuation characteristics, emotion change characteristics, and recommended strategies. Dynamic matching refers to the process of calculating the similarity between the pattern features extracted in real time and the features in the template library.

[0084] In the embodiments of the present application, extract feature parameters such as node type, intensity, and connection line density from the pattern evolution fragment; compare these parameters item by item with the feature dimensions in the teaching scenario template; calculate the matching scores of each template, and screen out a set of candidate templates with scores higher than the threshold.

[0085] Step 503: In the case where the matching results are multiple teaching scenario templates, generate content adjustment instructions and interactive guidance strategies according to the priority weights of different teaching scenario templates in the matching results.

[0086] In step 503, the priority weight is a strategy recommendation index comprehensively calculated based on the matching score and the importance of the teaching stage, and is used to determine the order of strategy execution in multi-template matching.

[0087] In the embodiments of the present application, calculate the product of the matching score of each candidate template and the weight of the current teaching stage to obtain the priority weight; sort them in descending order of the weight value, and select the top three templates with the highest weights; respectively extract the content adjustment methods and interactive strategies recommended by these templates, and perform fusion and optimization.

[0088] Step 504: Combine the content adjustment instructions and the interactive guidance strategies to generate an intelligent teaching analysis result.

[0089] In the embodiments of the present application, arrange and combine the selected content adjustment instructions in the logical order of knowledge points; at the same time, classify the interactive guidance strategies according to the implementation difficulty and expected effects; finally, generate a complete teaching analysis report including specific implementation steps and time arrangements.

[0090] The following is a specific example: In a mathematics classroom scenario, the system extracts a pattern evolution segment for the 20 - 30 minute teaching period from the dynamic learning state map of student A. This segment contains 3 consecutive attention fluctuation nodes and 1 emotion change node. The average correlation intensity value of the attention nodes is 15, and the amplitude change of the emotion node is 1.8 times the reference value. When matching this segment feature with the teaching strategy library, the calculated matching degree with the "difficulty in understanding knowledge points" template is 80 points (the matching degree calculation formula is the number of nodes multiplied by the average intensity coefficient 15 divided by 10 equals 45, plus the amplitude change coefficient 1.8 multiplied by 20 equals 36, and the total score of 81 points is rounded), and the matching degree with the "low classroom participation" template is 75 points. The system sets the current teaching stage as "new knowledge explanation" and the importance coefficient as 1.1. The calculated priority weights are 80 multiplied by 1.1 equals 88 and 75 multiplied by 1.1 equals 82.5 respectively. Select the "difficulty in understanding knowledge points" template with the highest weight, generate a content adjustment instruction of "increase the number of example demonstrations" and an interactive guidance strategy of "adopt heuristic questioning", and finally combine them into an intelligent teaching analysis result of "first supplement 2 transition examples and then ask 'how did you think about this problem-solving step'".

[0091] In the embodiment of the present application, through the precise extraction of the pattern evolution segment and multi-dimensional strategy matching, the dynamic adaptation of teaching strategies to the actual state of students is achieved. The generated intelligent analysis results not only consider the systematicness of knowledge transfer but also take into account the pertinence of teacher-student interaction, effectively improving the flexibility and effectiveness of classroom teaching.

[0092] To further improve the accuracy and applicability of teaching strategy generation, in some embodiments, step 503: generating a content adjustment instruction and an interactive guidance strategy according to the priority weights of different teaching scenario templates in the matching result includes: Step 601: Assign a first priority weight to the teaching scenario template in the matching result where the attention fluctuation frequency exceeds the preset attention fluctuation threshold range and the emotion change amplitude is within the preset emotion fluctuation tolerance range.

[0093] In step 601, the first priority weight refers to the weight value of the teaching strategy that is preferentially adopted when the student's attention fluctuates but the emotion is relatively stable, and this weight focuses on solving the problem of knowledge understanding.

[0094] In the embodiment of the present application, first, count the number of attention fluctuation nodes in the pattern evolution segment and calculate the fluctuation frequency per unit time; then detect whether the amplitude change of the emotion change node is within the preset normal fluctuation range; finally, assign a priority weight calculated based on the fluctuation frequency value to the teaching template that simultaneously meets the conditions of excessive attention frequency and normal emotion amplitude.

