An intelligent teaching auxiliary system based on a large language model

Abstract

The application discloses an intelligent teaching auxiliary system based on a large language model, and relates to the technical field of intelligent teaching.The system comprises a progress management module, a resource management module, a teaching service module, a test question test module, a consultation evaluation module and a result auditing module.The intelligent teaching auxiliary system based on the large language model judges the authenticity of the learning condition of a student by matching the test question test result of the student with input information, and corrects the score record by manually analyzing and auditing the test score and the evaluation score, so that the learning condition of the student is more real and accurate, and the system can adjust the teaching arrangement according to the real mastery of the student, thereby effectively avoiding improper teaching arrangement caused by incorrect judgment of the student, and helping to improve the teaching effect and the learning experience.

Description

Technical Field

[0001] This invention relates to the field of intelligent teaching technology, specifically to an intelligent teaching support system based on a large language model. Background Technology

[0002] Intelligent teaching refers to providing students with personalized and intelligent teaching services using artificial intelligence (AI) technology. This teaching model achieves accurate identification and satisfaction of students' individual needs through in-depth analysis and mining of data such as students' learning behavior and grades, combined with intelligent algorithms and big data technology. Moreover, intelligent teaching emphasizes the flexibility and adaptability of the teaching process, and can automatically adjust the teaching content, methods, and pace according to the different learning characteristics, progress, and interests of different students.

[0003] Intelligent teaching support systems, commonly known as computer-aided instruction (CAI) systems, are innovative teaching tools that utilize computer and network technologies to improve teaching efficiency and student learning experience. Their development benefits from the rapid advancements in information technologies such as artificial intelligence, big data, cloud computing, and the Internet of Things. These technologies provide strong support for the digital transformation of education, enabling intelligent teaching support systems to more accurately analyze student learning data and provide more personalized teaching services. Simultaneously, with the continuous development of technologies such as virtual reality (VR) and augmented reality (AR), intelligent teaching support systems can also create more immersive learning experiences for students, enhancing the fun and interactivity of learning. In summary, intelligent teaching support systems play a crucial role in intelligent teaching, providing teachers and students with a more efficient and convenient teaching and learning environment through personalized learning, enriched teaching resources, teaching effectiveness evaluation, continuous feedback and guidance, and teaching management and monitoring.

[0004] However, in actual teaching, although the interactive functions of intelligent teaching systems have improved compared to traditional CAI, they still lack a natural and smooth communication channel between students and computers. The system mainly judges students' mastery level through the information they input, and cannot judge students' true situation through natural communication and observation like human teachers can. This may lead to inappropriate teaching arrangements due to incorrect judgment of students' situation, affecting teaching effectiveness and learning experience.

[0005] Therefore, there is an urgent need to improve this shortcoming. This invention is to study and improve the existing technology and its deficiencies, and provide an intelligent teaching assistance system based on a large language model. Summary of the Invention

[0006] The purpose of this invention is to provide an intelligent teaching assistance system based on a large language model to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an intelligent teaching assistance system based on a large language model, comprising:

[0008] The module includes: progress management, resource management, teaching services, test questions, consultation and evaluation, and result review.

[0009] The progress management module is responsible for developing personalized teaching plans, monitoring students' learning progress, automatically updating students' learning progress, and adjusting teaching content and pace.

[0010] The resource management module is responsible for managing various teaching resources, including textbooks, courseware, videos, and exercises. Based on the teaching plan and students' individual needs, it provides suitable teaching resources to the teaching service module to ensure that the provided teaching resources are in line with the current learning progress and specifically meet students' learning needs.

[0011] The teaching service module is responsible for providing teaching services directly to students, including knowledge explanation, Q&A, and learning guidance. It uses a large language model to generate natural and fluent text or voice responses, simulating the communication style of human teachers.

[0012] The test module is responsible for automatically generating appropriate test questions based on students' learning progress and teaching content to test students' mastery of knowledge points. After the test, the module automatically grades the test papers and provides test scores and detailed explanations.

[0013] The consultation and assessment module is responsible for receiving students' automatically input personal learning information (such as study notes, self-assessment, etc.), analyzing and evaluating it using natural language processing technology, and providing assessment scores and personalized learning suggestions.

[0014] The result review module is responsible for comprehensively analyzing test scores and evaluation scores to determine the student's true learning situation and level. Based on the analysis results, it provides feedback to the teaching service module on the student's learning progress, difficulties, and common mistakes, providing decision support for subsequent teaching arrangements and guiding the optimization and personalized adjustment of teaching services.

