Virtual reality technology-based health knowledge science popularization interaction system establishment method

Abstract

The invention discloses a health knowledge science popularization interaction system establishment method based on a virtual reality technology, and the method comprises the steps: S1, obtaining multi-dimensional data of a science popularization activity process, and constructing a health knowledge science popularization quality dynamic evaluation index system; s2, establishing a health knowledge recommendation engine, and distributing multi-modal science popularization data to students according to the health knowledge recommendation engine; s3, constructing a science popularization scene based on a virtual reality technology; and S4, acquiring popular science learning action and state data of the student, carrying out feature description on the student, and analyzing the learning process and effect data of the student. According to the method, the characteristics of the students can be accurately described, personalized science popularization experience meeting individual requirements is provided, precise measurement of the science popularization activity effect is achieved, the learning quality of audiences in the health knowledge science popularization process is remarkably improved, quantitative analysis is conducted on the science popularization effect, and powerful support is provided for improvement and optimization of science popularization education activities.

Description

Technical Field

[0001] The present invention relates to the technical field of science popularization education system construction, and in particular to a method for establishing a health knowledge science popularization interactive system based on virtual reality technology. Background Art

[0002] Virtual Reality (VR) technology is a computer-based technology that combines relevant scientific and technological advances to create a digital environment that closely resembles a real-world environment within a certain range in terms of vision, hearing, touch, and other aspects. Students use the necessary equipment to interact with objects in the digital environment, gaining the feeling and experience of being in the real environment. Currently, VR technology has been widely used in multiple disciplines and industries, including health education. There are many successful VR health education cases both domestically and internationally, such as disease simulation, surgical training, and mental health intervention. However, VR applications specifically targeting the popularization and dissemination of health knowledge, especially through gamification, are relatively rare.

[0003] Although there are some health science popularization activities on the market, there are still the following difficulties in deeply integrating VR technology, scientifically disseminating health knowledge, motivating students to actively participate and form lasting behavioral changes, etc.: there is no accurate portrait of students, so it is impossible to provide targeted science popularization activity resources that meet the characteristics and requirements of students, so as to improve the quality of science popularization activities; the multi-dimensional dynamic interaction time series data of students continuously generated in science popularization activities are not fully analyzed and utilized, and it is impossible to design efficient and accurate resource recommendation algorithms, and it is impossible to achieve people-centered high-quality science popularization activities in the process of science popularization activities; the resource recommendation algorithm is of low quality and cannot accurately provide suitable science popularization content, resulting in a decline in student experience, hindered knowledge dissemination, and an inability to build students' innovative ability system. Summary of the Invention

[0004] The purpose of this invention is to provide a method for establishing an interactive health knowledge popularization system based on virtual reality technology. This method aims to address the problems encountered in existing health knowledge popularization activities, such as the lack of accurate student profiling, the failure to utilize the multi-dimensional dynamic interaction data generated by students during these activities, and the inability to provide relevant popular science content through recommendation algorithms.

[0005] To achieve the above object, the present invention is implemented through the following technical solutions:

[0006] The present invention provides a method for establishing a health knowledge popularization interactive system based on virtual reality technology, comprising the following steps:

[0007] Step S1: Acquire multidimensional data of the science popularization activity process, construct a dynamic evaluation index system for the quality of health knowledge popularization, and conduct quantitative analysis on the quality of the science popularization activities to be evaluated;

[0008] Step S2: Based on the health knowledge popularization quality dynamic evaluation index system, a health knowledge recommendation engine is established; based on the health knowledge recommendation engine, multimodal popular science data is distributed to students for personalized health popular science; and based on the health knowledge recommendation engine, the health knowledge popular science quality dynamic evaluation index system is updated;

[0009] Step S3: Based on the health knowledge popularization quality dynamic evaluation index system and the health knowledge recommendation engine, as well as virtual reality technology, a popularization scene is constructed to provide an immersive health knowledge popularization learning scenario;

[0010] Step S4: Obtain the student's science popularization learning action and status data, characterize the student, analyze the student's learning process and effect data, obtain analysis results, and update the health knowledge science popularization quality dynamic evaluation index system and the health knowledge recommendation engine based on the analysis results.

[0011] Preferably, the multi-dimensional data of the science popularization activity process includes: dynamic data of human phenotypic characteristics, interactive behavior data and learning stage goal achievement data in the health knowledge science popularization interactive activities.

[0012] Preferably, in step S1, the construction of a dynamic evaluation index system for the quality of health knowledge popularization specifically includes the following steps:

[0013] Step S1.1: Build a health knowledge base, obtain the total number of science popularization activity material types in the health knowledge base, calculate the resource type richness of the science popularization activity to be evaluated based on the total number of science popularization activity material types using a first scoring function, and perform normalization processing. The expression of the first scoring function is:

[0014]

[0015] Among them, KPtypes represents the number of types of science popularization activity materials, and TotalKPtypes represents the total number of science popularization material types;

[0016] Step S1.2: Obtain the total number of popular science knowledge points in the health knowledge base. Based on the total number of popular science knowledge points, calculate the knowledge coverage of the popular science activity to be evaluated using a second scoring function and perform normalization processing. The expression of the second scoring function is:

[0017]

[0018] Among them, KPs is the number of knowledge points of science popularization activities to be evaluated, and TotalKPs is the total number of science popularization knowledge points;

