An AI-assisted interactive teaching method and system for resource imbalance scenarios

Abstract

The application provides an AI-assisted interactive teaching method and system for resource imbalance scenarios, which comprises the following steps: first, establishing a real-time audio and video connection based on the course information created by the teaching end. Second, during the learning process, the question information of the learning end is received, the AI module generates and returns the answer content based on the adaptive knowledge base, and the question is synchronized to the teaching end. Third, the test module generates test information based on knowledge points, grades the answer information after receiving it, and generates a learning situation report. Then, according to the learning situation report and the historical learning track, the matching learning resources are selected from the hierarchical resource library and pushed to the learning end. Finally, the teaching and research module creates a teaching and research activity, receives the adaptive materials uploaded by the teaching end, and synchronizes them to the adaptive knowledge base and / or hierarchical resource library after the activity is completed. This method realizes the closed loop of teaching, answering questions, testing, resource pushing and teaching and research activities, can provide personalized learning support, improve the efficiency of teaching interaction and the pertinence of resource utilization, and promote the continuous accumulation and sharing of high-quality teaching resources.

Description

Technical Field

[0001] This invention relates to the field of educational technology, and in particular to an AI-assisted interactive teaching method and system for scenarios with uneven resource distribution. Background Technology

[0002] In the field of online education technology, especially in educational scenarios with uneven resource distribution, effectively integrating teaching, interaction, assessment, and resource provision to improve teaching quality and learning outcomes is an important research direction. These technologies are typically applied to scenarios such as remote teaching, after-school tutoring, and teacher training where network conditions vary and teacher resources are unevenly distributed. The aim is to bridge the gap in educational resources through technological means and achieve personalized learning support.

[0003] In existing technologies, online live streaming or pre-recorded courses are commonly used to achieve remote interactive teaching. Instructors deliver lectures via audio and video streams, while students interact by asking questions through text chat or voice communication. To assess learning outcomes, some systems offer online quizzes after the course. Regarding resource provision, a common practice is to provide a unified resource library for students to download or browse learning materials independently. Furthermore, teachers' research activities are often conducted through independent conferencing systems or file-sharing platforms, making it difficult to integrate high-quality teaching resources into the daily teaching process in a timely and systematic manner.

[0004] However, the aforementioned existing technical solutions have several limitations in practical applications. First, they lack interactive teaching and suffer from delayed feedback. In traditional online classrooms, students' questions often require manual responses from teachers. In scenarios with uneven resource distribution, teachers may not be able to handle a large number of personalized questions in a timely manner, leading to unresolved learning obstacles and affecting the continuity of learning. Second, learning assessment and resource delivery are disconnected. Existing testing functions are usually independent of the teaching process, and the generated learning reports are relatively simple and difficult to deeply integrate with students' historical learning data. This makes it impossible to dynamically and accurately deliver personalized learning resources tailored to students' current level and weaknesses. Third, teaching research results are difficult to effectively translate. The high-quality teaching materials and methods generated in teachers' teaching research activities lack an effective mechanism for accumulation and systematic application in subsequent daily teaching and resource database updates, resulting in a waste of knowledge assets and failing to form a virtuous cycle of "teaching-teaching research-resource optimization."

[0005] Therefore, there is an urgent need for an interactive teaching method for educational scenarios with uneven resource distribution, in order to solve the problems of delayed interactive feedback in teaching, inaccurate personalized learning resource delivery, and difficulty in effectively accumulating and feeding back teaching results in existing technologies. Summary of the Invention

[0006] This invention relates to the field of educational technology, and in particular to an AI-assisted interactive teaching method for educational scenarios with uneven resource distribution.

[0007] In traditional online teaching practices, especially in educational scenarios with uneven resource distribution, there are common problems such as insufficient real-time interaction between teachers and students, difficulty in providing personalized learning resources based on individual student differences, and difficulty in effectively accumulating and applying high-quality teaching materials generated from teaching and research activities to subsequent teaching loops. The purpose of this invention is to provide an AI-assisted interactive teaching method to solve at least one of the above-mentioned technical problems.

