Geography teaching device and interaction system based on holographic projection

Abstract

The invention provides a holographic projection-based geography teaching device and interaction system, and the device comprises a data collection module which is used for obtaining interaction data, carrying out the timestamp marking and sensor data fusion processing of the interaction data, and generating a multidimensional behavior data set containing a timestamp; the behavior analysis module is used for performing view preference feature extraction and operation habit analysis processing on the multi-dimensional behavior data set to generate a student learning deviation degree; the dynamic adjustment module is used for performing parameter adjustment processing on the teaching content mapping relation of the holographic projection system according to the student learning deviation degree to generate a scene change instruction set; and the feedback optimization module is used for performing trend analysis processing on the student response data after execution of the scene change instruction set, and generating an optimized interaction experience parameter set. By adopting the device, the geography teaching level and education experience can be improved.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent teaching equipment, and particularly relates to a geography teaching device and an interaction system based on holographic projection. Background Art

[0002] As a frontier direction in the field of educational technology, holographic projection geography teaching interaction combines virtual reality and spatial interaction technologies, which is of crucial significance for enhancing the learning experience and knowledge transfer efficiency of the geography discipline. Its importance lies in breaking the limitations of traditional planar teaching through intuitive three-dimensional presentation and dynamic interaction, enabling students to deeply participate in the exploration of geographical phenomena in an immersive environment. However, current solutions mostly remain at the technical demonstration level, and there is generally a problem of insufficient data utilization. Traditional teaching systems often only record the explicit behaviors of students, such as multiple-choice answers or class attendance rates, while ignoring the implicit data during the interaction process, such as perspective preferences and operation habits, resulting in rough teaching feedback and difficulty in accurately adapting to teaching needs.

[0003] Although holographic projection technology has made breakthroughs in visual presentation, the development of its interaction data system still faces significant challenges. The core problems focus on how to effectively collect and analyze the interaction data between students and holographic content, and how to convert this data into a basis for operable teaching adjustments. Specifically, the comprehensiveness of data, the real-time nature of analysis, and the personalization of applications are the three major technical bottlenecks. The comprehensiveness of data is limited by the system's ability to capture students' multi-dimensional behaviors. For example, the frequency of perspective selection and question asking is often ignored; the real-time nature of analysis is difficult to meet the dynamic needs of the classroom due to the lag of big data processing technology; the personalization of applications is also limited due to the lack of pattern recognition for individual learning characteristics. These unresolved technical factors prevent the holographic interaction system from fully exerting its teaching potential and achieving a seamless match between content presentation and students' needs.

[0004] Traditional holographic geography teaching systems cannot capture multi-dimensional interaction data in real time and lack the ability to dynamically adjust teaching content, resulting in unsatisfactory teaching effects. Summary of the Invention

[0005] Based on this, in view of the above technical problems, it is necessary to provide a geography teaching device and an interaction system based on holographic projection, which can collect and analyze the interaction data between students and holographic geography content in real time, generate an interaction experience parameter set to adjust the holographic projection content, so as to improve the geography teaching level and educational experience.

[0006] In a first aspect, the present application provides a geography teaching device based on holographic projection, including a data acquisition module, a behavior analysis module, a dynamic adjustment module, and a feedback optimization module: The data acquisition module is used to obtain interaction data, perform timestamp marking and sensor data fusion processing on the interaction data, and generate a multi-dimensional behavior dataset containing timestamps; the interaction data is used to characterize the interaction behavior between students and holographic projection geographical content; The behavior analysis module is used to extract perspective preference features and analyze operation habits from the multi-dimensional behavior dataset, and generate the student learning deviation; The dynamic adjustment module is used to perform parameter adjustment processing on the teaching content mapping relationship of the holographic projection system according to the student learning deviation, and generate a scene transformation instruction set; The feedback optimization module is used to perform trend analysis processing on the student response data after the execution of the scene transformation instruction set, and generate an optimized interaction experience parameter set, and the interaction experience parameter set is used to instruct the holographic projection device to generate holographic images.

[0007] In a possible embodiment, the data acquisition module includes: The multi-source data acquisition unit is used to obtain the perspective movement trajectory of the student, and perform quantization processing on the touch operation force based on the pressure sensor, and generate an original behavior data stream; The data preprocessing unit is used to perform sensor noise compensation processing on the original behavior data stream, and generate a denoised original dataset; The timestamp generation unit is used to perform time synchronization processing on the original dataset, align the time axis of the original dataset based on the time axis alignment algorithm, and generate a multi-dimensional behavior dataset with timestamp marking.

[0008] In a possible embodiment, the behavior analysis module includes: The data cleaning unit is used to filter out outliers from the multi-dimensional behavior dataset, and eliminate noise data through the sliding window algorithm, and generate a standardized behavior dataset; The perspective preference analysis unit is used to extract the sequence of fixation point coordinates from the standardized behavior dataset, and perform spatial aggregation processing on the sequence of fixation point coordinates based on the density clustering algorithm, and generate a perspective preference pattern; The operation habit analysis unit is used to perform frequent pattern mining processing on the operation instruction sequence in the standardized behavior dataset, and match it with the preset behavior template through the dynamic time warping algorithm, and generate an operation habit feature vector; The deviation calculation unit is used to map the perspective preference pattern and the operation habit feature vector to the node space of the pre-trained geographical knowledge graph to obtain the student learning vector, calculate the semantic distance between the student learning vector and the teaching target vector based on the graph neural network, and generate the student learning deviation according to the semantic distance.

[0009] In a possible embodiment, the deviation calculation unit includes: A difference comparison subunit, configured to map the perspective preference pattern to the node space of the geographical knowledge graph, calculate the matching degree with the target knowledge node based on the cosine similarity, and generate a first deviation coefficient; A regression analysis subunit, configured to perform multivariate linear regression processing on the operation habit feature vector, and generate a second deviation coefficient based on the proportional relationship between the regression residual and the preset threshold; A weight fusion subunit, configured to perform dynamic weight allocation on the first deviation coefficient and the second deviation coefficient through the entropy weight method, and generate the student learning deviation degree.