[0019] Step S1.3: Calculate the knowledge difficulty of the science popularization activity to be evaluated by using a third scoring function to evaluate the knowledge difficulty of the science popularization activity to be evaluated. The expression of the third scoring function is:

[0020]

[0021] Among them, i represents the knowledge point, h i Indicates the difficulty score of the i-th knowledge point in the activity, KPT i Indicates the number of times the i-th knowledge point appears in the activity, μ i Represents the attenuation coefficient related to the frequency of occurrence;

[0022] Step S1.4: Based on the health knowledge base, a knowledge graph of health knowledge popularization is constructed, and the knowledge connectivity of the popularization activity to be evaluated is calculated using a fourth scoring function. The expression of the fourth scoring function is:

[0023]

[0024] Among them, A ij represents the weight score of the connection relationship between the i-th and j-th knowledge points; n represents all knowledge points;

[0025] Step S1.5: Obtain the student's personal data, analyze the student's personal data, and define the student's profile to provide personalized science popularization activities for the student.

[0026] Preferably, in step S2, the establishment of a health knowledge recommendation engine and the distribution of multimodal science popularization data to students according to the health knowledge recommendation engine specifically include the following steps:

[0027] Step S2.1: Dynamically optimizing the knowledge graph based on the health knowledge base;

[0028] Step S2.2: Based on the knowledge graph, a recommendation model is established to enhance student learning drive, for dynamically matching popular science content with student status;

[0029] Step S2.3: Use the cross-modal attention mechanism to achieve feature alignment of multi-source data and distribute multimodal science data to students.

[0030] Preferably, in step S2.1, dynamically optimizing the knowledge graph based on the health knowledge base specifically includes:

[0031] Structuring the health knowledge base;

[0032] Defining the hierarchical structure and logical association of knowledge points in the health knowledge base according to different subject classifications;

[0033] Through Graph Neural Networks (GNN), the dynamic optimization of the knowledge graph is realized, and the update function of the knowledge graph is obtained as follows:

[0034]

[0035] Where Δw ij represents the association weight increment between knowledge points i and j, U represents the set of all students, R ui represents the student u’s mastery score of knowledge point i, R′ ui represents the student u’s historical mastery score of knowledge point i, D i represents the global heat value of knowledge point i, α1 represents the dynamic learning rate parameter,

[0036] The update function of the knowledge graph adjusts the connection strength between knowledge points in real time through the interactive behavior data in the multidimensional data;

[0037] In step S2.2, the establishment of a recommendation model that strengthens student learning drive specifically includes:

[0038] defining a student feature vector, and using the student feature vector as a state space;

[0039] Use one-hot encoding to combine into a 128-dimensional action vector and establish the following reward function:

[0040]

[0041] Among them, ΔK represents the improvement value of knowledge mastery, which is calculated by the difference between the pre-test and post-test scores; e -λt represents the attenuation term of interaction duration; λ is a constant representing the memory attenuation coefficient; HRV std represents the standard deviation of heart rate variability, which is used to evaluate the stability of physiological load;

[0042] The recommendation model is trained by the Deep Deterministic Policy Gradient (DDPG) algorithm to dynamically match popular science content with student status.

[0043] Feedback data is fed into the knowledge engine as input. Through the reinforcement learning model in the health knowledge recommendation engine, it can learn how to optimize content recommendation strategies based on students' real-time feedback, dynamically adjust science popularization activities, and improve the quality of science popularization activities.

[0044] In step S2.3, the cross-modal attention mechanism is used to achieve feature alignment of multi-source data, specifically including:

[0045] The attention weight calculation function is defined as:

[0046]

[0047] Among them, Q is the student's real-time status query vector, K / V is the key-value pair matrix of multimodal resources;

[0048] Based on the attention weight calculation function, a personalized composite content package is generated, and the correlation analysis between the trainees' physiological indicators and knowledge mastery is displayed in real time through an information panel in a virtual reality space.

[0049] Preferably, in step S3, the construction of a science popularization scenario based on the health knowledge popularization quality dynamic evaluation index system and the health knowledge recommendation engine, as well as virtual reality technology, specifically includes:

[0050] Step S3.1: Based on virtual reality technology, convert the knowledge graph into a three-dimensional virtual space;

[0051] Step S3.2: Constructing a "learning-practice-assessment" interactive process in the three-dimensional virtual space to monitor the trainees' operation accuracy and response speed in real time;

[0052] Step S3.3: Enhance the learning experience through cross-modal technology and establish a multi-sensory immersion enhancement system.

[0053] Preferably, in step S4, obtaining the student's science popularization learning action and status data, characterizing the student, and analyzing the student's learning process and effect data specifically includes:

[0054] Step S4.1: Establish an evaluation indicator system for students’ learning process and effect data;

[0055] Step S4.2: Obtain various sensor data during the science popularization activity and quantify the students' attention to the target objects in the game scene during the interaction.

[0056] Preferably, in step S4.1, the evaluation index system of the student learning process and effect data includes: knowledge acquisition evaluation, interactive activity evaluation, activity result quality evaluation, student emotion evaluation, sustained attention evaluation and task achievement evaluation.