[0008] To achieve the above objectives, this invention provides an AI-assisted interactive teaching method for educational scenarios with uneven resource distribution. The method includes the following steps: First, receiving course information created by the teaching terminal and establishing a real-time audio-visual connection between the teaching terminal and at least one learning terminal based on the course information, wherein the course information includes knowledge points and teaching materials. Second, during the real-time audio-visual connection, receiving question information input from the learning terminal, generating answer content based on an adapted knowledge base associated with the knowledge points using an AI module and returning it to the learning terminal, while simultaneously synchronizing the question information to the teaching terminal. Third, generating test information from an adapted question bank based on the knowledge points using a testing module and sending it to the learning terminal, receiving and grading the answers submitted by the learning terminal, and generating a learning report containing information on learning mastery. Fourth, selecting matching learning resources from a hierarchical resource library and pushing them to the learning terminal based on the learning report and the learning terminal's historical learning trajectory. The fifth step involves creating teaching and research activities through the teaching and research module, receiving the adaptation materials uploaded by the teaching terminal, and synchronizing the adaptation materials to the adaptation knowledge base and / or the hierarchical resource base after the teaching and research activities are completed.

[0009] In one possible implementation, the first step of establishing the real-time audio and video connection includes: the teaching end initiating a device test to detect the status of the audio and video device, and dynamically enabling a network adaptation mode to adjust the audio and video transmission quality based on the network bandwidth; the learning end entering the corresponding course and receiving the real-time audio and video connection, wherein the learning end supports raising its hand to request and submitting questions during the connection process. Further, the network adaptation mode dynamically switches the encoding protocol or transmission resolution of the audio and video streams based on the real-time detected network bandwidth.

[0010] In one possible implementation, the second step, where the AI ​​module generates Q&A content based on an adapted knowledge base, includes: performing semantic analysis on the question information input from the learning end to extract the core question; matching the core question with the adapted knowledge base to generate multimodal Q&A content containing at least one of text, voice, or visual elements; counting the frequency of the question information, and generating a reminder message and sending it to the teaching end when the frequency exceeds a preset threshold. Further, the reminder message is sent to the teaching end's interactive interface in the form of a pop-up window, and the reminder message contains a list of frequently asked questions.

[0011] In one possible implementation, the third step of generating test information and sending it to the learning terminal through the testing module includes: automatically generating test papers from the adapted question bank based on the knowledge points selected by the teaching terminal or the auxiliary terminal and the set test parameters; publishing the test information to the learning terminal and setting the answering time limit; receiving the answer information submitted by the learning terminal, using a rule engine to grade the objective questions, and using keyword matching to generate a scoring reference for the subjective questions.

[0012] In one possible implementation, the fourth step, selecting matching learning resources from the hierarchical resource library and pushing them to the learning end, includes: determining resource push logic based on the knowledge gaps identified in the learning report and the historical learning trajectory; automatically pushing matching resources with scene tags from the hierarchical resource library to the learning end based on the resource push logic; receiving feedback records from the learning end regarding the pushed resources, and adjusting the content of subsequently pushed resources based on the feedback records. Further, adjusting the content of subsequently pushed resources based on the feedback records includes: if the feedback record indicates a lack of understanding of the currently pushed resource, then selecting a more basic level resource from the hierarchical resource library for pushing; if the feedback record indicates understanding of the currently pushed resource, then selecting a higher level of enhancement resource from the hierarchical resource library for pushing.

[0013] In one possible implementation, the fifth step of creating a teaching and research activity through the teaching and research module includes: the initiating end creating the teaching and research activity and setting the activity theme and format, and inviting at least one teaching end as a participating end; the participating end uploading adaptation materials to the shared space of the teaching and research module; during the teaching and research activity, recording the interaction content between the participating ends; after the activity ends, storing the adaptation materials and recorded interaction content by category, and synchronizing them to the adaptation knowledge base and / or the hierarchical resource base.

[0014] In one possible implementation, the learning report displays the knowledge mastery rate and list of incorrect questions on the learning end in a visual form, and the visualization results are used in the resource push logic in the fourth step.

