Intelligent education management system based on multi-user cooperation

Abstract

The invention belongs to the technical field of education management, and particularly relates to an intelligent education management system based on multi-user collaboration, which comprises a user terminal group, a collaboration management server, a cloud database, an intelligent analysis engine and a visual interaction interface, the user terminal group is distributed access equipment composed of a teacher terminal, a student terminal and an administrator terminal; the collaborative management server is configured with a dynamic authority distribution engine and a real-time communication coordination module; a user portrait library and a teaching resource pool are stored in the cloud database; based on setting of a user terminal group, multi-user collaboration and real-time communication can be realized, space-time limitation can be broken, collaboration efficiency can be improved, and aiming at the situation that the requirement of dynamic permission change cannot be met due to static permission allocation in the prior art, the dynamic permission allocation engine can be utilized to realize dynamic permission allocation according to actual conditions through a collaboration management server. Authority is dynamically allocated according to requirements and role changes, and the requirements of online collaborative learning are met.

Description

Technical Field

[0001] The present invention belongs to the technical field of education management, and specifically is a smart education management system based on multi-user collaboration. Background Art

[0002] With the rapid development of information technology, the application scope of smart education management system, as an important support for modern teaching, is constantly expanding. However, in practical applications, the existing smart education management system still faces many technical challenges, which to some extent limit the improvement of teaching efficiency and the guarantee of teaching quality.

[0003] In the current existing technology, the system mostly adopts a static permission allocation method, that is, fixed permissions are pre-set according to user roles (such as teachers, students, and administrators). By defining different roles and corresponding permissions, users are assigned to corresponding roles to achieve permission control. When teaching needs change, teachers or administrators need to manually adjust student permissions; In the traditional teaching environment, the communication and collaboration between teachers and students are often restricted by time and space, resulting in low efficiency. Specifically, the current system mostly adopts a static permission allocation mechanism, that is, fixed permissions are pre-set according to user roles (such as teachers, students, administrators). This mechanism cannot flexibly adapt to dynamically changing teaching scenarios, resulting in teachers spending a lot of time and energy to manually adjust student permissions. Static permission allocation not only increases the workload of teachers, but may also affect students' learning progress and experience due to untimely permission adjustments. At the same time, the student interaction methods in the existing system are often limited to text chats or simple voice exchanges, lacking multimodal and immersive interaction methods. The single nature of interaction limits the deep collaboration and knowledge sharing between students, reducing the effectiveness and fun of collaborative learning.

[0004] To this end, the present invention provides a smart education management system based on multi-user collaboration. Summary of the invention

[0005] In order to make up for the deficiencies of the prior art, at least one technical problem raised in the background technology is solved.

[0006] The technical solution adopted by the present invention to solve the technical problem is: a smart education management system based on multi-user collaboration described in the present invention comprises: User terminal group, a distributed access device consisting of teacher terminals, student terminals and administrator terminals; A collaborative management server, configured with a dynamic rights allocation engine and a real-time communication coordination module; The cloud database stores user portraits and teaching resource pools; An intelligent analysis engine, used for optimizing collaborative learning, connected to the collaborative management server and configured with: Group behavior modeling unit, used to train GNN models and optimize configuration schemes; A knowledge graph construction unit, used to construct a knowledge graph; Personalized recommendation unit, used to provide personalized learning paths; Visual interactive interface, providing a collaborative operation platform for switching between multiple views.

[0007] Preferably, the dynamic permission allocation engine is used to update the permission level according to the user role, learning behavior and real-time environment; the dynamic permission allocation engine includes: A role identification module, used to resolve user identity identifiers; Context-aware module, collecting learning behavior and network environment in real time; The permission generation module dynamically adjusts the operation permission level based on fuzzy logic algorithm; The method for the dynamic permission allocation engine to dynamically update the permission level is: Parse and obtain user roles based on the role identification module. If the user role belongs to the student role set, the first permission set or the second permission set is assigned; If the user role does not belong to the student role set, the third permission set is assigned; Then collect real-time data based on the context awareness module; Finally, the user's permission level is dynamically updated based on the fuzzy logic algorithm; Preferably, the method for dynamically adjusting the user's authority level based on the fuzzy logic algorithm is: Obtain the user's learning behavior and network environment; Calculate learning behavior scores based on real-time data , Network environment score ; According to the formula: ; in, Score the frequency of student interaction. The maximum score for student interaction frequency and the score for test performance are The highest score for the test. , is the weight factor; ; is the real-time network speed, is the maximum network speed; is the total number of monitoring times during the monitoring period, is the number of network fluctuations; Scoring based on acquired learning behavior , Network environment score Calculate the overall score: ; Then, based on the fuzzy logic algorithm and the obtained comprehensive score S, the permission level of the user role is output. : ; ; ; ; in, is the membership function of the low authority level, is the membership function of the medium authority level, is the membership function of the high authority level; , , , is the turning point of the membership function, , is the turning point of the membership function of the low authority level; , , , is the turning point of the membership function of the medium authority level, , is the turning point of the membership function of the high authority level; Output the permission level corresponding to the user role based on calculation; If the user role belongs to the student role set, then: , assigning a first permission set; , assign a second permission set.