[0057] Preferably, the student knowledge acquisition evaluation includes: evaluating the student's knowledge acquisition degree according to the feedback result of human-computer interaction, and the fifth scoring function used for the student knowledge acquisition evaluation is:

[0058]

[0059] Among them, hi Indicates the difficulty of the i-th knowledge point in the activity, KPscores represents the student's ability level in the current knowledge area. i Indicates the student's test score on the i-th knowledge point;

[0060] The interactive activity evaluation includes: evaluating the activity of the trainees during the interaction process based on the analysis of the interactive feedback data. The sixth scoring function used for the interactive activity evaluation is:

[0061]

[0062] Among them, HandlerAction i Indicates the number of times the student uses the handle to perform effective operations on the i-th knowledge point in the activity; VisualFocus i It represents the number of times the learner’s visual focus is on the correct target during the i-th knowledge point activity;

[0063] The quality evaluation of the science popularization activity results includes: evaluating the improvement of students' comprehensive abilities through participating in the science popularization activities based on the feedback from the science popularization activities. The seventh scoring function used for the quality evaluation of the science popularization activity results is:

[0064]

[0065] Among them, S post,i , S pre,i They represent the students’ post-test and pre-test scores on the i-th knowledge point, w i Represents the weight of the knowledge point, h i Knowledge point difficulty coefficient, α2 and β1 both represent difficulty adjustment coefficients, A app represents the application ability score, γ1 represents the application ability weight coefficient, λ represents the memory decay coefficient, and t represents the time interval between the test and the end of the activity.

[0066] Preferably, in step S4.2, the quantification of the trainee's attention to the target object in the game scene during the interaction is achieved through a final attention score, thereby achieving a comprehensive quantification of the trainee's attention to the target object in the game scene during the interaction. The expression of the final attention score A is:

[0067] A=α3×H+β2×C+γ2×S

[0068] Among them, α3, β2, and γ2 represent weight coefficients; H represents the head posture index, which is used to measure whether the user's head is facing the target object; C represents the handle interaction index, which is used to measure whether the user's operation is directed at the target object; S represents the spatial relationship index, which is used to measure whether the user is actively approaching the target object.

[0069] Compared with the prior art, the present invention has the following beneficial effects:

[0070] The present invention provides a method for establishing a health knowledge popularization interactive system based on virtual reality technology. Through the application of virtual reality technology, it can accurately portray the characteristics of students, provide personalized popularization experience that meets individual needs, achieve accurate measurement of the effectiveness of popularization activities, significantly improve the learning quality of the audience in the process of health knowledge popularization, and quantify the popularization effect, providing strong support for the improvement and optimization of popularization education activities. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for the description. Obviously, the drawings described below are one embodiment of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort:

[0072] Figure 1 A flowchart of a method for establishing a health knowledge popularization interactive system based on virtual reality technology provided by one embodiment of the present invention;

[0073] Figure 2 This is an architectural diagram of a health knowledge popularization interactive system based on virtual reality technology provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0074] The following is combined with Figure 1-2 The method for establishing a health knowledge popularization interactive system based on virtual reality technology proposed by the present invention is further described in detail with reference to the specific implementation methods. According to the following description, the advantages and features of the present invention will be clearer. It should be noted that the drawings are in a very simplified form and use non-precise proportions, which are only used to conveniently and clearly assist in explaining the purpose of the implementation methods of the present invention. In order to make the purposes, features and advantages of the present invention more obvious and easy to understand, please refer to the drawings. It should be noted that the structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the implementation conditions of the present invention, so they have no technical substantive significance. Any modification of the structure, change in the proportional relationship or adjustment of the size should still fall within the scope of the technical content disclosed by the present invention without affecting the efficacy and purpose that can be achieved by the present invention.

[0075] In view of the problems that existing technologies in health science popularization activities do not accurately profile students, do not utilize the multi-dimensional dynamic interaction data of students generated in science popularization activities, and cannot provide suitable science popularization content through recommendation algorithms.

[0076] refer to Figure 1 and Figure 2 As shown, this embodiment provides a method for establishing a health knowledge popularization interactive system based on virtual reality technology, including the following steps:

[0077] Step S1: Acquire multidimensional data of the science popularization activity process, construct a dynamic evaluation index system for the quality of health knowledge science popularization, and conduct quantitative analysis on the quality of the science popularization activity to be evaluated. The multidimensional data of the science popularization activity process includes: dynamic data of human phenotypic characteristics in the health knowledge science popularization interactive activities, interactive behavior data, and learning stage goal achievement data. The interactive behavior data includes: handle button action information, handle trajectory data, eye movement data, heart rate and blood pressure, etc. The construction of the dynamic evaluation index system for the quality of health knowledge science popularization specifically includes the following steps:

[0078] Step S1.1: Build a health knowledge base, obtain the total number of science popularization activity material types in the health knowledge base, calculate the resource type richness of the science popularization activity to be evaluated based on the total number of science popularization activity material types using a first scoring function, and perform normalization processing. The expression of the first scoring function is:

[0079]

[0080] Among them, KPtypes represents the number of types of science popularization activity materials, and TotalKPtypes represents the total number of science popularization material types.

[0081] Step S1.2: Obtain the total number of popular science knowledge points in the health knowledge base. Based on the total number of popular science knowledge points, calculate the knowledge coverage of the popular science activity to be evaluated using a second scoring function and perform normalization processing. The expression of the second scoring function is:

[0082]

[0083] Among them, KPs is the number of knowledge points of science popularization activities to be evaluated, and TotalKPs is the total number of science popularization knowledge points.