[0015] The teaching service module receives information from the progress management module and the result review module, adjusts teaching strategies and content, and interacts with the test module and the consultation and evaluation module to obtain student learning feedback and further improve teaching effectiveness.

[0016] Furthermore, the progress management module specifically includes the following sub-modules:

[0017] The task statistics submodule is responsible for collecting and analyzing students' learning plan tasks. These tasks may come from the teacher's syllabus, the learning paths recommended by the system, or the learning content chosen by the students themselves. It records basic information such as the name, type (e.g., video learning, exercises, reading materials, etc.) and estimated completion time of each task.

[0018] The catalog creation submodule is responsible for using the information provided by the task statistics submodule to divide the learning tasks into different progress nodes and create a detailed learning progress catalog accordingly. The learning progress catalog should clearly show the learning objectives, main tasks and expected results of each stage.

[0019] Progress tracking submodule: Responsible for tracking students' learning progress in real time, specifically by evaluating their learning progress through their online learning behaviors (such as video viewing time, completion of exercises, and submission time of assignments), and generating progress reports;

[0020] Progress Update Submodule: Responsible for automatically updating the learning progress catalog based on the real-time progress information of students provided by the progress tracking submodule; when it is found that a student's progress is ahead or behind, this submodule can automatically adjust the difficulty of subsequent tasks or recommend supplementary learning resources to maintain the continuity and effectiveness of the learning progress.

[0021] Furthermore, the progress management module specifically includes the following sub-modules:

[0022] The resource download submodule is responsible for automatically downloading the teaching resources required for the upcoming learning stage from the cloud resource library and storing them in the local resource library. This submodule ensures that students can obtain the learning materials they need in a timely manner during the learning process without having to manually search or wait, thus improving the continuity and efficiency of learning. The resource download submodule supports the function of resuming interrupted downloads. If the download is interrupted, it will continue to download after reconnection.

[0023] The compression and upload submodule is responsible for compressing teaching resources that have been studied, are no longer frequently used, or have expired in the local resource library and uploading them to cloud storage. This submodule helps to free up local storage space while maintaining a long-term backup of teaching resources in the cloud, facilitating subsequent management and reuse. For resources that have been partially uploaded, the compression and upload submodule supports incremental upload functionality, uploading only the newly added or modified parts, reducing upload time and data consumption.

[0024] Furthermore, the workflow of the resource download submodule is as follows:

[0025] Analyze the learning progress catalog: Analyze the current learning progress catalog to determine the upcoming learning stages and corresponding teaching resource requirements;

[0026] Cloud resource retrieval: Based on your needs, retrieve relevant teaching resources from the cloud resource library, including video tutorials, e-books, exercise books, etc.

[0027] Resource download and verification: Download the retrieved resources and perform integrity verification to ensure the integrity and correctness of the resources;

[0028] Local storage: Downloaded and verified resources are stored in the local resource library for students to use in their subsequent studies.

[0029] Furthermore, the workflow of the compressed upload submodule is as follows:

[0030] Resource filtering: Filter resources that need to be compressed and uploaded based on preset rules (such as learning completion time, access frequency, etc.);

[0031] Resource compression: The selected resources are compressed to reduce file size, making them easier to store and transmit;

[0032] Cloud upload: Upload the compressed resources to the cloud resource library and ensure the accessibility of the uploaded resources in the cloud;

[0033] Local cleanup: After confirming that the resources have been successfully uploaded to the cloud, delete the corresponding resource files from the local resource library to free up storage space.

[0034] Furthermore, both the test module and the consultation assessment module convert the scores to a percentage system using a conversion formula before outputting the results. The conversion formula is as follows:

[0035]

[0036] Furthermore, the result review module specifically includes the following sub-modules:

[0037] The score calculation submodule is responsible for calculating the match rate between the test score and the evaluation score. It receives the test score from the test module and the evaluation score from the consultation evaluation module, then uses a formula to calculate the degree of match between the two and outputs the match rate between the test score and the evaluation score.

[0038] The matching judgment submodule is responsible for judging the authenticity of the student's learning situation reflected by the test results and the evaluation results based on the matching rate between the test scores and the evaluation scores.

[0039] Manual review submodule: When the matching judgment submodule determines that the test score and the evaluation score do not match, a human teacher is invited to participate in the review process to review the test score and the evaluation score. The human teacher judges whether the score is accurate based on their professional knowledge and experience, and makes necessary adjustments or corrections.