[0010] In a possible embodiment, the dynamic adjustment module includes: An adjustment coefficient generation unit, configured to optimize the teaching content mapping function through the gradient descent algorithm based on the student learning deviation degree, and generate an initial adjustment coefficient of the scene transformation parameter; An incremental learning unit, configured to input the initial adjustment coefficient into the incremental learning model, and perform fusion processing on the historical interaction data based on the stochastic weight averaging algorithm to generate a dynamically updated personalized pattern; An instruction generation unit, configured to match the spatial interaction interface protocol of the holographic projection system according to the personalized pattern, and convert the adjustment coefficient into a scene rendering instruction set, where the scene rendering instruction set is used to instruct the holographic projection device to control the laser parameters and / or the light field angle.

[0011] In a possible embodiment, the incremental learning unit includes: A historical retention evaluation subunit, configured to evaluate the historical knowledge retention rate after model update through sliding window cross-validation, and generate a stability verification result; A weight adjustment subunit, configured to dynamically adjust the model weight distribution according to the stability verification result, and generate a coverage threshold parameter, where the coverage threshold parameter is used to limit the coverage range of new knowledge to historical data; A threshold constraint subunit, configured to truncate the parameter update gradient of the incremental learning model based on the coverage threshold parameter, and generate a personalized pattern after catastrophic forgetting suppression, where the personalized pattern is used to update the teaching content mapping relationship of the holographic projection system.

[0012] In a possible embodiment, the feedback optimization module includes: A response evaluation unit, configured to extract the interaction frequency and operation accuracy rate from the student response data after the execution of the scene rendering instruction set, and perform modeling processing on the response change trend based on the hidden Markov model to generate a response evaluation value; An instruction optimization unit, configured to, when the response evaluation value is lower than the preset threshold, perform partitioning processing on the feature space through the C4.5 decision tree algorithm, and generate an improved parameter set driven by classification rules; A real-time update unit for inputting an improved parameter set into a streaming data processing engine, dynamically loading and processing geographical scene resources based on a time-sensitive hash index, and generating an optimized interactive experience parameter set.

[0013] In a second aspect, the present application also provides a geographical teaching interaction method based on holographic projection, including: Obtaining interaction data, performing timestamp marking and sensor data fusion processing on the interaction data to generate a multi-dimensional behavior data set containing timestamps; the interaction data is used to characterize the interaction behavior between students and holographic projection geographical content; Performing perspective preference feature extraction and operation habit analysis processing on the multi-dimensional behavior data set to generate a student learning deviation degree; Performing parameter adjustment processing on the teaching content mapping relationship of the holographic projection system according to the student learning deviation degree to generate a scene transformation instruction set; Performing trend analysis processing on the student response data after the execution of the scene transformation instruction set to generate an optimized interactive experience parameter set, and the interactive experience parameter set is used to instruct the holographic projection device to generate a holographic image.

[0014] In a third aspect, the present application also provides a geographical teaching interaction system based on holographic projection, including: A holographic projection device, an action capture sensor array, and a terminal processor, wherein the terminal processor is configured to implement the geographical teaching interaction based on holographic projection in the previous embodiment; The terminal processor is respectively communicatively connected to the holographic projection device and the action capture sensor array.

[0015] In a fourth aspect, the present application also provides a computer device, including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the steps of the geographical teaching method based on holographic projection as described above are implemented.

[0016] The above-mentioned holographic projection-based geography teaching device and interaction system timestamp the interaction data between students and holographic projection geography content and perform multi-sensor data fusion processing to generate a multi-dimensional behavior data set containing spatio-temporal correlation information, breaking through the limitations of traditional systems in capturing implicit interaction behaviors; based on the multi-dimensional behavior data set, extract perspective preference features and analyze operation habits to generate the learning deviation degree of students, quantifying the dynamic difference between the current cognitive state of students and the teaching objectives; the dynamic adjustment module parametrically adjusts the teaching content mapping relationship of the holographic projection system according to the learning deviation degree of students to generate a scene transformation instruction set, realizing the real-time adaptation of teaching content to individual learning needs; the feedback optimization module generates an optimized interaction experience parameter set by analyzing the trend of students' response data after the execution of the scene transformation instruction set, driving the holographic projection device to dynamically update the geography scene. The above method can improve the utilization rate of interaction data in the holographic geography teaching system and enhance the teaching effect through the real-time analysis of multi-dimensional behavior data and the dynamic iteration of teaching parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 It is a schematic structural diagram of a holographic projection-based geography teaching device provided by an embodiment of the present invention; Figure 2 It is a schematic flowchart of a holographic projection-based geography teaching interaction method provided by an embodiment of the present invention; Figure 3 It is a schematic structural diagram of a holographic projection-based geography teaching interaction system provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] In order to make the objectives, technical solutions and advantages of the present application more clear, the following further details the present application in conjunction with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0020] First, a brief introduction is given to the nouns involved in the embodiments of the present application.

[0021] Holographic projection is a technology that records and reproduces the three-dimensional image of an object by using the principles of light interference and diffraction. Through special optical devices and materials, it captures the light wave information of the object, including amplitude and phase, and stores it in the holographic medium. When the holographic medium is illuminated by a specific light source, it can reconstruct the three-dimensional stereoscopic image of the object, enabling people to observe the visual effect almost the same as the real object from different angles. Holographic projection technology is widely used in fields such as display, education, and entertainment, bringing immersive visual experiences to people.

[0022] A geographical knowledge graph is a semantic network that represents geographical knowledge in the form of a graph structure. By structuring and storing geographical entities (such as cities, mountains, rivers, etc.), geographical concepts (such as terrain types, climate zones, etc.), and their complex relationships (such as geographical location associations, administrative subordination relationships, etc.), it realizes the efficient organization and semantic association of geographical information. It can provide powerful support for the query, analysis, reasoning, and visualization of geographical information, and is widely used in fields such as geographical education, intelligent navigation, environmental monitoring, and urban planning, helping users to understand and utilize geographical knowledge more deeply.