[0008] Preferably, the real-time communication coordination module implements: Integration of multimodal communication channels, including establishing communication channels based on text, voice, video and virtual whiteboard; Delay-sensitive QoS guarantee mechanism, used to set the priority queue for teaching instruction transmission; The collaboration status synchronization engine uses an operation log incremental synchronization algorithm to update collaboration content.

[0009] Preferably, the user portrait library is used to dynamically describe the characteristics of the user; the user portrait library includes: Multidimensional feature vector set, including storage learning behavior features and historical permission level set; Dynamic update module, which tracks the evolution of the image based on the hidden Markov model; The method for dynamically describing the characteristics of the user is: Initialize the HMM model and train the HMM model based on user data and the Baum-Welch algorithm; Regularly collect the user's feature vector, where the feature vector is the value of the user in each feature dimension of the multidimensional feature vector set; Use the HMM model to predict the user's future feature vector based on the historical feature vector and the current feature vector; Update the predicted future feature vector to the user portrait library.

[0010] Preferably, the role identification module parsing the user identity identifier further comprises: Based on the user portrait library, obtain the user's historical permission level set; Process the user's historical permission level set to obtain the average permission level; If the average value of the user's permission level is not less than 0.5 and the user role belongs to the student role set, the second permission set is assigned; Otherwise, the first permission set is assigned.

[0011] Preferably, the method for optimizing collaborative learning by the intelligent analysis engine is: Build a student interaction network based on group configuration scheme; Training GNN model based on student interaction network; Use the GNN model to analyze the knowledge propagation path in the student interaction network; Optimize the team configuration plan based on the analysis results; The method for optimizing collaborative learning with an intelligent analysis engine also includes: Extract key knowledge points based on natural language processing technology, including concepts, definitions and formulas; The extracted key knowledge points are used to represent the graph structure and construct a knowledge graph, including taking the knowledge points as nodes and the relationships between the knowledge points as edges; the relationships between the knowledge points include hierarchical relationships and associative relationships; Based on the user portrait library, a multi-dimensional feature vector set is obtained, the knowledge graph is dynamically updated, and a personalized learning path is provided.

[0012] Preferably, the visual interactive interface includes: A three-dimensional virtual classroom space rendering module is used to provide a three-dimensional virtual classroom space; Multi-focus collaborative attention guidance interface, used to process discussion content and display it differently; A real-time collaboration efficiency heat map generation component is used to display the activity level of group collaboration in real time.

[0013] Preferably, it also includes: Edge computing nodes are deployed on the user terminal side and are used for localized behavior feature extraction, differential privacy protection in collaborative scenarios, and offline operation log caching and synchronization.

[0014] Preferably, the teaching resource pool is used for: Intelligent matching of multi-version resources, automatically associating heterogeneous resources based on semantic similarity analysis; Dynamic copyright management uses blockchain technology to record resource transfer paths.

[0015] The beneficial effects of the present invention are as follows: The smart education management system based on multi-user collaboration described in the present invention can realize multi-user collaboration and real-time communication through the setting of user terminal groups, break the time and space constraints, and improve collaboration efficiency. In view of the inability of static permission allocation in the prior art to meet the needs of dynamic permission changes, through the collaborative management server, it can use the dynamic permission allocation engine according to actual conditions, dynamically allocate permissions according to needs and role changes, and meet the needs of online collaborative learning. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention will be further described below in conjunction with the accompanying drawings.

[0017] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0018] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0019] like Figure 1 As shown, a smart education management system based on multi-user collaboration described in an embodiment of the present invention includes a user terminal group, a collaborative management server, a cloud database, an intelligent analysis engine and a visual interactive interface; the user terminal group is a distributed access device consisting of a teacher terminal, a student terminal and an administrator terminal; the collaborative management server is configured with a dynamic authority allocation engine and a real-time communication coordination module; the cloud database stores a user portrait library and a teaching resource pool; the intelligent analysis engine is used to optimize collaborative learning, is connected to the collaborative management server and is configured with: Group behavior modeling unit, used to train GNN models and optimize configuration schemes; A knowledge graph construction unit, used to construct a knowledge graph; Personalized recommendation unit, used to provide personalized learning paths; Visual interactive interface, providing a collaborative operation platform for switching between multiple views.