[0095] Step 602: Assign a second-priority weight to the teaching scenario templates in the matching results where the amplitude of emotional change exceeds the preset emotional fluctuation tolerance range and the attention fluctuation frequency is within the preset attention fluctuation threshold range.

[0096] In step 602, the second-priority weight refers to the weight value of the teaching strategy that is preferentially adopted when the student's emotions fluctuate but the attention is relatively stable. This weight focuses on solving the psychological state problem.

[0097] In the embodiment of the present application, it is detected whether the amplitude of the emotional change node exceeds the upper tolerance limit; at the same time, it is verified whether the attention fluctuation frequency is within the normal threshold; for the teaching templates that meet this condition, a priority weight calculated based on the amplitude of the emotional change is assigned, and the greater the amplitude, the higher the weight.

[0098] Step 603: Assign a third-priority weight to the teaching scenario templates in the matching results where the amplitude of emotional change exceeds the preset emotional fluctuation tolerance range and the attention fluctuation frequency exceeds the preset attention fluctuation threshold range.

[0099] In step 603, the third-priority weight refers to the weight value of the teaching strategy adopted when the student has both attention fluctuations and emotional fluctuations. This weight needs to take into account both cognitive and emotional factors.

[0100] In the embodiment of the present application, the abnormal periods that simultaneously exceed the attention frequency threshold and the emotional amplitude tolerance are identified; the weighted sum of the two abnormal indicators is calculated; and a priority weight that comprehensively considers the two abnormal degrees is assigned to the matching composite teaching templates.

[0101] Step 604: Generate a content adjustment instruction and an interaction guidance strategy according to the sorting results of the first-priority weight, the second-priority weight, and the third-priority weight.

[0102] In the embodiment of the present application, the teaching templates are selected in the order of the first, second, and third priorities; the content adjustment methods and interaction strategies recommended by each template are extracted; and the strategies are integrated according to the characteristics of the current teaching stage to generate an executable instruction combination.

[0103] The following is a specific example: In a mathematics classroom scenario, when the system analyzes the pattern evolution segment of student A in the 20 - 30 minute period, it detects that the attention fluctuation frequency of this student reaches 0.7 times per minute, exceeding the preset threshold of 0.5 times. The amplitude of emotional change is 1.8 times the baseline value and is within the preset tolerance range of 1.5 - 2.0 times, meeting the conditions for assigning the first - priority weight. According to the formula, the first - priority weight is equal to the matching degree multiplied by the teaching - stage coefficient. Taking the matching degree of the difficult - to - understand knowledge - point template as 80 points and the new - knowledge - explanation stage coefficient as 1.1, the obtained weight is 88 points. At the same time, it is found that the amplitude of emotional change of this student suddenly increases to 2.3 times the baseline value in the 25 - 28 minute period, exceeding the upper limit of the tolerance of 2.0 times. However, the attention fluctuation frequency of 0.3 times is within the normal range of less than 0.5 times, meeting the second - priority weight conditions. Taking the matching degree of the low - classroom - participation template as 75 points and multiplying it by 1.1, the obtained weight is 82.5 points. The system preferentially selects the difficult - to - understand knowledge - point template with a weight of 88 points, generates a content - adjustment instruction to increase 2 transitional examples. The specific number of examples is obtained by dividing the attention - fluctuation duration of 3 minutes by the preset baseline interval of 1.5 minutes. A heuristic - questioning interaction strategy is adopted, and the questioning interval is set to 5 minutes (the time required for the emotion to subside) divided by the preset questioning - density coefficient of 2, which is equal to 2.5 minutes.

[0104] In the embodiments of the present application, the core problem types of students are accurately identified through a hierarchical weight mechanism, enabling the generated teaching strategies to not only target and solve the main contradictions but also take into account secondary problems, achieving an accurate transformation from state recognition to strategy implementation, and effectively improving the timeliness and effectiveness of teaching intervention.

[0105] To further improve the accuracy and timeliness of physiological and behavioral data analysis, in some embodiments, step 102: The instant pattern recognition of the real - time physiological monitoring data and the behavioral monitoring data respectively to generate a physiological - state index sequence and an action - feature vector set includes: Step 701: Divide the real - time physiological monitoring data into multiple physiological signal segments according to a preset time window.

[0106] In step 701, a physiological signal segment refers to a time - period data unit obtained by dividing continuous physiological monitoring data according to a fixed duration, and each segment contains the complete heart - rate fluctuation - cycle characteristics.