[0023] Graph Neural Networks (GNNs) are a class of deep learning models for processing graph-structured data, aiming to capture the topological structure and node feature information of the graph by learning the representations of nodes, edges, and their relationships in the graph. It updates the feature representation of the current node by aggregating the information of neighboring nodes, thereby achieving feature extraction and modeling of graph-structured data. Graph neural networks perform well in multiple fields such as social network analysis, molecular structure prediction, recommendation systems, and knowledge graphs, and can effectively solve tasks such as node classification, graph classification, and link prediction, providing a powerful tool for processing complex graph-structured data.

[0024] According to the above noun explanations, the implementation environment of the geographical teaching device 10 based on holographic projection provided in the embodiments of the present application is described. Schematically, the implementation environment includes: a holographic projection device, an action capture sensor array, and a terminal processor. Among them, the terminal processor is signal-connected to the action capture sensor and the holographic projection device through a network device; the holographic projection device can be a holographic projector, a holographic projection screen, or a holographic projection system; the action capture sensor array includes, but is not limited to, an infrared camera array, a depth camera array, an inertial measurement unit sensor, an optical motion capture system, and a gesture recognition sensor, etc.; the terminal processor can be a graphics processor, a multi-core central processing unit, an embedded system processor, or a smart terminal processor, etc., which is not limited here.

[0025] Combined with the above-mentioned noun explanations and implementation environments, the application scenarios of the embodiments of this application will be described. The holographic projection-based geography teaching device 10 provided in the embodiments of this application can be applied to the following scenarios including but not limited to: In a geography class, the teacher uses a holographic projection device to display three-dimensional geographical content such as the Earth model, terrain and landforms, and climate distribution. Students can observe the geographical features of different regions by gesture operations (such as rotating and zooming the Earth model). The data acquisition module captures the students' interaction behaviors in real time. The behavior analysis module analyzes the students' operation habits and perspective preferences. The dynamic adjustment module automatically adjusts the display content of the holographic projection according to the students' learning deviation degree. For example, it highlights the regions that students pay less attention to and provides more detailed explanations. The feedback optimization module further optimizes the interaction experience according to the students' subsequent reactions to ensure that students can better understand and participate in learning. By using this device, students can learn geographical knowledge in a more intuitive and active way, and teachers can adjust teaching strategies according to real-time feedback to improve teaching effects.

[0026] In a virtual geography exploration scenario, students can "enter" a virtual geographical environment, such as exploring the Amazon rainforest or climbing Mount Everest. By using this device, the holographic projection device generates a realistic three-dimensional geographical scene. The motion capture sensor array captures the students' actions in real time, enabling students to interact with the virtual environment through natural body language, such as touching virtual plants with their hands and climbing virtual mountains. The behavior analysis module analyzes the students' operation habits and learning deviation degree. The dynamic adjustment module adjusts the scene difficulty or provides hint information according to the analysis results. The feedback optimization module further optimizes the interaction experience according to the students' reactions to ensure that students always maintain a high degree of participation and learning interest during the exploration process, making students feel as if they are in a real geographical environment, enhancing the immersion and interest of learning, and at the same time dynamically adjusting the content according to the students' learning status to ensure learning effects.

[0027] Illustratively, the holographic projection-based geography teaching device 10 provided in the embodiments of this application can also be applied to other application scenarios. Here, only examples are given and the specific application scenarios are not limited.

[0028] In an exemplary embodiment, as Figure 1 shown, a holographic projection-based geography teaching device 10 is provided. In this embodiment, an example is given where the device is applied to the terminal processor in the foregoing implementation environment. A holographic projection-based geography teaching device 10 provided in this embodiment includes a data acquisition module 11, a behavior analysis module 12, a dynamic adjustment module 13, and a feedback optimization module 14: The data acquisition module 11 is used to obtain interaction data, perform timestamp marking and sensor data fusion processing on the interaction data, and generate a multidimensional behavior data set containing timestamps; the interaction data is used to characterize the interaction behavior between students and holographic projection geography content.

[0029] For example, multiple motion capture sensors arranged around the holographic projection device can be used to obtain real-time interaction data between students and holographic projection geographical content, and the interaction data can be timestamped. Through multi-sensor fusion and timestamp synchronization technology, the limitations of traditional single behavior records can be broken through, and multi-dimensional implicit data such as viewing angle trajectory and operation strength can be accurately captured, thereby improving the comprehensiveness of behavior analysis.

[0030] The behavior analysis module 12 is used to extract perspective preference features and analyze operation habits of the multi-dimensional behavior data set to generate student learning deviation.

[0031] Specifically, we can extract perspective preference features from the data, analyze the time and frequency of students' gaze, and analyze students' operating habits, such as gestures, touch frequency, and interaction paths, to generate student learning deviations, which are used to assess the degree of deviation between students' learning status and the preset standard learning model. By combining density clustering with knowledge graph reasoning, discrete interactive behaviors are mapped into quantifiable learning deviation indicators, solving the problem of the extensiveness of traditional evaluation methods that rely on explicit answer data, and achieving accurate diagnosis of students' cognitive status.

[0032] The dynamic adjustment module 13 is used to perform parameter adjustment processing on the teaching content mapping relationship of the holographic projection system according to the student learning deviation, and generate a scene change instruction set.

[0033] Specifically, the teaching content mapping relationship of the holographic projection system can be adjusted according to the students' learning deviation. For example, if the students' learning deviation is high, the device can adjust the display parameters of the holographic projection equipment, such as increasing the highlighting of key areas, adjusting the viewing angle, or adding auxiliary instructions, etc., to generate a scene change instruction set to guide the holographic projection equipment to update the display method of the teaching content, so as to respond to the students' learning needs in real time, dynamically adjust the presentation method of the teaching content, ensure that the teaching process always fits the students' learning progress and understanding, and improve the teaching effect and students' learning enthusiasm.