[0020] In traditional teaching environments, communication and collaboration between teachers and students are often limited by time and space, resulting in low efficiency. Specifically, in a collaborative learning environment, different roles need to match different permissions, which leads to cumbersome permission settings. Static permission allocation cannot meet the needs of dynamic changes, resulting in low efficiency of online collaborative learning. In one embodiment of the present invention, a smart education management system is set up, including a user terminal group, which is composed of a teacher terminal, a student terminal and an administrator terminal, wherein the administrator terminal is only used to monitor the process of online collaborative learning. Based on the setting of the user terminal group, multi-user collaboration and real-time communication can be achieved, breaking the time and space limitations and improving the collaboration efficiency. In view of the fact that the static permission allocation in the prior art cannot meet the needs of dynamic permission changes, in the present invention, through the collaborative management server, according to the actual situation, the dynamic permission allocation engine can be used to dynamically allocate permissions according to needs and role changes to meet the needs of online collaborative learning. In addition, in the present invention, a cloud database is also set up to store user portrait libraries and teaching resource pools to achieve centralized management and sharing of resources. The intelligent analysis engine helps students optimize collaborative learning and provide personalized learning support through group behavior modeling, knowledge graph construction and personalized recommendation units. The visual interactive interface provides a collaborative operation platform to enhance user experience and collaboration effects. Exemplary: A university is carrying out an interdisciplinary collaborative learning project involving teachers and students from multiple disciplines; teachers upload course materials to the cloud database through teacher terminals; students log in to the system through student terminals, view course materials and participate in discussions; administrators monitor the system operation status through administrator terminals to ensure security and stability; the cloud database stores rich teaching resources and student portraits.

[0021] The collaborative management server dynamically allocates permissions to teachers, students, and administrators based on course requirements.

[0022] For example, in the discussion session, all students are given the right to speak; in the homework submission session, only teachers have the authority to correct; students can obtain relevant information from the resource pool according to their interests and needs. The intelligent analysis engine models the group behavior of students' discussion content to discover hot topics and potential problems; the knowledge graph construction unit builds a knowledge system based on course content and student interaction; the personalized recommendation unit recommends relevant resources and activities based on students' learning behavior and preferences; the visual interactive interface provides multiple views such as classroom view, discussion view, resource view, etc.; students can move around and communicate freely in the classroom view; focus on key points in the discussion view; find and share information in the resource view; During the discussion session, student A expressed an opinion in the discussion view. The real-time communication coordination module immediately synchronized this opinion to all student terminals, triggering a heated discussion. As the discussion deepened, the teacher temporarily granted some students the authority to host the discussion through the dynamic permission allocation engine to guide the discussion direction. The intelligent analysis engine found that Student B mentioned "AI ethics" many times in the discussion, so the personalized recommendation unit recommended relevant research papers and courses to Student B; after the discussion, Student C switched to the resource view to find materials related to the discussion topic and prepare for subsequent learning.