[0084] Step S1.3: Calculate the knowledge difficulty of the science popularization activity to be evaluated by using a third scoring function to evaluate the knowledge difficulty of the science popularization activity to be evaluated. The expression of the third scoring function is:

[0085]

[0086] Where i represents the knowledge point of the science popularization activity to be evaluated, h i KPT represents the difficulty score of the i-th knowledge point in the popular science activity to be evaluated. iIndicates the number of times the i-th knowledge point appears in the activity, μ i Represents the attenuation coefficient related to the frequency of occurrence.

[0087] Step S1.4: Based on the health knowledge base, a knowledge graph of health knowledge popularization is constructed, and the knowledge connectivity of the popularization activity to be evaluated is calculated using a fourth scoring function. The expression of the fourth scoring function is:

[0088]

[0089] Among them, A ij represents the weight score of the connection relationship between the i-th and j-th knowledge points, and n represents all knowledge points.

[0090] Step S1.5: Obtain the student's personal data, analyze the student's personal data, and define the student's profile to provide personalized science popularization activities for the student.

[0091] The collection of students' personal information is introduced with a specific embodiment: a questionnaire is designed before the science popularization activity to collect seven categories of students' personal status information for the purpose of describing the students' static characteristics. By analyzing the students' personal related data, a character portrait is created for them to be used for providing personalized science popularization activities to the students in the future. The seven categories of status information are shown in Table 1 below.

[0092] Table 1

[0093]

[0094] Step S2: Based on the health knowledge popularization quality dynamic evaluation index system, a health knowledge recommendation engine is established. Based on the health knowledge recommendation engine, multimodal popular science data is distributed to students for personalized health popularization. Based on the health knowledge recommendation engine, the health knowledge popularization quality dynamic evaluation index system is updated. Specifically, the following steps are included:

[0095] Step S2.1: Dynamically optimize the knowledge graph based on the health knowledge base, which specifically includes: structuring the health knowledge base; defining the hierarchical architecture and logical associations of knowledge points in the health knowledge base according to different subject classifications; and dynamically optimizing the knowledge graph through a graph neural network. The update function of the knowledge graph is:

[0096]

[0097] Where Δw ij represents the association weight increment between knowledge points i and j, R ui represents the mastery score of student u on knowledge point i, U represents the set of all students, R′ uirepresents the student u’s historical mastery score of knowledge point i, D i represents the global heat value of knowledge point i, and α1 represents the dynamic learning rate parameter.

[0098] The update function of the knowledge graph adjusts the connection strength between knowledge points in the knowledge graph in real time through the interactive behavior data in the multidimensional data.

[0099] Step S2.2: Based on the knowledge graph, establish a recommendation model that strengthens student learning drive, which is used for dynamic matching of popular science content and student status. The establishment of the recommendation model that strengthens student learning drive specifically includes: defining a student feature vector, using the student feature vector as the state space, and defining the student feature vector S t =[static portrait, real-time heart rate, eye movement focus entropy, knowledge gap matrix], where the knowledge gap matrix is generated by comparing the student test results with the knowledge graph node weights.

[0100] Use one-hot encoding to combine into a 128-dimensional action vector and establish the following reward function:

[0101] R t =0.6×ΔK+0.3×e -λt +0.1×HRV std (6)

[0102] Among them, ΔK represents the improvement value of knowledge mastery, which is calculated by the difference between the pre-test and post-test scores; e -λt represents the attenuation term of interaction duration; λ represents the memory attenuation coefficient, λ is a constant; HRV std represents the standard deviation of heart rate variability, which is used to evaluate the stability of physiological load.

[0103] The recommendation model is trained through a deep deterministic policy gradient algorithm to dynamically match the popular science content with the student status. Action Space A t It includes three parameters: multimodal resource type (e.g. video / game / graphics), knowledge difficulty level, and interaction complexity. t It is a set of specific actions that the model can take at time step t (such as when the student is learning for the tth time), and its function is similar to an "operating handle". The system is based on the student's feature vector S in the current state space. t Select Action A t (such as "pushing video + difficulty L2 + high interaction complexity"), by executing A t It affects students’ learning path and triggers environmental feedback (reward function calculation) to form a closed-loop optimization.

[0104] Step S2.3: Use the cross-modal attention mechanism to achieve feature alignment of multi-source data and distribute multi-modal science popularization data to students. Define the attention weight calculation function ATT(Q,K,V) as:

[0105]

[0106] Here, Q is the student's real-time status query vector, and K / V is the key-value matrix of multimodal resources. The key matrix K is used to store the "index features" of multimodal resources (such as semantic tags, modality types, and knowledge point relevance), and is used to perform similarity matching with the student's real-time status query vector Q to determine the resources relevant to the current needs. The value matrix V stores the "content features" of multimodal resources (such as the visual features of video frames, the geometric parameters of 3D models, and the semantic embedding of text), directly carrying the knowledge entities to be distributed.