[0034] The feedback optimization module 14 is used to perform trend analysis on the student response data after the scene change instruction set is executed, and generate an optimized interactive experience parameter set, which is used to instruct the holographic projection device to generate a holographic image.

[0035] Specifically, after the holographic projection device executes the scene transformation instruction set, the response data of the students to the new scene is obtained in real time, which may include new interaction behaviors and changes in line of sight, etc. Trend analysis processing is performed on these response data to evaluate the acceptance degree and learning effect of the students on the adjusted teaching content. According to the analysis results, the module generates an optimized set of interactive experience parameters and further adjusts the display parameters of the holographic projection device to optimize the learning experience of the students.

[0036] The above-mentioned geographical teaching device based on holographic projection marks the interaction data between students and holographic projection geographical content with time stamps and performs multi-sensor data fusion processing to generate a multi-dimensional behavior data set containing spatio-temporal correlation information, breaking through the limitations of traditional systems in capturing implicit interaction behaviors; based on the multi-dimensional behavior data set, perspective preference feature extraction and operation habit analysis processing are performed to generate the learning deviation degree of students, quantifying the dynamic difference between the current cognitive state of students and the teaching objectives; according to the learning deviation degree of students, parametric adjustment is performed on the teaching content mapping relationship of the holographic projection system to generate a scene transformation instruction set, realizing the real-time adaptation of teaching content to individual learning needs; through trend analysis of the student response data after the execution of the scene transformation instruction set, an optimized set of interactive experience parameters is generated to drive the holographic projection device to dynamically update the geographical scene. The above method can improve the utilization rate of interactive data in the holographic geography teaching system and improve the teaching effect through real-time analysis of multi-dimensional behavior data and dynamic iteration of teaching parameters.

[0037] In a possible embodiment, the data acquisition module 11 may include: A multi-source data acquisition unit 111, configured to acquire the perspective movement trajectory of students and perform quantization processing on the touch operation force based on a pressure sensor to generate an original behavior data stream.

[0038] Specifically, the perspective movement trajectory of students and the touch operation force can be obtained through a camera and a pressure sensor to generate an original behavior data stream, so as to comprehensively capture the interaction behaviors of students, including not only visual interaction but also tactile interaction, ensuring that every detail of the teaching process can be effectively recorded and providing high-information-density original data for subsequent analysis.

[0039] A data preprocessing unit 112, configured to perform sensor noise compensation processing on the original behavior data stream to generate a denoised original data set.

[0040] Exemplarily, an adaptive filtering algorithm can be adopted to suppress noise in the original behavior data stream: First, a high-frequency jitter noise in the infrared tracking data is eliminated by a moving average filter, and a low-frequency effective displacement signal is retained; a wavelet threshold denoising algorithm is used to separate the environmental electromagnetic interference component in the touch pressure signal, and a smooth pressure change curve is reconstructed; the denoised perspective trajectory and touch force data are normalized, mapped to a unified dimension space, and a denoised standardized original data set is generated, improving the accuracy and reliability of the data and avoiding data distortion caused by sensor errors or external interferences.

[0041] The timestamp generation unit 113 is configured to perform time synchronization processing on the original data set, align the time axis of the original data set based on a time axis alignment algorithm, and generate a multi-dimensional behavior data set marked with timestamps.

[0042] Specifically, a global time reference can be constructed based on a precision clock source: the local clocks of each sensor are synchronized through the IEEE 1588 Precision Time Protocol (PTP), and a timestamp label with nanosecond-level accuracy is attached to each behavior record in the standardized original data set. Further, a Dynamic Time Warping (DTW) algorithm can be adopted to perform non-linear alignment on the time axes of multi-source data streams, eliminate the timing deviation caused by sensor response delays, generate a multi-dimensional behavior data set with strictly synchronized timestamps and spatially aligned, solve the time drift problem of multi-sensor data, ensure the spatio-temporal consistency of behavior data, and provide a reliable basis for temporal correlation analysis.

[0043] In a possible embodiment, the behavior analysis module 12 may include: The data cleaning unit 121 is configured to perform outlier filtering processing on the multi-dimensional behavior data set, and eliminate noise data through a sliding window algorithm, and generate a standardized behavior data set.

[0044] Specifically, outlier filtering processing is performed on the multi-dimensional behavior data set, data points that significantly do not conform to normal interaction behaviors are identified and removed, such as perspective angles or operation forces that exceed reasonable ranges, the remaining data is processed through a sliding window algorithm, the data within the window is smoothed, and noise data is eliminated, generating a standardized behavior data set, providing a clean and accurate data basis for subsequent analysis.

[0045] The perspective preference analysis unit 122 is configured to extract a sequence of fixation point coordinates from the standardized behavior data set, and perform spatial aggregation processing on the sequence of fixation point coordinates based on a density clustering algorithm, and generate a perspective preference pattern.

[0046] Specifically, extract the sequence of fixation point coordinates of students, perform spatial aggregation processing on these sequences of fixation point coordinates based on the density clustering algorithm, and through cluster analysis, divide the discrete fixation point coordinates into multiple high-density aggregation regions; by calculating the centroid coordinates and coverage area of each cluster, generate a perspective preference pattern characterizing the distribution characteristics of students' attention foci, including hot spot area coordinates, proportion of residence time, and scanning path shape parameters, to achieve quantitative characterization of the perspective behavior pattern.

[0047] The operation habit analysis unit 123 is used to perform frequent pattern mining processing on the operation instruction sequences in the standardized behavior dataset, and match them with the preset behavior templates through the dynamic time warping algorithm to generate operation habit feature vectors.

[0048] Specifically, a pattern mining algorithm with prefix projection can be used to perform frequent subsequence mining on the operation instruction sequences to identify recurring behavior patterns; at the same time, calculate the time morphological difference degree between the student operation sequence and the preset standard template through the dynamic time warping (DTW) algorithm to generate a multi-dimensional operation habit feature vector including operation frequency, timing deviation, and path similarity; this feature vector is processed by Z-score standardization to eliminate the dimension difference and form a quantifiable and comparable behavior feature index.