[0023] The dynamic permission allocation engine is used to update the permission level according to the user role, learning behavior and real-time environment; the dynamic permission allocation engine includes: A role identification module, used to resolve user identity identifiers; Context-aware module, collecting learning behavior and network environment in real time; The permission generation module dynamically adjusts the operation permission level based on fuzzy logic algorithm; The method for the dynamic permission allocation engine to dynamically update the permission level is: Parse and obtain user roles based on the role identification module. If the user role belongs to the student role set, the first permission set or the second permission set is assigned; If the user role does not belong to the student role set, the third permission set is assigned; Then collect real-time data based on the context awareness module; Finally, the user's permission level is dynamically updated based on the fuzzy logic algorithm; In the process of online learning, based on the traditional permission allocation principle, students' permissions generally only include basic viewing, editing their own documents, viewing learning materials and speaking, etc., and it is precisely because the traditional permission allocation principle is static permission that teachers have too heavy tasks in the remote teaching process and the interactivity between students is too single. In one embodiment of the present invention, according to the dynamic permission allocation engine, the data collection of students' learning behavior and real-time environment can be used to dynamically adjust the permission level of some students, reduce the difficulty of teachers' tasks and improve the participation of some students. Among them, the function of the first permission set can be set to view and edit their own documents and participate in discussions; the function of the second permission set can be set to view and edit other users' documents; for example, in the process of test grading, teachers allow some students to participate in the grading process, thereby increasing the interactivity between students and reducing the difficulty of teachers' tasks; The method for dynamically adjusting the user's authority level based on the fuzzy logic algorithm is: Obtain users’ learning behavior and network environment; Calculate learning behavior scores based on real-time data , Network environment score ; According to the formula: ; in, Score the frequency of student interaction. The maximum score for student interaction frequency, To score the test, The highest score for the test. , is the weight factor; ; is the real-time network speed, is the maximum network speed; is the total number of monitoring times during the monitoring period, is the number of network fluctuations; Scoring based on acquired learning behavior , Network environment score Calculate the overall score: ; For example, the learning behavior score of the user is calculated first. Assuming that the full score of the interaction frequency and the test score is 100, and the interaction frequency is calculated to be 80 points, and the test score is calculated to be 90 points, the learning behavior score of the student is expressed as follows: ; Since the interaction frequency cannot truly reflect the students' cognitive level, the interaction frequency is pre-set to 0.3, and the test score is pre-set to 0.7 accordingly; Then calculate the user's network environment score and obtain the real-time network speed of the user's mobile device based on the permissions. Assuming that the ratio of the user's current network speed to the maximum network speed is 0.2, the total number of monitoring times during the monitoring period is 10, and the number of network fluctuations is 5, then: ; Based on the above, the comprehensive score can be obtained as 0.87+0.08=0.95; Among them, the quantification of the interaction frequency score is based on the cumulative number of speeches, questions, etc., and is obtained through weighted calculation.

[0024] Then, based on the fuzzy logic algorithm and the obtained comprehensive score S, the permission level of the user role is output. : ; ; ; ; in, is the membership function of the low authority level, is the membership function of the medium authority level, is the membership function of the high authority level; , , , is the turning point of the membership function; Output the permission level corresponding to the user role based on calculation; If the user role belongs to the student role set, then: , assigning a first permission set; , assign a second permission set.

[0025] In the process of online learning, based on the traditional principle of authority allocation, students' authority generally only includes basic viewing, editing their own documents, viewing learning materials and speaking, etc., and it is precisely because the traditional principle of authority allocation is static authority that teachers have too heavy tasks in the link of distance learning, and the interaction between students is too single. In one embodiment of the present invention, after obtaining the comprehensive score of the student, when calculating the input comprehensive score based on the fuzzy logic algorithm, the turning point of the membership function is set first, that is, the above-mentioned , , , ; In this embodiment, the turning point is set , , , It is only used to demonstrate how to output the corresponding permission level and adjust it when the comprehensive score of different sizes is input into the fuzzy logic algorithm. It can be adaptively adjusted in actual application; specifically, , is the turning point of the membership function of the low authority level; , , , is the turning point of the membership function of the medium authority level, , is the turning point of the membership function of the high authority level; Assume that , , , ; but , , ; but

[0026] Based on the above, the comprehensive score When, based on the pre-set turning point , , , , the output permission level is 0.5; Set the threshold range of the R value. For example, if the user role belongs to the student role set, then: is the first permission set; is the second permission set; First, according to the user's role, if the user belongs to the student role set, the system will assign him an initial permission set, namely the first permission set, which includes basic permissions such as viewing and editing his own documents and participating in discussions; when the user performs an operation, the system will calculate the R value according to his learning behavior and network environment; according to the calculated R value, the system will dynamically adjust the user's permission level; when the user's R value is lower than 0.5, it means that his current state is not suitable for high-authority operations, so the system will maintain his first permission set; when , the user's R value is higher than or equal to 0.5, which means that the current status is good and can perform more advanced operations, so the system will upgrade the permission to the second permission set; for example, the user is allowed to view and modify other users' documents; In addition, it should be noted that after the permission takes effect, the user's permission level adjustment takes effect, the user can perform operations allowed in the second permission set, but still cannot perform operations with higher permissions, such as changing the user role, etc.; Based on the above, according to the students' comprehensive scores and the use of fuzzy logic algorithm to calculate the students' authority levels, according to the pre-set judgment logic, the user role can be modified to the authority set of individual students in the student role set, so that a part of students with better cognitive levels can be screened and these students can be given higher authority, which reduces the difficulty of teachers' tasks in the teaching process and enriches the diversity of interaction between students.

[0027] The real-time communication coordination module implements: Integration of multimodal communication channels, including establishing communication channels based on text, voice, video and virtual whiteboard; Delay-sensitive QoS guarantee mechanism, used to set the priority queue for teaching instruction transmission; The collaboration status synchronization engine uses an operation log incremental synchronization algorithm to update collaboration content.