[0107] In the embodiments of the present application, first, a time - window length suitable for the classroom scenario is set, then the real - time collected physiological data is equally divided according to this window to ensure that each segment can reflect a complete state - change cycle, and finally, quality verification is performed on the divided segments to eliminate invalid segments with incomplete signals.

[0108] Step 702: Identify physiological - state indicators associated with attention fluctuations from each physiological signal segment.

[0109] In step 702, the physiological state indicators associated with attention fluctuations refer to the quantitative features extracted from physiological parameters such as heart rate variability and blood oxygen saturation changes, which can characterize the changes in the degree of attention concentration.

[0110] In the embodiments of the present application, waveform feature analysis is performed on each valid physiological signal segment, the rhythm change pattern therein is detected, the characteristic parameters that can reflect the activity of the autonomic nervous system are calculated, and the corresponding relationship between these parameters and the attention state is established to form a standardized state indicator.

[0111] Step 703: Arrange all the physiological state indicators in chronological order to form a physiological state indicator sequence.

[0112] In step 703, chronological order and the preset time window are not the same concept: the preset time window is a fixed interval for data segmentation, and chronological order is the sequence relationship of the segmented segments arranged in sequence.

[0113] In the embodiments of the present application, the attention-related indicators calculated for each time window are arranged in the order of acquisition time, and the timestamp information is retained at the same time to form a time series data sequence that can reflect the continuous changes in the attention state.

[0114] Step 704: Split the limb movements captured in the behavior monitoring data into multiple action units.

[0115] In step 704, an action unit refers to a limb movement segment with complete semantics segmented from continuous behavior data, and each unit contains a complete action process of start, duration, and end. The structure of the action unit contains the coding information of the direction dimension and the amplitude dimension, which is used to describe the spatial change characteristics of the limb movement.

[0116] In the embodiments of the present application, by analyzing the motion trajectory and posture change characteristics in the behavior monitoring data, the key frames of the start and end of the action are identified, the continuous behavior stream is segmented into independent units with clear action intentions, and the action type and spatio-temporal characteristics are marked for each unit.

[0117] Step 705: Identify the action feature vectors associated with emotional changes from each action unit.

[0118] In step 705, the action feature vectors associated with emotional changes refer to the combination of motion features extracted from the action units, which can reflect the changes in the emotional state.

[0119] In the embodiments of the present application, the basic features such as the motion amplitude, speed, and direction of each action unit are analyzed, the statistics and change patterns of these features are calculated, and the feature combinations with stable correlations with typical emotional states are screened out and encoded into multi-dimensional feature vectors.

[0120] Step 706: Combine all the action feature vectors into an action feature vector set.

[0121] In the embodiment of the present application, the emotion feature vectors obtained by analyzing each action unit are arranged in the time sequence of action occurrence, while retaining the time correspondence with the physiological data, to form a feature set for subsequent multimodal analysis.

[0122] The following is a specific example: In a mathematics classroom scenario, the system divides the real-time physiological monitoring data of student A into segments according to a preset time window of 30 seconds, and obtains 10 physiological signal segments in the time period of 20 - 25 minutes. Calculate the heart rate change rate for each segment. The formula for the heart rate change rate is equal to the maximum heart rate in the window minus the minimum heart rate divided by the length of the time window. For example, in the first window, the heart rate is measured to rise from 82 beats per minute to 98 beats per minute, and the change rate is (98 - 82) / 30 = 0.53. The change rate of the second window is 0.67, generating a sequence of 0.53, 0.67. When synchronously processing the behavior data, the 20 - 25 minute video stream is decomposed into 32 action units, including 15 head-turning actions and 5 body-leaning-forward actions. Calculate the change amount of the direction angle for each head-turning action. For example, an action of turning left by 20 degrees for 3 seconds is encoded as a vector 20, 3. Calculate the change amount of the forward inclination angle for the body-leaning-forward action. For example, an action of leaning forward by 30 degrees for 2 seconds is encoded as a vector 30, 2. In particular, at 24 minutes and 30 seconds, a hugging-arm action is detected, and its sudden increase in amplitude is calculated to be 2.1 times the reference value and lasts for 4 seconds, which is encoded as a special emotion feature vector 2.1, 4. Finally, the system generates a physiological state index sequence containing 10 heart rate change rate indicators and an action feature vector set of 32 action feature vectors, providing a data basis for subsequent construction of a dynamic learning state map.