[0049] The deviation calculation unit 124 is used to map the perspective preference pattern and the operation habit feature vector to the node space of the pre-trained geographical knowledge graph to obtain the student learning vector, calculate the semantic distance between the student learning vector and the teaching target vector based on the graph neural network, and generate the student learning deviation according to the semantic distance.

[0050] Specifically, input the perspective preference pattern and the operation habit feature vector into the embedding model of the pre-trained geographical knowledge graph, map them to the node space of the knowledge graph through the graph attention network (GAT) to generate a 128-dimensional student learning vector integrating spatial behavior and cognitive features; calculate the similarity between the teaching target vector (pre-generated by curriculum standard knowledge points) constructed based on the graph neural network and the student learning vector to obtain a semantic matching degree in the range of 0-1; quantify (1 - semantic matching degree) as the student learning deviation index, which is used to characterize the difference between the student's current cognitive state and the teaching target, provide a quantitative basis for dynamically adjusting teaching content, ensure that the teaching process always fits the student's learning progress and understanding level, and improve the teaching effect.

[0051] In a possible embodiment, the deviation calculation unit 124 may include: The difference comparison subunit 1241 is used to map the perspective preference pattern to the node space of the geographical knowledge graph, calculate the matching degree with the target knowledge node based on the cosine similarity, and generate the first deviation coefficient.

[0052] Specifically, map the perspective preference pattern to the node space of the pre-trained geographical knowledge graph, compare the perspective preference pattern with the target knowledge nodes in the knowledge graph, calculate the matching degree between the perspective preference pattern and the target knowledge nodes based on the cosine similarity, and generate a first deviation coefficient by quantifying the matching degree, which is used to characterize the difference degree between the student's perspective preference and the teaching objective.

[0053] The regression analysis subunit 1242 is used to perform multivariate linear regression processing on the operation habit feature vector, and generate a second deviation coefficient based on the proportional relationship between the regression residual and the preset threshold.

[0054] Specifically, perform multivariate linear regression processing on the operation habit feature vector, analyze the linear relationship between the feature vector and the preset teaching objective feature vector, and generate a second deviation coefficient based on the proportional relationship between the regression residual and the preset threshold, which is used to characterize the deviation degree between the student's operation habit and the teaching objective.

[0055] The weight fusion subunit 1243 is used to perform dynamic weight allocation on the first deviation coefficient and the second deviation coefficient by the entropy weight method to generate the student learning deviation degree.

[0056] Exemplarily, the entropy weight method can be used to determine the dynamic weights of the first deviation coefficient and the second deviation coefficient: calculate the information entropy values of each coefficient in the historical dataset, and the lower the information entropy, the higher the weight; allocate the weight coefficients according to the calculation results of the entropy values (perspective deviation weight) and (operation deviation weight), satisfying ; generate a comprehensive evaluation index through the weighted summation formula [student learning deviation degree = × first deviation coefficient + × second deviation coefficient], and this index is output to the dynamic adjustment module 13 after Z-score standardization. The entropy weight method objectively quantifies the contribution degrees of different deviation dimensions, avoids the subjective deviation of manually setting weights, ensures that the calculation of the student learning deviation degree is more scientific and reasonable, provides a more accurate quantitative evaluation for the subsequent dynamic adjustment of teaching content, and improves the personalization and adaptability of teaching.

[0057] In a possible embodiment, the dynamic adjustment module 13 includes: The adjustment coefficient generation unit 131 is used to optimize the teaching content mapping function by the gradient descent algorithm based on the student learning deviation degree, and generate an initial adjustment coefficient of the scene transformation parameter.

[0058] Specifically, an optimization objective for constructing a teaching content mapping function can be based on the learning deviation of students: using the deviation value as the weight factor of the loss function, calculating the partial derivative directions of the terrain complexity parameter, dynamic demonstration speed parameter, and information density through the gradient descent algorithm; for example, iteratively updating the parameter values with a learning rate of 0.01 until the loss function converges, generating an initial adjustment coefficient set including hierarchical scaling coefficients, light intensity gradients, and animation frame rates, and mapping the numerical range of the coefficient set to the interval [0, 1] after Min-Max normalization to dynamically adjust the presentation mode of teaching content according to the learning deviation of students.

[0059] The incremental learning unit 132 is used to input the initial adjustment coefficients into the incremental learning model, and perform fusion processing on historical interaction data based on the Stochastic Weight Averaging (SWA) algorithm to generate a dynamically updated personalized pattern.

[0060] Specifically, the initial adjustment coefficients can be input into the historical parameter pool of the incremental learning model, and the model fusion is performed using the Stochastic Weight Averaging (SWA) algorithm: for example, extracting the most recent 100 groups of valid samples from the historical interaction data to construct a sliding window data set, and calculating the exponentially weighted moving average of the weights of each model within the window; smoothing the high-frequency fluctuations in the parameter update process through the momentum factor μ = 0.9 to generate a dynamically updated personalized pattern that combines historical experience and real-time feedback.

[0061] The instruction generation unit 133 is used to match the spatial interaction interface protocol of the holographic projection system according to the personalized pattern, and convert the adjustment coefficients into a scene rendering instruction set, which is used to instruct the holographic projection device to control the laser parameters and / or light field angles.

[0062] Specifically, according to the personalized pattern generated by the incremental learning unit 132, the spatial interaction interface protocol of the holographic projection system is matched, and the adjustment coefficients are converted into a scene rendering instruction set, which is used to instruct the holographic projection device to control the laser parameters and / or light field angles. In this way, the holographic projection device can dynamically adjust the display content according to the personalized needs of students and optimize the teaching effect.

[0063] In a possible embodiment, the incremental learning unit 132 includes: The historical retention evaluation sub-unit 1321 is used to evaluate the historical knowledge retention rate after model update through sliding window cross-validation to generate a stability verification result.

[0064] Specifically, by using the sliding window technique to select historical data from different time periods, dividing it into multiple validation sets, verifying the performance of the model after each update, and calculating the degree of retention of historical knowledge by the model, so as to generate a stability verification result to provide a basis for subsequent weight adjustment.