[0040] The results analysis submodule is responsible for analyzing test scores and evaluation scores to extract detailed information about students' learning progress. It considers factors such as students' strengths, weaknesses, learning progress, and mastery level, and outputs a comprehensive learning evaluation report, including learning outcomes, existing problems, and improvement suggestions.

[0041] Furthermore, the formula for calculating the matching rate of the score calculation submodule is as follows:

[0042]

[0043] Wherein, CR represents the matching rate, Score1 represents the test score output by the test module, and Score2 represents the evaluation score output by the consultation and assessment module.

[0044] Furthermore, the matching judgment submodule's judgment condition is "CR≥95%?", and the corresponding instruction for the judgment result is as follows: if CR≥95%, the test score and the evaluation score are considered to match, and the student's learning situation reflected is real, directly triggering the result analysis submodule; conversely, if CR<95%, the test score and the evaluation score are considered to not match, and the student's learning situation reflected is incorrect, triggering the manual review submodule.

[0045] Furthermore, the workflow of the manual review submodule is as follows: receiving instructions and relevant data from the matching judgment submodule, assigning review tasks to appropriate human teachers, having the human teachers review the test scores and evaluation scores, provide review comments, and then updating the score records based on the review comments.

[0046] This invention provides an intelligent teaching assistance system based on a large language model, which has the following beneficial effects:

[0047] 1. This invention judges the authenticity of the student's learning situation by matching the student's test results with the input information. In cases where test scores and assessment scores do not match, human teachers conduct manual analysis and review to correct the score records, thereby obtaining a more realistic and accurate understanding of the student's learning situation. This allows the system to adjust the teaching arrangements according to the student's true mastery of the teaching content, effectively avoiding inappropriate teaching arrangements caused by misjudgments of student situations, and helping to improve teaching effectiveness and learning experience.

[0048] 2. This invention is designed with a teaching resource loading control mechanism, which can load subsequent teaching resources in an appropriate amount according to the student's learning progress. This ensures that the teaching resources are sufficient to support the student's learning progress, and effectively avoids the problem of affecting the device's operating speed due to loading too many teaching resources at once. Furthermore, expired teaching resources are compressed and uploaded to the cloud, further reducing the device's resource consumption and improving the operating speed of the intelligent teaching device. Attached Figure Description

[0049] Figure 1 This is a logical block diagram of an intelligent teaching assistance system based on a large language model according to the present invention.

[0050] Figure 2 This is a schematic diagram illustrating the operation flow of the progress management module of an intelligent teaching support system based on a large language model according to the present invention.

[0051] Figure 3 This is a schematic diagram of the operation flow of the result review module of an intelligent teaching assistance system based on a large language model according to the present invention. Detailed Implementation

[0052] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0053] like Figures 1-3 As shown, an intelligent teaching assistance system based on a large language model includes:

[0054] Progress Management Module: This module creates personalized learning plans, monitors students' learning progress, automatically updates students' learning progress, and adjusts teaching content and pace. The Progress Management Module specifically includes the following sub-modules:

[0055] Task Statistics Submodule: Collects and analyzes students' learning plan tasks. These tasks may come from the teacher's syllabus, the learning path recommended by the system, or the learning content chosen by the students themselves. It records basic information such as the name, type (e.g., video learning, exercises, reading materials, etc.) and estimated completion time of each task.

[0056] The information provided by the task statistics submodule divides the learning tasks into different progress nodes and creates a detailed learning progress catalog accordingly. The learning progress catalog should clearly show the learning objectives, main tasks and expected results of each stage.

[0057] Progress tracking submodule: Tracks students' learning progress in real time, specifically by evaluating their learning progress through their online learning behaviors (such as video viewing time, completion of exercises, and homework submission time), and generates progress reports;

[0058] Progress Update Submodule: Based on the real-time progress information of students provided by the Progress Tracking Submodule, the learning progress catalog is automatically updated. When it is found that a student's progress is ahead or behind, this submodule can automatically adjust the difficulty of subsequent tasks or recommend supplementary learning resources to maintain the continuity and effectiveness of the learning progress.

[0059] Resource Management Module: Manages various teaching resources, including textbooks, courseware, videos, and exercises. Based on the teaching plan and students' individual needs, it provides suitable teaching resources to the teaching service module, ensuring that the provided resources align with the current learning progress and specifically meet students' learning needs. The progress management module specifically includes the following sub-modules:

[0060] The resource download submodule automatically downloads the teaching resources required for the upcoming learning stage from the cloud resource library and stores them in the local resource library. This submodule ensures that students can obtain the learning materials they need in a timely manner during the learning process without having to manually search or wait, thus improving the continuity and efficiency of learning. The resource download submodule supports the function of resuming interrupted downloads. If the download is interrupted, it will continue to download after reconnection.