[0015] By adopting the above technical solution, the present invention can achieve at least the following beneficial effects:

[0016] First, by establishing real-time audio and video connections that support network adaptation and integrating multiple aspects such as AI-powered Q&A, intelligent testing, personalized resource delivery, and sharing of teaching and research results, a complete teaching loop is constructed, effectively improving the interactivity and continuity of teaching in scenarios with uneven resource distribution. Second, utilizing the AI ​​module for multimodal Q&A based on an adapted knowledge base enables timely responses to student questions, reducing the burden of repetitive Q&A for teachers. Simultaneously, by statistically analyzing frequently asked questions and alerting teachers, it helps teachers dynamically adjust teaching focus and improve the relevance of instruction. Third, the testing module automatically generates test papers, intelligently grades them, and produces visualized learning reports, enabling rapid and objective evaluation of learning outcomes and providing accurate data for personalized resource delivery. Fourth, based on learning reports and learning trajectories, matching resources are intelligently delivered from a tiered resource library, and the resource hierarchy is dynamically adjusted based on student feedback, achieving truly personalized learning path planning and effectively compensating for the shortcomings of the "one-size-fits-all" approach to resource delivery in traditional teaching. Fifth, the teaching and research module allows high-quality, relevant materials generated during teaching and research activities to be stored and synchronized to the knowledge base and resource base, enabling the teaching and research results to directly support daily teaching, promoting the continuous optimization and sharing of teaching resources, and is especially beneficial to improving teaching conditions in resource-scarce areas. Attached Figure Description

[0017] Figure 1 A flowchart illustrating an AI-assisted interactive teaching method for educational scenarios with uneven resource distribution, provided as an embodiment of this application;

[0018] Figure 2 A schematic diagram illustrating a specific operation process of an AI-assisted interactive teaching method for educational scenarios with uneven resource distribution, provided in an embodiment of this application;

[0019] Figure 3 This is another schematic diagram illustrating the specific operation process of an AI-assisted interactive teaching method for educational scenarios with uneven resource distribution, provided as an embodiment of this application. Detailed Implementation

[0020] Example 1

[0021] This embodiment provides a closed-loop method for AI-assisted interactive teaching in scenarios with uneven resource distribution. This method aims to address the problems of uneven distribution of educational resources, insufficient teacher-student interaction, and the difficulty in achieving personalized teaching by constructing a system integrating a teaching end, a learning end, an AI module, a testing module, a resource module, and a teaching research module.

[0022] like Figure 1As shown, the system architecture supports interactive teaching between multiple teaching and learning terminals through real-time connection, and is supported by intelligent modules such as AI, testing, resources and teaching research in the background, forming a complete closed loop from teaching, interaction, assessment to resource push and teaching research accumulation.

[0023] Specifically, the AI-assisted interactive teaching closed-loop method for resource imbalance scenarios provided in this embodiment includes the following steps:

[0024] S1. Receive course information created by the teaching terminal, and establish a real-time audio and video connection between the teaching terminal and at least one learning terminal based on the course information. The course information includes knowledge points and teaching materials.

[0025] S2. During the real-time audio and video connection process, the system receives the question information input by the learning terminal, generates the answer content based on the adaptive knowledge base associated with the knowledge point through the AI ​​module, and returns it to the learning terminal. At the same time, the question information is synchronized to the teaching terminal.

[0026] S3. The testing module generates test information from the adaptation question bank based on the knowledge points and sends it to the learning terminal. After receiving the answer information submitted by the learning terminal, it grades the answer and generates a learning report containing the learning mastery status.

[0027] S4. Based on the learning progress report and the historical learning trajectory of the learning terminal, select matching learning resources from the hierarchical resource library and push them to the learning terminal;

[0028] S5. Create teaching and research activities through the teaching and research module, receive the adaptation materials uploaded by the teaching terminal, and synchronize the adaptation materials to the adaptation knowledge base and / or the hierarchical resource base after the teaching and research activities are completed.