[0123] In the embodiment of the present application, through a standardized data segmentation and feature extraction process, the unified representation of multi-source heterogeneous data is realized, providing a reliable feature basis for accurately identifying the learning state of students, and at the same time ensuring the real-time performance and interpretability of data analysis.

[0124] Figure 2 FIG. is a schematic structural diagram of an intelligent teaching analysis system based on big data provided by an embodiment of the present application. As Figure 2 shown, the system includes: An acquisition module 21, configured to acquire real-time physiological monitoring data and behavior monitoring data of students during the classroom process.

[0125] An identification module 22, configured to perform real-time pattern recognition on the real-time physiological monitoring data and the behavior monitoring data respectively, and generate a physiological state index sequence and an action feature vector set.

[0126] An analysis module 23 for performing multimodal time-series correlation analysis on the physiological state index sequence and the action feature vector set to generate a dynamic learning state map.

[0127] A generation module 24 for generating an intelligent teaching analysis result based on the dynamic learning state map, where the intelligent teaching analysis result includes a content adjustment instruction and an interaction guidance strategy for the current teaching stage.

[0128] Figure 2 The described intelligent teaching analysis system based on big data can execute Figure 1 The described intelligent teaching analysis method based on big data in the illustrated embodiment, and its implementation principle and technical effects will not be elaborated further. For the intelligent teaching analysis system based on big data in the above embodiment, the specific ways in which each module and unit perform operations have been described in detail in the embodiment related to the method, and will not be elaborated here.

[0129] In a possible design, Figure 2 The intelligent teaching analysis system based on big data in the illustrated embodiment can be implemented as a computing device, such as Figure 3 shown, and the computing device can include a storage component 31 and a processing component 32; The storage component 31 stores one or more computer instructions, where the one or more computer instructions are called and executed by the processing component 32.

[0130] The processing component 32 is used for the Figure 1 intelligent teaching analysis method based on big data in the above

[0131] embodiment. Among them, the processing component 32 can include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component can also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components for executing the above method.

[0132] The storage component 31 is configured to store various types of data to support the operations of the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable programmable read only memory (EPROM), programmable read only memory (PROM), read only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disc.

[0133] Of course, the computing device may also necessarily include other components, such as input / output interfaces, display components, communication components, etc.

[0134] The input / output interface provides an interface between the processing component and the peripheral interface module, and the above-mentioned peripheral interface module can be an output device, an input device, etc.

[0135] The communication component is configured to facilitate communication between the computing device and other devices in a wired or wireless manner, etc.

[0136] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. At this time, the computing device can refer to a cloud server, and the above-mentioned processing component, storage component, etc. can be basic server resources leased or purchased from the cloud computing platform.

[0137] The embodiment of the present application also provides a computer storage medium storing a computer program, and when the computer program is executed by the computer, it can implement the above Figure 1 shown embodiment of an intelligent teaching analysis method based on big data.

[0138] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0139] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative effort.

[0140] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0141] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present application.

Claims

1. An intelligent teaching analysis method based on big data, characterized in that, Including: During the class, obtaining the real-time physiological monitoring data and behavior monitoring data of students; Performing immediate pattern recognition on the real-time physiological monitoring data and the behavior monitoring data respectively to generate a physiological state index sequence and an action feature vector set; Performing multimodal time-series correlation analysis on the physiological state index sequence and the action feature vector set to generate a dynamic learning state map; Based on the dynamic learning state map, generating an intelligent teaching analysis result, where the intelligent teaching analysis result includes a content adjustment instruction and an interaction guiding strategy for the current teaching stage.

2. The method according to claim 1, wherein The performing multimodal time-series correlation analysis on the physiological state index sequence and the action feature vector set to generate a dynamic learning state map includes: Establishing a cross-modal mapping relationship between the physiological state index sequence and the action feature vector set in the time dimension; Generating a dynamic learning state map according to the correlation strength between the change of the physiological state index and the change of the action feature vector in the cross-modal mapping relationship.