[0065] The weight adjustment subunit 1322 is configured to perform dynamic adjustment processing on the model weight distribution according to the stability verification result, and generate a coverage threshold parameter, which is used to limit the coverage range of new knowledge over historical data.

[0066] Specifically, according to the stability verification result generated by the historical retention evaluation subunit, dynamic adjustment processing is performed on the weight distribution of the incremental learning model. This subunit adjusts the weights of each parameter in the model according to the stability verification result to optimize the model's ability to retain historical knowledge, and generates a coverage threshold parameter, which is used to limit the coverage range of new knowledge over historical data, ensuring that the model does not overly forget historical knowledge when learning new knowledge.

[0067] The threshold constraint subunit 1323 is configured to perform truncation processing on the parameter update gradient of the incremental learning model based on the coverage threshold parameter, and generate a personalized pattern after catastrophic forgetting suppression, which is used to update the teaching content mapping relationship of the holographic projection system.

[0068] Exemplarily, truncation processing is performed on the parameter update gradient of the incremental learning model based on the coverage threshold parameter generated by the weight adjustment subunit. Specifically, by restricting the gradient amplitude of the parameter update, it is prevented that the model causes catastrophic forgetting of historical knowledge when learning new knowledge, thereby generating a personalized pattern after catastrophic forgetting suppression. This personalized pattern is used to update the teaching content mapping relationship of the holographic projection system, ensuring that the dynamic adjustment of the teaching content not only adapts to the new needs of students but also retains important historical knowledge.

[0069] In a possible embodiment, the feedback optimization module 14 includes: The response evaluation unit 141 is configured to extract the interaction frequency and operation accuracy rate from the student response data after the execution of the scene rendering instruction set, perform modeling processing on the response change trend based on the hidden Markov model, and generate a response evaluation value.

[0070] Specifically, the student response data can be extracted from the interaction interface of the holographic projection device through buried point log analysis, the touch event log after the execution of the scene rendering instruction set is parsed, and the interaction operation frequency within a unit time is counted; the similarity between the actual operation path of the student and the preset standard path is calculated through the operation sequence matching algorithm to generate an operation accuracy rate index in the range of 0 - 100%; further, the time series change of the above indexes is modeled based on the hidden Markov model (HMM), the state transition probability matrix is trained through the Baum-Welch algorithm, and the expected response gain within the next 3 time windows is calculated to generate a 0 - 1 normalized response evaluation value, where an evaluation value lower than 0.6 triggers the optimization mechanism.

[0071] An instruction optimization unit 142, configured to, when the response evaluation value is lower than a preset threshold, divide the feature space through a C4.5 decision tree algorithm to generate an improved parameter set driven by classification rules.

[0072] Specifically, it is started when the response evaluation value is lower than the preset threshold, and divides the feature space through a C4.5 decision tree algorithm. Based on the features in the student response data, this unit constructs a decision tree model and generates an improved parameter set driven by classification rules to accurately divide different learning groups according to the student features and learning situations, and generate targeted optimization suggestions for each group.

[0073] A real-time update unit 143, configured to input the improved parameter set into a streaming data processing engine, perform dynamic loading processing on geographical scene resources based on a time-sensitive hash index, and generate an optimized interactive experience parameter set.

[0074] Specifically, it inputs the improved parameter set into a streaming data processing engine, performs dynamic loading processing on geographical scene resources based on a time-sensitive hash index, and generates an optimized interactive experience parameter set. This unit can quickly respond to the requirements of the improved parameter set, dynamically adjust the display content and interactive parameters of the holographic projection device, and ensure the real-time update and optimization of teaching content. Through the time-sensitive hash index, this unit can efficiently load geographical scene resources, improve the response speed and operation efficiency of the system, and provide students with a smooth and optimized learning experience.

[0075] In summary, the holographic projection-based geography teaching device 10 provided by the embodiments of the present application captures the interaction behaviors between students and holographic geography content in real time through multi-source sensor fusion technology, including multi-dimensional data such as the perspective movement trajectory and touch operation intensity, and generates a high-precision multi-dimensional behavior data set through timestamp synchronization and noise suppression processing; based on the density clustering algorithm and the geographical knowledge graph inference engine, extracts the perspective preference pattern and operation habit feature vectors from the behavior data, and combines the graph neural network to calculate the semantic deviation degree between the student's cognitive state and the teaching goal; drives the incremental learning model based on the deviation index to dynamically optimize the teaching content mapping function, and generates a scene transformation instruction set for controlling the light field parameters and terrain rendering logic; performs real-time evaluation and parameter iteration on the teaching effect through the hidden Markov model and the streaming computing framework to form an adaptive personalized teaching strategy. The above technical solutions can improve the interactive experience of holographic projection teaching, ensure that the teaching process always meets the learning needs of students, and improve the learning enthusiasm and teaching effect of students.

[0076] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are sequentially shown according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear description in this article, there is no strict order limit for the execution of these steps, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0077] Based on the same inventive concept, an embodiment of the present application also provides a holographic projection-based geography teaching method for implementing the above-mentioned holographic projection-based geography teaching device 10. The implementation solution provided by this method for solving problems is similar to the implementation solution recorded in the above device. Therefore, the specific limitations in one or more embodiments of the following holographic projection-based geography teaching methods can refer to the limitations on the holographic projection-based geography teaching device 10 in the above text, and will not be repeated here.

[0078] In an exemplary embodiment, as Figure 2 shown, a holographic projection-based geography teaching interaction method is provided, including the following steps: S101, obtain interaction data, perform timestamp marking and sensor data fusion processing on the interaction data, and generate a multi-dimensional behavior data set containing timestamps; the interaction data is used to characterize the interaction behavior between students and holographic projection geography content.

[0079] S102, perform perspective preference feature extraction and operation habit analysis processing on the multi-dimensional behavior data set, and generate a student learning deviation degree.