[0029] First, the teaching end creates the course and establishes a real-time connection. The teaching end (e.g., the teacher's terminal device) creates the course through the system interface, setting the course topic, objectives, and syllabus. Once created, the system generates a course entry point, allowing the teaching end to invite or authorize designated learning ends (e.g., students' terminal devices) to join. When a learning end accepts the invitation, a real-time audio, video, and data connection is established between the teaching end and the learning end to support subsequent real-time teaching and interaction. During this process, the system can perform basic network checks to ensure connection stability.

[0030] Secondly, the learning client submits questions, which are then answered and synchronized by the AI ​​module. During real-time instruction, the learning client can submit questions in text or voice format through an interactive interface. The AI ​​module receives these questions in real time and analyzes them based on a pre-set knowledge base or large language model to generate preliminary answers. These answers can be immediately fed back to the learning client that submitted the question and simultaneously synchronized to the instructor's interface, such as in a list or pop-up window, allowing the instructor to promptly understand the students' points of confusion and provide supplementary explanations when appropriate.

[0031] Third, the testing module generates test papers, grades them, and produces learning reports. At specific stages of the course (such as after all knowledge points have been explained) or after the course ends, the teaching end can trigger a test. The testing module automatically or semi-automatically generates test questions based on the course content, creates a test paper, and pushes it to the learning end. After the learning end completes the test and submits their answers, the testing module automatically grades the answers and, based on the grading results and the learning end's interaction history during the course (such as the frequency of question submissions and Q&A feedback), generates a comprehensive learning report. This learning report reflects each learning end's mastery of the knowledge points, weaknesses, and learning characteristics.

[0032] Fourth, based on learning reports and interaction data, the resource module pushes tiered learning resources. The resource module receives learning reports from the testing module and interaction data from the AI ​​module. Based on this data, the resource module performs personalized analysis for each learning end and matches and pushes tiered learning resources from a pre-set resource library. For example, for learning ends showing lower mastery levels in their learning reports, more basic and detailed review materials or practice questions are pushed; for learning ends showing higher mastery levels, more challenging extension materials or advanced application cases are pushed. The pushed resources can be delivered to the learning end via message notifications or dedicated learning spaces.

[0033] Fifth, the teaching research module accumulates and synchronizes suitable materials to the knowledge base and resource library; the teaching research module also provides the teaching end with the function of organizing teaching research activities. The teaching end can initiate a teaching research activity based on the teaching process, student interaction data, learning reports, and final teaching results of this course. In the teaching research activity, the teaching end can organize the effective teaching materials generated in this course (such as optimized explanation segments, collections of answers to frequently asked questions, highly adaptable test questions, etc.) and classify and store them in the system's shared knowledge base. At the same time, high-quality resources that have been verified and optimized through teaching research can also be synchronously updated to the resource library of the resource module for future use by other courses or other teaching ends, thereby realizing the accumulation of teaching experience and the sharing and circulation of high-quality resources.

[0034] Through the sequential execution and iterative iteration of the above steps, the method in this embodiment constructs a complete teaching loop. It not only achieves real-time interaction, intelligent Q&A, and personalized assessment within a single lesson, but also transforms the experience of a single teaching session into reusable, systematic knowledge through resource delivery and teaching research accumulation, thereby promoting the balanced allocation of educational resources and the continuous improvement of teaching quality on a macro level.

[0035] In a preferred embodiment of this example, the preferred implementation method for network adaptation and device testing is...

[0036] This preferred embodiment follows the steps in Example 1 regarding the creation of courses and the establishment of real-time connections at the teaching end. It provides a detailed description of the specific implementation methods involved in this step, such as device testing, network adaptation, and retention of interactive content, in order to improve the stability of interactive teaching and the accessibility of core content in scenarios with uneven resources.

[0037] Before creating a course and establishing a real-time connection on the teaching end, the system first performs a device testing process. Specifically, the client software or application running on the teaching end device (such as the teacher's computer or mobile terminal) calls the local hardware detection interface to detect the working status of key audio and video input / output devices such as cameras, microphones, and speakers. This detection may include checking whether the device is correctly identified, whether the driver is working properly, and whether it has basic audio and video capture and playback capabilities. If any key device is detected to be abnormal (e.g., the camera cannot be turned on, or the microphone has no input signal), the system will generate and display corresponding abnormal prompts on the teaching end's user interface. These prompts can clearly indicate the abnormal device and its possible causes (e.g., "camera not connected" or "microphone permission disabled"), guiding the instructor to troubleshoot and repair, thereby ensuring the availability of basic interactive equipment before the formal teaching begins.