[0061] The compression upload submodule compresses and uploads teaching resources that have been studied, are no longer frequently used, or have expired from the local resource library to cloud storage. This submodule helps free up local storage space while maintaining a long-term backup of teaching resources in the cloud, facilitating subsequent management and reuse. For partially uploaded resources, the compression upload submodule supports incremental upload functionality, uploading only the newly added or modified parts, reducing upload time and bandwidth consumption.

[0062] The workflow of the resource download submodule is as follows:

[0063] Analyze the learning progress catalog: Analyze the current learning progress catalog to determine the upcoming learning stages and corresponding teaching resource requirements;

[0064] Cloud resource retrieval: Based on your needs, retrieve relevant teaching resources from the cloud resource library, including video tutorials, e-books, exercise books, etc.

[0065] Resource download and verification: Download the retrieved resources and perform integrity verification to ensure the integrity and correctness of the resources;

[0066] Local storage: Downloaded and verified resources are stored in the local resource library for students to use in subsequent learning;

[0067] The workflow of the compressed upload submodule is as follows:

[0068] Resource filtering: Filter resources that need to be compressed and uploaded based on preset rules (such as learning completion time, access frequency, etc.);

[0069] Resource compression: The selected resources are compressed to reduce file size, making them easier to store and transmit;

[0070] Cloud upload: Upload the compressed resources to the cloud resource library and ensure the accessibility of the uploaded resources in the cloud;

[0071] Local cleanup: After confirming that the resources have been successfully uploaded to the cloud, delete the corresponding resource files from the local resource library to free up storage space;

[0072] Teaching Service Module: Directly provides teaching services to students, including knowledge explanation, Q&A, and learning guidance. It uses a large language model to generate natural and fluent text or voice responses, simulating the communication style of human teachers. The Teaching Service Module receives information from the Progress Management Module and the Result Review Module to adjust teaching strategies and content. At the same time, it interacts with the Test Module and the Consultation and Assessment Module to obtain students' learning feedback and further improve teaching effectiveness.

[0073] Test module: Based on students' learning progress and teaching content, it automatically generates test questions of appropriate difficulty to test students' mastery of knowledge points. After the test, it automatically grades the test paper and gives the test score and detailed explanation.

[0074] Consultation and assessment module: Receives students' automatically input personal learning information (such as study notes, self-assessment, etc.), analyzes and evaluates it using natural language processing technology, and provides assessment scores and personalized learning suggestions;

[0075] Before outputting the scores, both the test module and the consultation assessment module convert the scores to a percentage system using a conversion formula:

[0076]

[0077] The results review module comprehensively analyzes test scores and assessment scores to determine the student's true learning situation and level. Based on the analysis results, it provides feedback to the teaching service module on the student's learning progress, difficulties, and common mistakes, providing decision support for subsequent teaching arrangements and guiding the optimization and personalized adjustment of teaching services. The results review module specifically includes the following sub-modules:

[0078] The score calculation submodule calculates the match rate between the test score and the evaluation score. It receives the test score from the test module and the evaluation score from the evaluation module, then uses a formula to calculate the degree of match between the two, and outputs the match rate. The formula for calculating the match rate is: Where CR represents the accuracy rate, Score1 represents the test score output by the test module, and Score2 represents the assessment score output by the consultation assessment module.

[0079] The matching judgment submodule: Based on the matching rate between test scores and assessment scores, it judges the authenticity of the student's learning situation reflected by the test results and assessment results. The judgment condition of the matching judgment submodule is "CR≥95%?", and the corresponding instructions for the judgment result are as follows: If CR≥95%, the test score and assessment score are considered to match, and the student's learning situation reflected is true, directly triggering the result analysis submodule; conversely, if CR<95%, the test score and assessment score are considered to not match, and the student's learning situation reflected is incorrect, triggering the manual review submodule.

[0080] The manual review submodule: When the matching judgment submodule determines that the test score and the evaluation score do not match, a human teacher is invited to participate in the review process to verify the test score and evaluation score. The human teacher uses their professional knowledge and experience to judge whether the score is accurate and makes necessary adjustments or corrections. The workflow of the manual review submodule is as follows: receiving instructions and relevant data from the matching judgment submodule, assigning review tasks to appropriate human teachers, reviewing the test score and evaluation score, providing review comments, and then updating the score record based on the review comments.