[0107] Based on the attention weight calculation function, a personalized composite content package is generated, and a real-time analysis of the correlation between the student's physiological indicators and knowledge mastery is displayed through an information panel in the virtual reality space. For example, for students with weak cardiovascular knowledge, a dynamic 3D heart model demonstration, an interactive blood pressure adjustment game, and a customized dietary guidance video are integrated, and a real-time analysis of the correlation between the student's physiological indicators and knowledge mastery is displayed through an information panel in the virtual reality space.

[0108] Step S3: Based on the dynamic evaluation index system for the quality of health knowledge popular science, the health knowledge recommendation engine, and virtual reality technology, a science popularization scene is constructed to provide an immersive health knowledge popular science learning scenario. The immersive health knowledge popular science learning scenario uses a virtual reality helmet display to design a science popularization learning process in a virtual reality space, including a science popularization learning process including learning-practice-evaluation links, constructs an immersive science popularization scene, and provides video information display, science popularization interactive games, and other different types of human-computer interaction modes. Specifically, the following steps are included:

[0109] Step S3.1: Based on virtual reality technology, the knowledge graph is converted into a three-dimensional virtual space. Each knowledge point corresponds to a specific theme scene. In one embodiment, the spatial distance between scenes is inversely proportional to the strength of knowledge association. For example, the knowledge points of hypertension and stroke are closely related, and the distance between the corresponding scene entrances is set to within 3 meters; while the knowledge points of hypertension and fracture rehabilitation are weakly related, and the scene distance is extended to more than 10 meters. After the trainee wears the virtual reality device, he can trigger the scene switch through handle ray interaction or gaze, and the system automatically optimizes the path navigation according to the knowledge graph weight.

[0110] Step S3.2: Construct a "learning-practice-evaluation" three-in-one interactive process in the three-dimensional virtual space to monitor the trainees' operation accuracy and response speed in real time. In one embodiment, if the task completion rate is higher than 85% for three consecutive times, the challenge level will be automatically increased (such as shortening the task time limit, adding interference items); if it is detected that the trainee's heart rate is abnormally high or the error rate increases sharply, the auxiliary guidance mode will be activated (such as highlighting key operation areas and reducing interaction complexity). Knowledge mastery is comprehensively scored through multiple-choice questions and practical task completion, of which the practical score accounts for 60%, focusing on the ability to apply knowledge.

[0111] Step S3.3: Enhance the learning experience through cross-modal technology and establish a multi-sensory immersion enhancement system. The cross-modal technology-enhanced learning experience specifically includes: (1) Visual enhancement: Using layered rendering technology, a semi-transparent pathological state contrast layer is superimposed on the normal anatomical structure; (2) Auditory feedback: The background sound effects change dynamically with the student's cognitive state. When the system detects that the student's concentration has decreased (scattered eye movement trajectory, operation delay), it automatically switches to a fast-paced prompt sound; when the knowledge test is correct, a motivational sound effect is played; (3) Tactile interaction: Simulate operation feedback through handle vibration. For example, when the cardiopulmonary resuscitation compression depth is completed correctly, the handle will produce a short vibration; if the compression position is offset, a continuous vibration warning is triggered.

[0112] Step S4: Obtain the student's science popularization learning action and status data, characterize the student, analyze the student's learning process and effect data, obtain analysis results, and update the health knowledge science popularization quality dynamic evaluation index system and health knowledge recommendation engine based on the analysis results to provide closed-loop feedback for the health knowledge science popularization interactive system.

[0113] Through the sensors built into the virtual reality helmet and the necessary sensor equipment provided to students during the popular science process, the behavioral data of students during the popular science process is continuously collected. By analyzing the behavioral data, the current status indicators of the students are extracted and feedback data is provided to the health knowledge recommendation engine. The feedback data is sent as input to the health knowledge recommendation engine. In the health knowledge recommendation engine, the recommendation model is trained through the deep deterministic policy gradient algorithm. It can learn how to optimize the content recommendation strategy based on the real-time feedback of students, dynamically adjust the popular science activities, and improve the quality of popular science activities. Specifically including:

[0114] Step S4.1: Establish an evaluation indicator system for students’ learning process and effect data.

[0115] The evaluation index system for student learning process and effect data includes: knowledge acquisition evaluation, interactive activity evaluation, activity result quality evaluation, student emotion evaluation, sustained attention evaluation and task achievement evaluation.

[0116] The evaluation of the student's knowledge acquisition degree includes: evaluating the student's knowledge acquisition degree according to the feedback result of human-computer interaction. The fifth scoring function used for the evaluation of the student's knowledge acquisition degree is:

[0117]

[0118] Among them, h i Indicates the difficulty of the i-th knowledge point in the activity, KPscores represents the student's ability level in the current knowledge area. i Indicates the student's test score on the i-th knowledge point.

[0119] The interactive activity evaluation includes: evaluating the activity of the trainees during the interaction process based on the analysis of the interactive feedback data. The sixth scoring function used for the interactive activity evaluation is:

[0120]

[0121] Among them, HandlerAction i Indicates the number of times the student uses the handle to perform effective operations on the i-th knowledge point in the activity; VisualFocus i It indicates the number of times the learner’s visual focus is on the correct target during the i-th knowledge point activity.