[0038] After the equipment tests pass, the system further performs network bandwidth detection. The teaching end sends a series of test data packets to the server or the learning end, and evaluates the current network link's bandwidth status and stability based on the round-trip time, packet loss rate, and transmission rate. Based on the detected network bandwidth results, the system dynamically selects or switches the encoding protocol and transmission parameters of the audio and video streams. For example, when the detected network bandwidth is high and stable (e.g., greater than 2Mbps), the system can use a high-resolution, high-frame-rate video encoding protocol (e.g., H.264 High Profile or H.265) for transmission to provide a clearer picture; while when the detected network bandwidth is low or unstable (e.g., below 512Kbps), the system automatically switches to a low-resolution, low-bitrate encoding protocol (e.g., H.264 Baseline Profile), and may reduce the frame rate, or even prioritize the transmission of the audio stream to ensure the continuity of the teaching voice. This dynamic switching mechanism helps adapt to resource differences in different regions and network environments, ensuring basic smoothness of real-time audio and video interaction.

[0039] Furthermore, to address potential interactive difficulties in weak network environments, this preferred implementation also includes a core content retention mechanism. On the learning end, when network latency or jitter exceeds a preset threshold (e.g., latency greater than 500 milliseconds), the client activates a local caching strategy. Specifically, for real-time video streams pushed by the teaching end, the learning end strives to render the latest frames while caching key video frames and synchronized audio data received within the most recent period (e.g., the first 10 seconds) in local memory or storage. More importantly, for non-streaming core content closely related to the teaching progress, such as whiteboard notes written by the teacher, shared PPT document page-turning commands, or key text information sent through chat windows, the system prioritizes the reliable transmission of these commands and data, ensuring their correct presentation and persistent display on the learning end interface. Even if real-time video experiences stuttering or brief interruptions due to network issues, learning end users can still view and review received whiteboard notes, documents, and key text, thus avoiding complete loss of core teaching content. In some alternative implementations, the learning interface may also provide a "content review" area, which is dedicated to displaying the core teaching materials that have been successfully received and cached under the current network conditions.

[0040] Through the coordinated efforts of the aforementioned equipment testing, bandwidth adaptive encoding switching, and the learning-end core content retention mechanism, this preferred embodiment can provide a more stable and reliable basic communication environment for subsequent steps (such as real-time Q&A, testing, etc.) of the AI-assisted interactive teaching closed loop in resource-uneven scenarios where network and equipment conditions differ.

[0041] In a preferred embodiment of this example, the reminder and interaction process for frequently asked questions is described in detail. This process follows the steps in Example 1 where the learning end submits questions and the AI ​​module answers them synchronously. The aim is to improve the timeliness of the teaching end's response to common questions through real-time statistics and proactive reminders.

[0042] like Figure 2 As shown, after receiving question data from multiple learning endpoints, the AI ​​module initiates a real-time frequency statistics and judgment process. Specifically, this process may include:

[0043] S51, Question Keyword Extraction and Clustering Submodule, is used to perform natural language processing on the received question text, extract core knowledge points or concepts as keywords, and group semantically similar questions into the same cluster;

[0044] S52, Frequency Statistics and Threshold Judgment Submodule, is used to count in real time the number of times each question cluster appears within a preset time window (e.g., the last 5 minutes or the current course segment), and compare the number of times with the preset frequency threshold.

[0045] S53, the reminder generation and distribution submodule, is used to generate reminder information containing the high-frequency question content, the frequency of occurrence, and the related knowledge points when the occurrence frequency of any question cluster exceeds the frequency threshold, and distribute the reminder information to the teaching terminal.