3. The method according to claim 2, wherein The generating a dynamic learning state map according to the correlation strength between the change of the physiological state index and the change of the action feature vector in the cross-modal mapping relationship includes: Performing correlation strength calculation on the amplitude change rate for reflecting the change of the physiological state index and the direction dimension change amount for reflecting the change of the action feature vector in the direction dimension to generate a correlation strength value; Marking the first physiological signal segment with the correlation strength value exceeding the set correlation strength threshold as an attention fluctuation node; Marking the second physiological signal segment with the change amount of the amplitude dimension exceeding the preset fluctuation range as an emotion change node, where the change amount of the amplitude dimension is used to reflect the change degree of the action feature vector in the amplitude dimension; Constructing a dynamic learning state map according to the attention fluctuation node and the emotion change node.

4. The method according to claim 3, wherein The constructing a dynamic learning state map according to the attention fluctuation node and the emotion change node includes: Performing overlapping region detection on the start time of the attention fluctuation node and the duration of the emotion change node to generate a spatio-temporal interaction coefficient between the attention fluctuation node and the emotion change node; Generating a map connection line reflecting the correlation between attention fluctuation and emotion change according to the node pairs with the spatio-temporal interaction coefficient exceeding the preset coupling threshold; Counting the interaction times of the map connection line within a preset time window to generate a topological structure; Integrating the attention fluctuation node, the emotion change node, the map connection line, and the topological structure to construct a dynamic learning state map.

5. The method according to claim 1, wherein The generating an intelligent teaching analysis result based on the dynamic learning state map includes: Determining a mode evolution segment corresponding to the current teaching stage according to the dynamic learning state map; Performing dynamic matching on the mode evolution segment and the teaching scenario templates in the preset teaching strategy library; In the case of multiple teaching scenario templates in the matching result, generating a content adjustment instruction and an interaction guiding strategy according to the priority weights of different teaching scenario templates in the matching result; Combining the content adjustment instruction and the interaction guiding strategy to generate an intelligent teaching analysis result.

6. The method according to claim 5, characterized in that, Generating a content adjustment instruction and an interaction guidance strategy according to the priority weights of different teaching scenario templates in the matching result, including: Assigning a first priority weight to a teaching scenario template in the matching result where the attention fluctuation frequency exceeds the preset attention fluctuation threshold range and the emotional change amplitude is within the preset emotional fluctuation tolerance range; Assigning a second priority weight to a teaching scenario template in the matching result where the emotional change amplitude exceeds the preset emotional fluctuation tolerance range and the attention fluctuation frequency is within the preset attention fluctuation threshold range; Assigning a third priority weight to a teaching scenario template in the matching result where the emotional change amplitude exceeds the preset emotional fluctuation tolerance range and the attention fluctuation frequency exceeds the preset attention fluctuation threshold range; Generating a content adjustment instruction and an interaction guidance strategy according to the sorting result of the first priority weight, the second priority weight, and the third priority weight.

7. The method according to claim 1, wherein Performing immediate pattern recognition on the real-time physiological monitoring data and the behavior monitoring data respectively to generate a physiological state index sequence and an action feature vector set, including: Dividing the real-time physiological monitoring data into multiple physiological signal segments according to a preset time window; Identifying physiological state indicators associated with attention fluctuations from each physiological signal segment; Arranging all the physiological state indicators in chronological order to form a physiological state index sequence; Splitting the limb movements captured in the behavior monitoring data into multiple action units; Identifying action feature vectors associated with emotional changes from each action unit; Combining all the action feature vectors into an action feature vector set.

8. An intelligent teaching analysis system based on big data, characterized in that, Including: An acquisition module for acquiring the real-time physiological monitoring data and behavior monitoring data of students during the classroom process; An identification module for performing immediate pattern recognition on the real-time physiological monitoring data and the behavior monitoring data respectively to generate a physiological state index sequence and an action feature vector set; An analysis module for performing multi-modal time series correlation analysis on the physiological state index sequence and the action feature vector set to generate a dynamic learning state map; A generation module for generating an intelligent teaching analysis result based on the dynamic learning state map, where the intelligent teaching analysis result includes a content adjustment instruction and an interaction guidance strategy for the current teaching stage.

9. A computing device, characterized in that, Including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a big data-based intelligent teaching analysis method according to any one of claims 1 to 7.

10. A computer storage medium, characterized in that, A computer program is stored, and when the computer program is executed by a computer, it implements a big data-based intelligent teaching analysis method according to any one of claims 1 to 7.

Images