[0122] The quality evaluation of the science popularization activity results includes: evaluating the improvement of students' comprehensive abilities through participating in the science popularization activities based on the feedback from the science popularization activities. The seventh scoring function used for the quality evaluation of the science popularization activity results is:

[0123]

[0124] Among them, S post,i , S pre,i They represent the students’ scores on the post-test and pre-test of the i-th knowledge point (standardized to 0-1 points), w i Represents the weight of the knowledge point, (determined by expert evaluation or entropy weight method, satisfying ∑w i =1), h i represents the difficulty of the i-th knowledge point in the activity (based on the discrimination parameter of item response theory (IRT), or inferred from the test results), α2 and β1 both represent difficulty adjustment coefficients (preferably α2 = 0.7, β1 = 0.3, to avoid high-difficulty knowledge points completely inhibiting gain), A apprepresents the application ability score (obtained through behavioral assessments such as scenario simulation and open-ended problem solving, standardized from 0 to 1 points), γ1 represents the application ability weight coefficient (preferably γ1 = 0.4, adjusted according to the activity goals), λ represents the memory decay coefficient (preferably, λ = 0.05 / day), and t represents the time interval (in days) between the test and the end of the activity.

[0125] The student emotion assessment includes: evaluating emotions according to the following five dimensions: polarity (positive, negative), intensity (weak, strong), duration (short-term, long-lasting), complexity and functional orientation.

[0126] The sustained attention evaluation includes: evaluating according to the ratio of effective human-computer interaction time, with a score of 1 to 5.

[0127] The task achievement evaluation includes: whether the knowledge learning requirements of the science popularization activity task are achieved: the task completion degree is evaluated according to 0 to 5.

[0128] Preferably, step S4.2: obtaining various sensor data during the science popularization activity, and quantifying the students' attention to the target object in the game scene during the interaction process.

[0129] In step S4.2, the quantification of the trainee's attention to the target object in the game scene during the interaction is achieved through the final attention score, which is a comprehensive quantification of the trainee's attention to the target object in the game scene during the interaction. The expression of the final attention score A is:

[0130] A=α3γH+β2×C+γ2×S (11)

[0131] Among them, α3, β2, and γ2 represent weight coefficients; H represents the head posture index (directly taking the time proportion 0 to 1), which is used to measure whether the user's head is facing the target object. The effective gaze time is the cumulative time when the angle θ between the head direction and the target object direction is ≤ 30°. C represents the handle interaction index. Used to measure whether the user's operation is directed at the target object. The average operation deviation is the average Euclidean distance between the handle pointing / operation position and the center point of the target object (needs to be calculated in the game coordinate system). The number of effective operations is the number of effective operations on the target object during the game. max is the maximum possible operation deviation in the scene (e.g., the length of the scene diagonal), ensuring that C∈(0,1](0,1). S represents the spatial relationship index, which is used to measure whether the user actively approaches the target object.

[0132] The analysis result is the comprehensive quantification of the student's attention to the target object in the game scene during the interaction process. After completing the cross-modal attention score calculation, dynamic fusion and precise distribution of multimodal data are achieved based on the attention weight. By applying the calculated final attention score to multiple content features such as videos, 3D models, and interactive scripts, a composite science popularization content package adapted to the student's cognitive state is generated. Based on the analysis results, the multidimensional data of the science popularization activity process is updated, and the dynamic evaluation indicator system for the quality of health knowledge popularization is updated.

[0133] According to the analysis results, the fused student feature vector S t , input the deep deterministic policy gradient algorithm to train the recommendation model to trigger real-time policy updates, automatically adjust the knowledge density distribution (such as increasing the related content coverage of knowledge points in high-attention areas by 30%) and reset the interactive task difficulty parameters (dynamically compressing the task time limit based on the stability of attention focus) to achieve the update of the health knowledge recommendation engine.

[0134] These process data will drive the incremental update of the association weights of the knowledge graph and the reconstruction of the VR scene topology, thereby building a closed-loop response chain from attention calculation to scene optimization, achieving sub-second dynamic matching of popular science content and students' cognitive status (response delay <800ms), and significantly improving knowledge transfer efficiency and long-term retention rate.

[0135] In summary, this embodiment proposes a method for establishing a health knowledge popularization interactive system based on virtual reality technology. The method includes establishing a GNN-based dynamic adjustment model for knowledge association weights, developing a multimodal attention fusion mechanism, and constructing a VR scene topology with spatial logical mapping to convert knowledge association strength into three-dimensional spatial distance parameters. This virtual reality-based health knowledge popularization interactive framework achieves precision and intelligent popular science education through the integration of four-dimensional technologies. The system development method includes a dynamic quality assessment system (building a quantitative assessment model based on seven dimensions, including knowledge coverage, resource richness, and difficulty gradients), a reinforcement learning-driven knowledge recommendation engine (integrating a knowledge graph with a deep deterministic policy gradient algorithm to achieve personalized content distribution), a multimodal VR immersive scene (building a "learning-practice-evaluation" interactive closed loop and a cross-sensory feedback system), and a student profile system (integrating seven types of static features with 12-dimensional real-time physiological and behavioral data). Biometric features are incorporated into the recommendation algorithm's reward function to achieve dynamic evolution of the knowledge graph. This system improves the accuracy of popular science content matching and knowledge retention, providing a quantifiable and iterative intelligent solution for health science popularization education. Through the application of virtual reality technology, we can accurately portray the characteristics of students, provide personalized science popularization experiences that meet individual needs, accurately measure the effectiveness of science popularization activities, significantly improve the audience's learning quality in the process of health knowledge popularization, and conduct quantitative analysis of science popularization effects, providing strong support for the improvement and optimization of science popularization education activities.