[0046] The frequency statistics and threshold judgment submodule can dynamically adjust the length of the preset time window and the value of the frequency threshold. For example, for key and difficult chapters of a course, a shorter time window (e.g., 2 minutes) and a lower frequency threshold (e.g., 3 times) can be set to achieve more sensitive reminders; for general content, a longer time window and a higher frequency threshold can be set to avoid interference. Those skilled in the art will understand that the specific settings of the above parameters can be optimized and adjusted according to the course type, student size, and historical data experience.

[0047] When the reminder message arrives at the teaching platform, it will be highlighted as a pop-up in the user interface. The pop-up typically includes a text summary of frequently asked questions, the number (or frequency) of students asking the question, and the corresponding class time. In a typical interactive scenario, the teacher can choose to pause the current lecture and directly address the frequently asked question; alternatively, they can choose to temporarily ignore the pop-up, but the information will remain in the question list in the sidebar for later review. This design ensures that the teaching platform can promptly identify common student concerns, thereby adjusting the teaching pace and focus.

[0048] Furthermore, the frequently asked questions collected during the above process will be simultaneously written into the learning progress report. Specifically, when generating learning progress reports for classes or individuals, the system will incorporate the clustering results of frequently asked questions, their temporal distribution, and the final handling status at the teaching end (such as whether they have been explained) as important analytical dimensions into the report. This helps to identify weaknesses in the course from a macro perspective and provides data support for subsequent optimization of teaching resources and teaching research activities.

[0049] By introducing the aforementioned real-time statistics and pop-up reminder mechanism for frequently asked questions, this implementation method can further transform scattered student questions into centralized teaching feedback on the basis of AI-assisted Q&A, prompting the teaching end to intervene in real time. This can more effectively ensure the bottom line of overall teaching quality in scenarios with uneven resources, and enrich the data dimensions and value of learning reports.

[0050] In a preferred embodiment of this example, the feedback loop mechanism for tiered resource push is described in detail. This mechanism follows the step of "pushing tiered resources based on learning progress reports and trajectories" in Example 1. It aims to dynamically adjust subsequent push strategies by receiving feedback from the learning end on the pushed resources, thereby forming a personalized learning resource push closed loop.

[0051] like Figure 3 As shown, when the learning client receives the tiered learning resources pushed by the system, the interface provides clear feedback options, such as "Understood" and "Not Understood" buttons. Users can select and submit the appropriate feedback based on their understanding of the resource content.

[0052] Specifically, the feedback loop push mechanism includes the following processes:

[0053] S61. When a user selects "Don't understand" for the currently pushed resource, the system will record this feedback along with the corresponding resource identifier, user identifier, and timestamp. Subsequently, based on this feedback, the system will retrieve and push similar learning resources from the resource library that are of lower difficulty, more basic in explanation, or more intuitive in format (such as videos or animations). For example, if a user indicates they don't understand a video explanation of a word problem, the system can automatically push text and images explaining the basic concepts involved in the problem, or simpler example videos.

[0054] S62. When a user selects "Understood" for the currently pushed resource, the system also records this feedback. Based on this positive feedback, the system will determine that the user's current knowledge level may be higher than the initial assessment, and then retrieve and push advanced resources from the resource library that are slightly more difficult, have a broader scope of knowledge, or are more challenging. For example, after a user understands the explanation of a grammar point, the system can push comprehensive practice questions or related reading materials containing that grammar point.

[0055] All resource push records and corresponding user feedback are compiled into a detailed feedback trajectory and integrated into the user's learning report. Support users (such as teaching assistants or parents) can log in to the system to view this feedback trajectory. Support users can clearly see which resources the user understands or does not understand, and which resources the system subsequently automatically pushes. Based on this, support users can make manual judgments and interventions. For example, for knowledge points that the user still "does not understand" after multiple system pushes, support users can send text or voice messages to the learning end or recommend specific human-explained resources for supplementary guidance.

[0056] In some optional implementations, the system can also set a confidence threshold for feedback. For example, if a user continuously reports "don't understand" to multiple different forms of resources on the same knowledge point within a short period of time, the system can automatically increase the "weak level" of that knowledge point in the learning report and trigger a higher-priority reminder sent to the teaching or auxiliary end. Furthermore, the difficulty gradient adjustment range of resource recommendations can also be dynamically set based on the frequency and consistency of feedback. For example, after the first "don't understand" feedback, resources with a difficulty level reduced by one can be recommended; after two consecutive "don't understand" feedbacks, resources with a difficulty level reduced by two can be recommended.