[0028] In order to ensure the reliability, efficiency and security of real-time communication and meet the needs of online teaching and collaboration, in one embodiment of the present invention, through the integration of multimodal communication channels, text communication between users is allowed, such as sending messages, creating group chats, etc., real-time voice communication between users is allowed, such as online discussions, answering questions, etc., real-time video communication between users is allowed, such as online teaching, remote collaboration, etc., and whiteboard content is allowed to be edited together, such as brainstorming, drawing charts, etc.; in addition, in order to ensure the timely transmission of teaching instructions, the real-time communication coordination module adopts a delay-sensitive QoS guarantee mechanism and sets a teaching instruction transmission priority queue; exemplary: the teaching instructions are divided into different priority levels, for example: High priority: teaching instructions, such as instructions issued by teachers, questions asked by students, etc.

[0029] Different transmission strategies are adopted according to the priority level, for example: High priority: Use low-latency transmission methods, such as RTP protocol, to ensure timely transmission of teaching instructions; Low priority: Use common transmission methods, such as TCP protocol, to ensure the reliability of data transmission.

[0030] Finally, in order to ensure that all users can obtain the latest collaboration status in real time, the real-time communication coordination module adopts an operation log incremental synchronization algorithm, including recording all users' operations on the collaboration content, such as creation, modification, deletion, etc., and only synchronizing the changed operations to other users to avoid transmitting a large amount of redundant data and improve synchronization efficiency. The operations on the collaboration content are classified as the operations allowed in the first permission set. When concurrent operations cause conflicts, the corresponding conflict resolution strategy is adopted, such as: Last Writer Wins: The last written operation is used.

[0031] It also includes voting based on factors such as user role or operation time to select the final collaboration content.

[0032] The user portrait library is used to dynamically describe the characteristics of the user; the user portrait library includes: Multidimensional feature vector set, including storage learning behavior features and historical permission level set; Dynamic update module, which tracks the evolution of the image based on the hidden Markov model; The method for dynamically describing the characteristics of the user is: Initialize the HMM model and train the HMM model based on user data and the Baum-Welch algorithm; Regularly collect the user's feature vector, where the feature vector is the value of the user in each feature dimension of the multidimensional feature vector set; Use the HMM model to predict the user's future feature vector based on the historical feature vector and the current feature vector; Update the predicted future feature vector to the user portrait library.

[0033] In the process of online learning, user characteristics will change over time. Considering that the learning behavior and cognitive level of users in different time periods will change, it is impossible to match the current user with the historical characteristics of the user, so the user portrait needs to be updated; in one embodiment of the present invention, during the user's use, the platform can collect the user's learning data, wherein the learning behavior characteristics include learning time, learning frequency, learning resource access, learning task completion, and knowledge test scores; the platform uses the Baum-Welch algorithm and these data to train the HMM model, wherein the cognitive level of the student can be set in the HMM model, such as elementary, intermediate and advanced; by setting Time, regularly collect students' learning behavior characteristics, for example, collect them once a day, and convert them into feature vectors. For example, the platform can record students' daily learning time, learning frequency, types and number of learning resources visited, number of exercises completed and accuracy, and convert them into feature vectors. Then use the trained HMM model to predict students' cognitive level status based on the learning behavior feature vectors of historical events. For example, assuming that students' learning time, learning frequency and accuracy of exercises are all high, the model predicts that the student's cognitive level status is upgraded to advanced; based on this, the platform will update the predicted cognitive level status to the user portrait library, and form a complete user portrait with other feature information of the student; Among them, learning behavior characteristics include learning time, learning frequency, access to learning resources, completion of learning tasks, knowledge test scores, etc.; cognitive level parameters include knowledge mastery, learning ability, learning style, learning interest, etc.

[0034] The role identification module further comprises: Based on the user portrait library, obtain the user's historical permission level set; Process the user's historical permission level set to obtain the average permission level; If the average value of the user's permission level is not less than 0.5 and the user role belongs to the student role set, the second permission set is assigned; Otherwise, the first permission set is assigned.

[0035] When the dynamic permission allocation engine dynamically updates the user's permission level, in order to reduce the engine's workload, for users whose average permission level in the user portrait library is higher than 0.5, when the role identification module parses the user identity identifier, if the user role belongs to the student role set, it is determined to allocate the second permission set, thereby reducing the engine's computational workload.