[0080] S103, perform parameter adjustment processing on the teaching content mapping relationship of the holographic projection system according to the student learning deviation degree, and generate a scene transformation instruction set.

[0081] S104, perform trend analysis processing on the student response data after the execution of the scene transformation instruction set, and generate an optimized interaction experience parameter set. The interaction experience parameter set is used to instruct the holographic projection device to generate a holographic image.

[0082] In a possible embodiment, performing timestamp marking and sensor data fusion processing on the interaction data to generate a multi-dimensional behavior data set containing timestamps may include: S201, obtain the perspective movement trajectory of the student, and perform quantization processing on the touch operation force based on the pressure sensor to generate an original behavior data stream.

[0083] S202. Perform sensor noise compensation processing on the original behavior data stream to generate a denoised original data set.

[0084] S203. Perform time synchronization processing on the original data set, align the time axis of the original data set based on the time axis alignment algorithm, and generate a multi-dimensional behavior data set marked with time stamps.

[0085] In a possible embodiment, the behavior analysis module includes: S301. Perform outlier filtering processing on the multi-dimensional behavior data set, and eliminate noise data through the sliding window algorithm to generate a standardized behavior data set.

[0086] S302. Extract the gaze point coordinate sequence from the standardized behavior data set, perform spatial aggregation processing on the gaze point coordinate sequence based on the density clustering algorithm, and generate a perspective preference pattern.

[0087] S303. Perform frequent pattern mining processing on the operation instruction sequence in the standardized behavior data set, and match it with the preset behavior template through the dynamic time warping algorithm to generate an operation habit feature vector.

[0088] S304. Map the perspective preference pattern and the operation habit feature vector to the node space of the pre-trained geographical knowledge graph to obtain the student learning vector, calculate the semantic distance between the student learning vector and the teaching target vector based on the graph neural network, and generate the student learning deviation degree according to the semantic distance.

[0089] In a possible embodiment, calculating the semantic distance between the student learning vector and the teaching target vector based on the graph neural network, and generating the student learning deviation degree includes: S401. Map the perspective preference pattern to the node space of the geographical knowledge graph, calculate the matching degree with the target knowledge node based on the cosine similarity, and generate a first deviation coefficient.

[0090] S402. Perform multivariate linear regression processing on the operation habit feature vector, and generate a second deviation coefficient based on the proportional relationship between the regression residual and the preset threshold.

[0091] S403. Perform dynamic weight allocation on the first deviation coefficient and the second deviation coefficient through the entropy weight method to generate the student learning deviation degree.

[0092] In a possible embodiment, performing parameter adjustment processing on the teaching content mapping relationship of the holographic projection system according to the student learning deviation degree to generate a scene transformation instruction set includes: S501. Optimize the teaching content mapping function through the gradient descent algorithm based on the student learning deviation degree, and generate the initial adjustment coefficient of the scene transformation parameter.

[0093] S502. Input the initial adjustment coefficient into the incremental learning model, and fuse the historical interaction data based on the stochastic weight averaging algorithm to generate a dynamically updated personalized pattern.

[0094] S503. Match the personalized pattern with the spatial interaction interface protocol of the holographic projection system, and convert the adjustment coefficient into a scene rendering instruction set, which is used to instruct the holographic projection device to control the laser parameters and / or the light field angle.

[0095] In a possible embodiment, inputting the initial adjustment coefficient into the incremental learning model and fusing the historical interaction data based on the stochastic weight averaging algorithm to generate a dynamically updated personalized pattern includes: S601. Evaluate the retention rate of historical knowledge after model update through sliding window cross-validation to generate a stability verification result.

[0096] S602. Dynamically adjust the model weight distribution according to the stability verification result to generate a coverage threshold parameter, which is used to limit the coverage range of new knowledge to historical data.

[0097] S603. Truncate the parameter update gradient of the incremental learning model based on the coverage threshold parameter to generate a personalized pattern after catastrophic forgetting suppression, which is used to update the teaching content mapping relationship of the holographic projection system.

[0098] In a possible embodiment, performing trend analysis on the student response data after the execution of the scene transformation instruction set to generate an optimized set of interaction experience parameters includes: S701. Extract the interaction frequency and operation accuracy rate from the student response data after the execution of the scene rendering instruction set, and model the response change trend based on the hidden Markov model to generate a response evaluation value; S702. When the response evaluation value is lower than the preset threshold, divide the feature space through the C4.5 decision tree algorithm to generate an improved parameter set driven by classification rules; S703. Input the improved parameter set into the streaming data processing engine, and dynamically load the geographical scene resources based on the time-sensitive hash index to generate an optimized set of interaction experience parameters.

[0099] Based on the same inventive concept, an embodiment of the present application further provides a holographic projection-based geographical teaching interaction system for implementing the holographic projection-based geographical teaching interaction method involved above. The implementation solutions provided by this system for solving problems are similar to those recorded in the above method. Therefore, the specific limitations in one or more embodiments of the holographic projection-based geographical teaching interaction system provided below can refer to the limitations on the holographic projection-based geographical teaching interaction method in the foregoing text, and will not be elaborated here.

[0100] In an exemplary embodiment, as Figure 3 shown, a holographic projection-based geographical teaching interaction system is provided, including: a holographic projection device, an action capture sensor array, and a terminal processor, wherein the terminal processor is configured to implement the holographic projection-based geographical teaching interaction method in any of the foregoing embodiments.

[0101] The terminal processor is communicatively connected to the holographic projection device and the action capture sensor array respectively.

[0102] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of a holographic projection-based geographical teaching method as described above are implemented.

[0103] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can refer to the partial descriptions of the method embodiments. The device embodiments described above are only illustrative. The components 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 may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the present disclosure solution. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0104] The above embodiments only represent several implementation manners of the embodiments of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the embodiments of the application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the embodiments of the present application, several deformations and improvements can still be made, and these all belong to the protection scope of the embodiments of the present application.