[0057] Through the aforementioned feedback loop mechanism, the system can dynamically and accurately adjust the content and difficulty of the learning resources based on users' real-time understanding and feedback, achieving a shift from "one-way push" to "two-way interaction and closed-loop optimization." This helps to build personalized learning paths that continuously adapt to each learner's true level of understanding in scenarios with uneven resource distribution, improving resource utilization efficiency and teaching support effectiveness.

[0058] In a preferred embodiment of this example, the preferred embodiment of the teaching and research activity accumulation and sharing space is as follows;

[0059] This preferred embodiment builds upon the steps in Example 1 regarding the accumulation and synchronization of adaptable materials in the teaching and research module, further refining the processes for initiating, collaborating on, and accumulating knowledge in teaching and research activities. This approach aims to promote collaboration and experience sharing among teachers and efficiently integrate validated teaching materials into the system's knowledge base and resource repository, thereby continuously optimizing the quality and adaptability of teaching resources.

[0060] Specifically, the process of accumulating and sharing teaching and research activities can include the following steps:

[0061] First, the teaching and research activities are created by the initiator. The initiator can be a teacher delivering the lesson or a teaching and research administrator. When creating an activity, the initiator can set the activity's theme, objectives, participant permissions, and duration. Permission settings can include, but are not limited to, allowing participants to view shared spaces, upload materials, post comments, and edit shared documents. Through refined permission management, it can be ensured that teaching and research activities are conducted in an orderly and secure environment.

[0062] Secondly, the system generates a temporary shared space for this teaching and research activity. During the activity, participants can collaborate within this shared space. The shared space supports previewing and categorizing various interaction records, such as:

[0063] Participating teachers can upload and preview original or adapted teaching materials such as courseware, lesson plans, and exercise sets related to the activity theme.

[0064] The system can categorize and display interactive information such as discussion records, comments, voting results, and modification suggestions during the event, according to topic, time, or related materials.

[0065] With authorization, the shared space can be linked to and displayed teaching trajectory data of related courses (such as summaries of learning reports, statistics of frequently asked questions, etc.) to provide data support for teaching and research discussions.

[0066] Finally, after the teaching and research activities conclude, the system will automatically, or upon confirmation from the initiator, synchronize the effective outcomes of the activities to the system's knowledge base and resource repository. The synchronization process may include:

[0067] Teaching materials (including but not limited to courseware, exercises, explanatory videos, and tiered resource packages) that have been discussed, optimized, or verified through teaching and research activities, along with their metadata (such as applicable grade level, knowledge point tags, difficulty level, and description of suitable scenarios), will be stored in the corresponding categories of the resource library.

[0068] Valuable discussion conclusions, teaching strategies, and solutions to common problems generated during teaching and research activities are processed in a structured manner (e.g., extracting keywords and generating summaries) and stored in a knowledge base as a reference for subsequent AI Q&A or resource recommendations.

[0069] The teaching and research activity identifiers from which the materials and knowledge entries are synchronized can be recorded, facilitating traceability and subsequent iterative updates.

[0070] Through the aforementioned mechanism for accumulating and sharing teaching and research activities, the teaching wisdom and experience of individual teachers can be transformed into a system-wide, reusable, and iterative collective knowledge asset. This not only enriches the content of the resource and knowledge bases and improves the adaptability of resources to specific learning and teaching situations, but also provides a higher-quality and more practice-tested solution for addressing differentiated teaching needs in scenarios of uneven resource distribution. Those skilled in the art will understand that the specific functional interfaces of the shared space, the types of interactive tools, and the triggering conditions for material synchronization (such as automatic synchronization or manual confirmation synchronization) can be adjusted according to the actual system design, and these adjustments all fall within the scope of protection of this invention.