[0036] The method for optimizing collaborative learning by the intelligent analysis engine is: Build a student interaction network based on group configuration scheme; Training GNN model based on student interaction network; Use the GNN model to analyze the knowledge propagation path in the student interaction network; Optimize the team configuration plan based on the analysis results; The method for optimizing collaborative learning with an intelligent analysis engine also includes: Extract key knowledge points based on natural language processing technology, including concepts, definitions and formulas; The extracted key knowledge points are used to represent the graph structure and construct a knowledge graph, including taking the knowledge points as nodes and the relationships between the knowledge points as edges; the relationships between the knowledge points include hierarchical relationships and associative relationships; Based on the user portrait library, a multi-dimensional feature vector set is obtained, the knowledge graph is dynamically updated, and a personalized learning path is provided.

[0037] In the process of online learning, in order to optimize the learning process and increase the interaction between students, establishing groups can quickly improve learning efficiency. In the process of group collaborative learning, the learning efficiency between different groups will be different. Therefore, in order to improve the overall learning efficiency, it is also necessary to optimize the group configuration scheme according to the learning efficiency and interaction intensity of the group. In one embodiment of the present invention, a student interaction network is preferentially established according to the current group configuration scheme, that is, the interactive behavior between students is represented as a graph structure, and students are regarded as nodes, and the interactive behaviors between students, such as asking questions, answering, commenting and collaboration, are regarded as edges; based on the above construction of the interactive network, the GNN model is used to analyze the propagation path of knowledge in group discussions, aiming to find students who play a key role in the group and identify the cognitive gaps between students in the group. Finally, the group configuration scheme is optimized according to the analysis results, that is, students with complementary knowledge are assigned to the same group; exemplary: Student A often asks thoughtful questions and is able to give clear answers.

[0038] Student B actively participated in the discussion and was able to share his learning experiences and insights.

[0039] Student C does not have a deep understanding of a certain knowledge point and often asks related questions.

[0040] Student D does not have a firm grasp of a certain knowledge point and is unable to accurately explain or apply it.

[0041] According to the analysis results, the optimization group configuration plan can be understood as follows: Assign Student A and Student C to the same group so that Student A can help Student C understand and master the knowledge blind spots.

[0042] Assign Student B and Student D to the same group so that Student B can help Student D understand and apply the knowledge.

[0043] In the process of online teaching, since different students have different knowledge level improvements in the same time period, in order to help students with poor cognitive levels improve faster and not delay the learning of students with good cognitive levels, in one embodiment of the present invention, it is considered to use natural language processing technology to extract key knowledge points and construct a knowledge graph to help students with poor cognitive levels improve quickly. At the same time, for students with good cognitive levels, the knowledge graph can be updated for them, that is, personalized learning services are provided. It can be understood that based on the user portrait library, a multi-dimensional feature vector set corresponding to the student is obtained. If the student's cognitive level is advanced, the platform may recommend more difficult learning content and provide more challenging learning paths and exercises.

[0044] The visual interaction interface includes: A three-dimensional virtual classroom space rendering module is used to provide a three-dimensional virtual classroom space; Multi-focus collaborative attention guidance interface, used to process discussion content and display it differently; A real-time collaboration efficiency heat map generation component is used to display the activity level of group collaboration in real time.

[0045] In the process of online learning, in order to help students improve their learning experience and effect intuitively and efficiently, in one embodiment of the present invention, the system also includes a visual interactive interface, which provides an intuitive, efficient and convenient collaborative operation platform, supports multi-view switching and multiple interaction methods, and specifically includes a three-dimensional virtual classroom space rendering module for providing a three-dimensional virtual classroom space. After going online, students can move around freely, communicate and carry out collaborative learning activities in the three-dimensional virtual classroom space. Demonstrationally, the three-dimensional virtual classroom space can customize the layout of the classroom, the position of the seats and the configuration of teaching equipment. The interaction method includes students moving around freely and communicating in the form of text, voice or video; it also includes a multi-focus collaborative attention guidance interface. In the discussion session, the content of the discussion can be divided into different focus areas and labeled. In addition, in order to intuitively reflect different content, the focus areas should be visualized to students in different colors, shapes and sizes. Demonstrationally, for core ideas, eye-catching colors and fonts can be used to highlight them. For supporting arguments, lines or arrows can be used to connect core ideas and supporting arguments, while for opposing arguments, different colors and shapes are used to distinguish them. For questions and comments, bubbles or icons can be used to display them. It also includes a real-time collaboration performance heat map generation component, which is used to display the activity level of group collaboration in real time. According to the above, according to the interaction between students, the group configuration is optimized according to the intelligent analysis engine. In the discussion session, a heat map can be used to display the interaction of the group in the virtual classroom, including activity intensity and activity type. The activity intensity can be represented by red, for example, dark red indicates frequent activities, while light red indicates less activities. For activity types, different colors or shapes can be used to represent them, for example, green is used for voice communication, yellow is used for text communication, and blue is used for collaborative editing. During the whole process, the heat map is updated in real time to reflect the cone state of student activities. Teachers can view the activities of each group by clicking on the heat map, thereby increasing the interaction between students and teachers.