Claims

1. A geography teaching device based on holographic projection, characterized in that: The device includes a data acquisition module, a behavior analysis module, a dynamic adjustment module and a feedback optimization module: The data acquisition module is used to obtain interaction data, perform timestamp marking and sensor data fusion processing on the interaction data, and generate a multi-dimensional behavior data set containing timestamps; the interaction data is used to characterize the interaction behavior between students and holographic projection geography content; The behavior analysis module is used to extract the perspective preference features and analyze the operation habits of the multi-dimensional behavior data set to generate the student learning deviation; The dynamic adjustment module is used to perform parameter adjustment processing on the teaching content mapping relationship of the holographic projection system according to the student learning deviation, and generate a scene change instruction set; The feedback optimization module is used to perform trend analysis on the student response data after the scene change instruction set is executed, and generate an optimized interactive experience parameter set, and the interactive experience parameter set is used to instruct the holographic projection device to generate a holographic image.

2. The device according to claim 1, characterized in that The data acquisition module comprises: A multi-source data acquisition unit is used to acquire the student's visual angle movement trajectory, and quantify the touch operation force based on the pressure sensor to generate an original behavior data stream; A data preprocessing unit, used for performing sensor noise compensation processing on the original behavior data stream to generate a denoised original data set; The timestamp generation unit is used to perform time synchronization processing on the original data set, align the time axis of the original data set based on a time axis alignment algorithm, and generate a multi-dimensional behavior data set marked with a timestamp.

3. The device according to claim 2, characterized in that The behavior analysis module includes: A data cleaning unit, used to perform outlier filtering processing on the multidimensional behavior data set and eliminate noise data through a sliding window algorithm to generate a standardized behavior data set; A viewing angle preference analysis unit, configured to extract a gaze point coordinate sequence from the standardized behavior data set, perform spatial aggregation processing on the gaze point coordinate sequence based on a density clustering algorithm, and generate a viewing angle preference pattern; An operation habit parsing unit, used to perform frequent pattern mining processing on the operation instruction sequence in the standardized behavior data set, match it with the preset behavior template through a dynamic time warping algorithm, and generate an operation habit feature vector; A deviation calculation unit is used to map the perspective preference pattern and the operation habit feature vector to the node space of a pre-trained geographic knowledge graph to obtain a student learning vector, calculate the semantic distance between the student learning vector and the teaching target vector based on a graph neural network, and generate a student learning deviation based on the semantic distance.

4. The device according to claim 3, characterized in that The deviation calculation unit comprises: A difference comparison subunit, used for mapping the perspective preference pattern to the node space of the geographic knowledge graph, calculating the matching degree with the target knowledge node based on the cosine similarity, and generating a first deviation coefficient; A regression analysis subunit, used for performing multivariate linear regression processing on the operation habit feature vector, and generating a second deviation coefficient based on a proportional relationship between a regression residual and a preset threshold value; The weight fusion subunit is used to dynamically allocate weights to the first deviation coefficient and the second deviation coefficient through the entropy weight method to generate the student learning deviation.

5. The device according to claim 1, characterized in that The dynamic adjustment module comprises: An adjustment coefficient generating unit, configured to optimize the teaching content mapping function through a gradient descent algorithm based on the student learning deviation, and generate an initial adjustment coefficient of the scene transformation parameter; An incremental learning unit, used for inputting the initial adjustment coefficient into an incremental learning model, fusing the historical interaction data based on a random weighted average algorithm, and generating a dynamically updated personalized model; An instruction generation unit is used to convert the adjustment coefficient into a scene rendering instruction set according to the spatial interaction interface protocol of the personalized mode matching holographic projection system, and the scene rendering instruction set is used to instruct the holographic projection device to control laser parameters and / or light field angle.

6. The device according to claim 5, characterized in that The incremental learning unit comprises: The history retention evaluation subunit is used to evaluate the historical knowledge retention rate after the model is updated through sliding window cross-validation to generate stability verification results; A weight adjustment subunit, used to dynamically adjust the model weight distribution according to the stability verification result, and generate a coverage threshold parameter, wherein the coverage threshold parameter is used to limit the coverage of new knowledge on historical data; The threshold constraint subunit is used to truncate the parameter update gradient of the incremental learning model based on the coverage threshold parameter to generate a personalized model after catastrophic forgetting is suppressed, and the personalized model is used to update the teaching content mapping relationship of the holographic projection system.

7. The device according to claim 5, characterized in that The feedback optimization module comprises: A response evaluation unit, used to extract the interaction frequency and operation accuracy from the student response data after the scene rendering instruction set is executed, model the response change trend based on a hidden Markov model, and generate a response evaluation value; An instruction optimization unit, configured to partition the feature space by using a C4.5 decision tree algorithm to generate an improved parameter set driven by a classification rule when the response evaluation value is lower than a preset threshold; The real-time update unit is used to input the improved parameter set into the streaming data processing engine, dynamically load the geographic scene resources based on the time-sensitive hash index, and generate an optimized interactive experience parameter set.

8. A geography teaching interactive method based on holographic projection, characterized in that: include: Acquire interaction data, perform timestamp marking and sensor data fusion processing on the interaction data, and generate a multi-dimensional behavior data set containing timestamps; The interaction data is used to characterize the interaction behavior between the students and the holographic projection geography content; Extracting perspective preference features and analyzing operation habits of the multidimensional behavior data set to generate student learning deviation; According to the student learning deviation, parameter adjustment processing is performed on the teaching content mapping relationship of the holographic projection system to generate a scene change instruction set; The student response data after the scene change instruction set is executed is subjected to trend analysis and processing to generate an optimized interactive experience parameter set, wherein the interactive experience parameter set is used to instruct the holographic projection device to generate a holographic image.

9. A geography teaching interactive system based on holographic projection, characterized in that: The system comprises: A holographic projection device, a motion capture sensor array and a terminal processor, wherein the terminal processor is configured to implement the geography teaching interactive method based on holographic projection as described in claim 8; The terminal processor is communicatively connected with the holographic projection device and the motion capture sensor array respectively.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the method according to claim 8 is implemented.

Images