[0136] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0137] It should be noted that the devices and methods disclosed in the embodiments of this document may also be implemented in other ways. The device embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of the devices, methods, and computer program products according to the various embodiments of this document. In this regard, each box in the flowchart or block diagram may represent a module, program, or portion of code, wherein the module, program segment, or portion of code contains one or more executable instructions for implementing a specified logical function, and the module, program segment, or portion of code contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the boxes may also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes may actually be executed substantially in parallel, or they may sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system for performing the specified function or action, or may be implemented by a combination of dedicated hardware and computer instructions.

[0138] In addition, the functional modules in the various embodiments of this document may be integrated together to form an independent part, or each module may exist independently, or two or more modules may be integrated to form an independent part.

[0139] Although the present invention has been described in detail through the above preferred embodiments, it should be understood that the above description is not intended to limit the present invention. After reading the above description, various modifications and substitutions of the present invention will become apparent to those skilled in the art. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A method for establishing a health knowledge popularization interactive system based on virtual reality technology, characterized in that: The following steps are involved: Step S1: Acquire multidimensional data of the science popularization activity process, construct a dynamic evaluation index system for the quality of health knowledge popularization, and conduct quantitative analysis on the quality of the science popularization activities to be evaluated; Step S2: Based on the health knowledge popularization quality dynamic evaluation index system, a health knowledge recommendation engine is established; based on the health knowledge recommendation engine, multimodal popular science data is distributed to students for personalized health popular science; and based on the health knowledge recommendation engine, the health knowledge popular science quality dynamic evaluation index system is updated; Step S3: constructing a science popularization scenario based on the health knowledge popularization quality dynamic evaluation index system and the health knowledge recommendation engine, as well as virtual reality technology; Step S4: Obtain the student's science popularization learning action and status data, characterize the student, analyze the student's learning process and effect data, obtain analysis results, and update the health knowledge science popularization quality dynamic evaluation index system and the health knowledge recommendation engine based on the analysis results.

2. The method for establishing a health knowledge popularization interactive system based on virtual reality technology according to claim 1, characterized in that: The multi-dimensional data of the science popularization activity process includes: dynamic data of human phenotypic characteristics, interactive behavior data and learning stage goal achievement data in the health knowledge science popularization interactive activities.

3. The method for establishing a health knowledge popularization interactive system based on virtual reality technology according to claim 1, characterized in that: In step S1, the construction of a dynamic evaluation index system for the quality of health knowledge popularization specifically includes the following steps: Step S1.1: Build a health knowledge base, obtain the total number of science popularization activity material types in the health knowledge base, calculate the resource type richness of the science popularization activity to be evaluated based on the total number of science popularization activity material types using a first scoring function, and perform normalization processing. The expression of the first scoring function is: Among them, KPtypes represents the number of types of science popularization activity materials, and TotalKPtypes represents the total number of science popularization material types; Step S1.2: Obtain the total number of popular science knowledge points in the health knowledge base. Based on the total number of popular science knowledge points, calculate the knowledge coverage of the popular science activity to be evaluated using a second scoring function and perform normalization processing. The expression of the second scoring function is: Among them, KPs is the number of knowledge points of science popularization activities to be evaluated, and TotalKPs is the total number of science popularization knowledge points; Step S1.3: Calculate the knowledge difficulty of the science popularization activity to be evaluated by using a third scoring function to evaluate the knowledge difficulty of the science popularization activity to be evaluated. The expression of the third scoring function is: Among them, i represents the knowledge point, h i Indicates the difficulty score of the i-th knowledge point in the activity, KPT i Indicates the number of times the i-th knowledge point appears in the activity, μ i Represents the attenuation coefficient related to the frequency of occurrence; Step S1.4: Based on the health knowledge base, a knowledge graph of health knowledge popularization is constructed, and the knowledge connectivity of the popularization activity to be evaluated is calculated using a fourth scoring function. The expression of the fourth scoring function is: Among them, A ij represents the weight score of the connection relationship between the i-th and j-th knowledge points; n represents all knowledge points; Step S1.5: Obtain the student's personal data, analyze the student's personal data, and define the student's profile to provide personalized science popularization activities for the student.

4. The method for establishing a health knowledge popularization interactive system based on virtual reality technology according to claim 3, characterized in that: In step S2, the establishment of a health knowledge recommendation engine and the distribution of multimodal science popularization data to students based on the health knowledge recommendation engine specifically include the following steps: Step S2.1: Dynamically optimizing the knowledge graph based on the health knowledge base; Step S2.2: Based on the knowledge graph, a recommendation model is established to enhance student learning drive, for dynamically matching popular science content with student status; Step S2.3: Use the cross-modal attention mechanism to achieve feature alignment of multi-source data and distribute multimodal science data to students.