[0046] Also includes: Edge computing nodes are deployed on the user terminal side and are used for localized behavior feature extraction, differential privacy protection in collaborative scenarios, and offline operation log caching and synchronization.

[0047] In the process of online learning, it is inevitable that students may lose data due to device status or network environment problems. To solve this problem, in one embodiment of the present invention, an edge computing node is also set up. Through localized behavior feature extraction, the behavior of students when using the system can be monitored and analyzed in real time, and key behavior features such as learning time, interaction frequency, and browsing content can be extracted. In addition, a collaborative scenario differential privacy protection function is also set up. In the process of collaborative learning, differential privacy technology can be used to perturb students' behavior data to ensure that individual privacy will not be leaked during data sharing and analysis. It also includes offline operation log caching and synchronization functions. When students When in an offline state, the edge computing node will cache its completed learning tasks and interaction records. Once connected to the network, the data will be synchronized with the cloud database to ensure that the student's learning records are complete and continuous and are not affected by the network status. For example, when a student is learning online, there is a network interruption. During the network interruption, he continues to study in the virtual classroom and completes some tasks. The edge computing node monitors and records these behaviors in real time, and caches the learning tasks and interaction records in local storage. When the network is restored, the edge computing node can synchronize the cached offline operation logs with the cloud database and integrate these data to update the student's learning records and collaboration efficiency heat map.

[0048] The teaching resource pool is used for: Intelligent matching of multi-version resources, automatically associating heterogeneous resources based on semantic similarity analysis; Dynamic copyright management uses blockchain technology to record resource transfer paths.

[0049] The teaching resource pool can store multiple versions of teaching resources, and based on semantic similarity analysis technology, automatically associate and match resources of different structures to ensure that teachers and students can obtain the most relevant teaching resources and improve teaching efficiency. Among them, the dynamic copyright management module can use blockchain technology to record the circulation path of teaching resources, ensure that the copyright information of resources is clear and traceable, and aims to protect the copyright of the original creator, prevent the resources from being illegally copied and disseminated, and facilitate the legal sharing and trading of resources. The cross-platform resource conversion engine is used to convert and make teaching resources compatible between different platforms, so as to achieve extensive sharing and reuse of teaching resources. For example, when teachers are preparing a course on "Basics of Artificial Intelligence", they only need to search for "Basics of Artificial Intelligence" in the resource pool. The system can automatically match and display multiple versions of related resources, including textbooks, videos, and courseware through semantic similarity analysis. Teachers can choose the most appropriate resource version according to their needs. The resources selected by teachers carry copyright information recorded in the blockchain. The system will automatically verify the copyright status of the resources to ensure that the resources can be used legally. When teachers use resources, the system will record the usage behavior and update the circulation path of the resources.

[0050] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A smart education management system based on multi-user collaboration, characterized in that: include: User terminal group, a distributed access device consisting of teacher terminals, student terminals and administrator terminals; A collaborative management server, configured with a dynamic rights allocation engine and a real-time communication coordination module; The cloud database stores user portraits and teaching resource pools; An intelligent analysis engine, used for optimizing collaborative learning, connected to the collaborative management server and configured with: Group behavior modeling unit, used to train GNN models and optimize configuration schemes; A knowledge graph construction unit, used to construct a knowledge graph; Personalized recommendation unit, used to provide personalized learning paths; Visual interactive interface, providing a collaborative operation platform for switching between multiple views.

2. According to claim 1, a smart education management system based on multi-user collaboration is characterized in that: The dynamic permission allocation engine is used to update the permission level according to the user role, learning behavior and real-time environment; The dynamic rights allocation engine comprises: A role identification module, used to resolve user identity identifiers; Context-aware module, collecting learning behavior and network environment in real time; The permission generation module dynamically adjusts the operation permission level based on fuzzy logic algorithm; The method for the dynamic permission allocation engine to dynamically update the permission level is: Parse and obtain user roles based on the role identification module. If the user role belongs to the student role set, the first permission set or the second permission set is assigned; If the user role does not belong to the student role set, the third permission set is assigned; Then collect real-time data based on the context awareness module; Finally, the user's authority level is dynamically updated based on the fuzzy logic algorithm.