[0071] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. An AI-assisted interactive teaching method for educational scenarios with uneven resource distribution, characterized in that, Includes the following steps: Receive course information created by the teaching terminal, and establish a real-time audio and video connection between the teaching terminal and at least one learning terminal based on the course information, wherein the course information includes knowledge points and teaching materials; During the real-time audio and video connection process, the system receives questions input from the learning end, generates answers based on the adapted knowledge base associated with the knowledge point through the AI ​​module, and returns the answers to the learning end. At the same time, the questions are synchronized to the teaching end. The testing module generates test information from the adapted question bank based on the knowledge points and sends it to the learning terminal. After receiving the answer information submitted by the learning terminal, it grades the answer and generates a learning report containing the learning mastery status. Based on the learning progress report and the historical learning trajectory of the learning terminal, matching learning resources are selected from the hierarchical resource library and pushed to the learning terminal; The teaching and research module creates teaching and research activities, receives adaptation materials uploaded by the teaching terminal, and synchronizes the adaptation materials to the adaptation knowledge base and / or the hierarchical resource base after the teaching and research activities are completed.

2. The method according to claim 1, characterized in that, In the first step, establishing the real-time audio and video connection includes: The teaching end initiates a device test to detect the status of audio and video devices, and dynamically enables network adaptation mode based on network bandwidth to adjust the audio and video transmission quality; The learning client enters the corresponding course and receives the real-time audio and video connection. The learning client supports raising hands to apply and submitting questions during the connection process.

3. The method according to claim 1, characterized in that, In the second step, the AI ​​module generates Q&A content based on the adapted knowledge base, including: Semantic analysis is performed on the question information input from the learning end to extract the core questions; Based on the core question, the adapted knowledge base is matched to generate multimodal Q&A content containing at least one of text, voice, or visual elements; The frequency of the question information is counted, and when the frequency exceeds a preset threshold, a reminder message is generated and sent to the teaching terminal.

4. The method according to claim 1, characterized in that, The third step, which involves generating test information through the test module and sending it to the learning terminal, includes: Based on the knowledge points selected by the teaching terminal or auxiliary terminal and the test parameters set, the test information is automatically generated from the adapted question bank. The test information is published to the learning platform, and a time limit for answering the questions is set. It receives the answer information submitted by the learning end, uses a rule engine to grade the objective questions, and uses keyword matching to generate a scoring reference for the subjective questions.

5. The method according to claim 1, characterized in that, The fourth step, selecting matching learning resources from the hierarchical resource library and pushing them to the learning end, includes: Based on the knowledge gaps identified in the learning progress report and the historical learning trajectory, the resource recommendation logic is determined. Based on the resource push logic, matching resources with scene tags are selected from the hierarchical resource library and automatically pushed to the learning terminal; The system receives feedback records from the learning terminal regarding the pushed resources and adjusts the content of subsequent pushed resources based on these feedback records.

6. The method according to claim 1, characterized in that, The fifth step, which involves creating teaching and research activities through the teaching and research module, includes: The initiator creates a teaching and research activity and sets the theme and format of the activity, and invites at least one teaching end as a participating end; The participating terminal uploads the adapted materials to the shared space of the teaching and research module; During the teaching and research activities, the interaction content between the participants was recorded; After the event, the adaptation materials and recorded interactive content will be stored according to categories and synchronized to the adaptation knowledge base and / or the hierarchical resource library.

7. The method according to claim 2, characterized in that, The network adaptation mode dynamically switches the encoding protocol or transmission resolution of audio and video streams based on the real-time detected network bandwidth.

8. The method according to claim 3, characterized in that, The reminder message is sent to the interactive interface of the teaching terminal in the form of a pop-up window, and the reminder message contains a list of frequently asked questions.

9. The method according to claim 4, characterized in that, The learning progress report displays the knowledge mastery rate and list of incorrect questions on the learning end in a visual format, and the visualization results are used in the resource push logic in the fourth step.

10. The method according to claim 5, characterized in that, The adjustment of subsequent push resource content based on feedback records includes: If the feedback record indicates a lack of understanding of the currently pushed resource, then a more basic level resource is selected from the hierarchical resource library for pushing; If the feedback record indicates that the current push resource has been understood, then a higher-level upgrade resource is selected from the hierarchical resource library for push.

Images