5. The method for establishing a health knowledge popularization interactive system based on virtual reality technology according to claim 4, characterized in that: In step S2.1, the dynamic optimization of the knowledge graph based on the health knowledge base specifically includes: Structuring the health knowledge base; Defining the hierarchical structure and logical association of knowledge points in the health knowledge base according to different subject classifications; Through the graph neural network, the dynamic optimization of the knowledge graph is realized, and the update function of the knowledge graph is obtained as follows: Where Δw ij represents the association weight increment between knowledge points i and j, U represents the set of all students, R ui represents the student u’s mastery score of knowledge point i, R′ ui represents the student u’s historical mastery score of knowledge point i, D i represents the global heat value of knowledge point i, α1 represents the dynamic learning rate parameter, The updating function of the knowledge graph adjusts the connection strength between knowledge points in real time through the interactive behavior data in the multidimensional data; In step S2.2, the establishment of a recommendation model that strengthens student learning drive specifically includes: defining a student feature vector, and using the student feature vector as a state space; Use one-hot encoding to combine into a 128-dimensional action vector and establish the following reward function: R t =0.6×ΔK+0.3×e -λt +0.1×HR V std Among them, ΔK represents the improvement value of knowledge mastery, which is calculated by the difference between the pre-test and post-test scores; e -λt represents the attenuation term of interaction duration; λ is a constant representing the memory attenuation coefficient; HRV std represents the standard deviation of heart rate variability, which is used to evaluate the stability of physiological load; The recommendation model is trained through a deep deterministic policy gradient algorithm to dynamically match popular science content with student status; In step S2.3, the cross-modal attention mechanism is used to achieve feature alignment of multi-source data, specifically including: The attention weight calculation function is defined as: Among them, Q is the student's real-time status query vector, K / V is the key-value pair matrix of multimodal resources; Based on the attention weight calculation function, a personalized composite content package is generated, and the correlation analysis between the trainees' physiological indicators and knowledge mastery is displayed in real time through an information panel in a virtual reality space.

6. The method for establishing a health knowledge popularization interactive system based on virtual reality technology according to claim 1, characterized in that: In step S3, the construction of a science popularization scenario based on the health knowledge popularization quality dynamic evaluation index system and the health knowledge recommendation engine, as well as virtual reality technology, specifically includes: Step S3.1: Based on virtual reality technology, convert the knowledge graph into a three-dimensional virtual space; Step S3.2: Based on the dynamic evaluation index system for the quality of health knowledge popularization, a "learning-practice-evaluation" interactive process is constructed in the three-dimensional virtual space to monitor the students' operation accuracy and response speed in real time; Step S3.3: Enhance the learning experience through cross-modal technology and establish a multi-sensory immersion enhancement system.

7. The method for establishing a health knowledge popularization interactive system based on virtual reality technology according to claim 1, characterized in that: In step S4, the steps of obtaining the student's science popularization learning action and status data, characterizing the student, and analyzing the student's learning process and effect data specifically include: Step S4.1: Establish an evaluation indicator system for students’ learning process and effect data; Step S4.2: Obtain various sensor data during the science popularization activity and quantify the students' attention to the target objects in the game scene during the interaction.

8. The method for establishing a health knowledge popularization interactive system based on virtual reality technology according to claim 7, characterized in that: In step S4.1, the evaluation index system of the student learning process and effect data includes: knowledge acquisition evaluation, interactive activity evaluation, activity result quality evaluation, student emotion evaluation, sustained attention evaluation and task achievement evaluation.

9. The method for establishing a health knowledge popularization interactive system based on virtual reality technology according to claim 7, characterized in that: The evaluation of the student's knowledge acquisition degree includes: evaluating the student's knowledge acquisition degree according to the feedback result of human-computer interaction. The fifth scoring function used for the evaluation of the student's knowledge acquisition degree is: Among them, h i Indicates the difficulty of the i-th knowledge point in the activity, KPscores represents the student's ability level in the current knowledge area. i Indicates the student's test score on the i-th knowledge point; The interactive activity evaluation includes: evaluating the activity of the trainees during the interaction process based on the analysis of the interactive feedback data. The sixth scoring function used for the interactive activity evaluation is: Among them, HandlerAction i Indicates the number of times the student uses the handle to perform effective operations on the i-th knowledge point in the activity; VisualFocus i It represents the number of times the learner’s visual focus is on the correct target during the i-th knowledge point activity; The quality evaluation of the science popularization activity results includes: evaluating the improvement of students' comprehensive abilities through participating in the science popularization activities based on the feedback from the science popularization activities. The seventh scoring function used for the quality evaluation of the science popularization activity results is: Among them, S post,i , S pre,i They represent the students’ post-test and pre-test scores on the i-th knowledge point, w i Represents the weight of the knowledge point, h i Knowledge point difficulty coefficient, α2 and β1 both represent difficulty adjustment coefficients, A app represents the application ability score, γ1 represents the application ability weight coefficient, λ represents the memory decay coefficient, and t represents the time interval between the test and the end of the activity.

10. The method for establishing a health knowledge popularization interactive system based on virtual reality technology according to claim 7, characterized in that: In step S4.2, the quantification of the trainee's attention to the target object in the game scene during the interaction is achieved through the final attention score, which is a comprehensive quantification of the trainee's attention to the target object in the game scene during the interaction. The expression of the final attention score A is: A=α3×H+β2×C+γ2×S Among them, α3, β2, and γ2 represent weight coefficients; H represents the head posture index, which is used to measure whether the user's head is facing the target object; C represents the handle interaction index, which is used to measure whether the user's operation is directed at the target object; S represents the spatial relationship index, which is used to measure whether the user is actively approaching the target object.

Images