3. According to claim 2, a smart education management system based on multi-user collaboration is characterized in that: The method for dynamically adjusting the user's authority level based on the fuzzy logic algorithm is: Obtain users’ learning behavior and network environment; Calculate learning behavior scores based on real-time data , Network environment score ; According to the formula: ； in, Score the frequency of student interaction. The maximum score for student interaction frequency, To score the test, The highest score for the test. , is the weight factor; ； is the real-time network speed, is the maximum network speed; is the total number of monitoring times during the monitoring period, is the number of network fluctuations; Scoring based on acquired learning behavior , Network environment score Calculate the overall score: ； Then, based on the fuzzy logic algorithm and the obtained comprehensive score S, the permission level of the user role is output. : ； ； ； ； in, is the membership function of the low authority level, is the membership function of the medium authority level, is the membership function of the high authority level; , , , is the turning point of the membership function, , is the turning point of the membership function of the low authority level; , , , is the turning point of the membership function of the medium authority level, , is the turning point of the membership function of the high authority level; Output the permission level corresponding to the user role based on calculation; If the user role belongs to the student role set, then: , assigning a first permission set; , assign a second permission set.

4. According to claim 1, a smart education management system based on multi-user collaboration is characterized in that: The real-time communication coordination module implements: Integration of multimodal communication channels, including establishing communication channels based on text, voice, video and virtual whiteboard; Delay-sensitive QoS guarantee mechanism, used to set the priority queue for teaching instruction transmission; The collaboration status synchronization engine uses an operation log incremental synchronization algorithm to update collaboration content.

5. The intelligent education management system based on multi-user collaboration according to claim 1 is characterized in that: The user portrait library is used to dynamically describe the characteristics of the user; the user portrait library includes: Multidimensional feature vector set, including storage learning behavior features and historical permission level set; Dynamic update module, which tracks the evolution of the image based on the hidden Markov model; The method for dynamically describing the characteristics of the user is: Initialize the HMM model and train the HMM model based on user data and the Baum-Welch algorithm; Regularly collect the user's feature vector, where the feature vector is the value of the user in each feature dimension of the multidimensional feature vector set; Use the HMM model to predict the user's future feature vector based on the historical feature vector and the current feature vector; Update the predicted future feature vector to the user portrait library.

6. The intelligent education management system based on multi-user collaboration according to claim 5 is characterized in that: The role identification module further comprises: Based on the user portrait library, obtain the user's historical permission level set; Process the user's historical permission level set to obtain the average permission level; If the average value of the user's permission level is not less than 0.5 and the user role belongs to the student role set, the second permission set is assigned; Otherwise, the first permission set is assigned.

7. The intelligent education management system based on multi-user collaboration according to claim 1 is characterized in that: The method for optimizing collaborative learning by the intelligent analysis engine is: Build a student interaction network based on group configuration scheme; Training GNN model based on student interaction network; Use the GNN model to analyze the knowledge propagation path in the student interaction network; Optimize the team configuration plan based on the analysis results; The method for optimizing collaborative learning with an intelligent analysis engine also includes: Extract key knowledge points based on natural language processing technology, including concepts, definitions and formulas; The extracted key knowledge points are used to represent the graph structure and construct a knowledge graph, including taking the knowledge points as nodes and the relationships between the knowledge points as edges; the relationships between the knowledge points include hierarchical relationships and associative relationships; Based on the user portrait library, a multi-dimensional feature vector set is obtained, the knowledge graph is dynamically updated, and a personalized learning path is provided.

8. The intelligent education management system based on multi-user collaboration according to claim 1 is characterized in that: The visual interaction interface includes: A three-dimensional virtual classroom space rendering module is used to provide a three-dimensional virtual classroom space; Multi-focus collaborative attention guidance interface, used to process discussion content and display it differently; A real-time collaboration efficiency heat map generation component is used to display the activity level of group collaboration in real time.

9. The intelligent education management system based on multi-user collaboration according to claim 8, characterized in that: Also includes: Edge computing nodes are deployed on the user terminal side and are used for localized behavior feature extraction, differential privacy protection in collaborative scenarios, and offline operation log caching and synchronization.

10. The intelligent education management system based on multi-user collaboration according to claim 1, characterized in that: The teaching resource pool is used for: Intelligent matching of multi-version resources, automatically associating heterogeneous resources based on semantic similarity analysis; Dynamic copyright management uses blockchain technology to record resource transfer paths.

Images