[0081] The results analysis submodule delves into test scores and evaluation scores to extract detailed information about students' learning, taking into account factors such as students' strengths, weaknesses, learning progress, and mastery levels, and outputs a comprehensive learning evaluation report, including learning outcomes, existing problems, and improvement suggestions.

[0082] The embodiments of the present invention are given for illustrative and descriptive purposes only, and are not intended to be exhaustive or to limit the invention to the forms disclosed. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to better illustrate the principles and practical application of the invention, and to enable those skilled in the art to understand the invention and to design various embodiments with various modifications suitable for a particular purpose.

Claims

1. An intelligent teaching support system based on a large language model, characterized in that, include: The module includes: progress management, resource management, teaching services, test questions, consultation and evaluation, and result review. The progress management module is responsible for developing personalized teaching plans, monitoring students' learning progress, automatically updating students' learning progress, and adjusting teaching content and pace. The resource management module is responsible for managing various teaching resources, including textbooks, courseware, videos, and exercises. It provides appropriate teaching resources based on the teaching plan and students' individual needs, ensuring that the provided teaching resources are in line with the current learning progress. The teaching service module is responsible for providing teaching services directly to students, using a large language model to generate natural and fluent text or voice responses, simulating the communication style of human teachers. The test module is responsible for automatically generating test questions of appropriate difficulty based on students' learning progress and teaching content, and automatically grading the test papers after the test, providing test scores and detailed explanations; The consultation and assessment module is responsible for receiving students' automatically inputted personal learning information, analyzing and evaluating it using natural language processing technology, and providing assessment scores and personalized learning suggestions. The result review module is responsible for comprehensively analyzing test scores and evaluation scores to determine the student's true learning situation and level. Based on the analysis results, it provides feedback to the teaching service module on the student's learning progress, difficulties, and common mistakes, providing decision support for subsequent teaching arrangements and guiding the optimization and personalized adjustment of teaching services. The teaching service module receives information from the progress management module and the result review module, adjusts teaching strategies and content, and interacts with the test module and the consultation and evaluation module to obtain students' learning feedback and further improve teaching effectiveness. The progress management module specifically includes the following sub-modules: Task Statistics Submodule: Responsible for collecting and analyzing students' learning plan tasks, and recording basic information for each task; The catalog creation submodule is responsible for using the information provided by the task statistics submodule to divide the learning tasks into different progress nodes and create a detailed learning progress catalog accordingly. Progress tracking submodule: Responsible for tracking students' learning progress in real time, specifically by assessing their learning progress through their online learning behavior and generating progress reports; Progress Update Submodule: Responsible for automatically updating the learning progress catalog based on the real-time progress information of students provided by the progress tracking submodule; when it is found that a student's progress is ahead or behind, it will automatically adjust the difficulty of subsequent tasks or recommend supplementary learning resources. The progress management module specifically includes the following sub-modules: The resource download submodule is responsible for automatically downloading the teaching resources required for the upcoming learning stage from the learning progress directory in the cloud resource library and storing them in the local resource library. The resource download submodule supports the function of resuming interrupted downloads. If the download is interrupted, it will continue to download after reconnection. The compression and upload submodule is responsible for compressing teaching resources that have been studied, are no longer frequently used, or have expired in the local resource library and uploading them to cloud storage. For resources that have been partially uploaded, the compression and upload submodule supports incremental upload functionality, uploading only the newly added or modified parts.

2. The intelligent teaching assistance system based on a large language model according to claim 1, characterized in that, The workflow of the resource download submodule is as follows: Analyze the learning progress catalog: Analyze the current learning progress catalog to determine the upcoming learning stages and corresponding teaching resource requirements; Cloud resource retrieval: Based on your needs, retrieve relevant teaching resources from the cloud resource library, including video tutorials, e-books, and exercise sets; Resource download and verification: Download the retrieved resources and perform integrity verification to ensure the integrity and correctness of the resources; Local storage: Downloaded and verified resources are stored in the local resource library for students to use in their subsequent studies.

3. The intelligent teaching assistance system based on a large language model according to claim 2, characterized in that, The workflow of the compressed upload submodule is as follows: Resource filtering: Filter out resources that need to be compressed and uploaded based on preset rules; Resource compression: The selected resources are compressed to reduce file size; Cloud upload: Upload the compressed resources to the cloud resource library and ensure the accessibility of the uploaded resources in the cloud; Local cleanup: After confirming that the resources have been successfully uploaded to the cloud, delete the corresponding resource files from the local resource library to free up storage space.

4. The intelligent teaching assistance system based on a large language model according to claim 1, characterized in that, Before outputting the scores, both the test module and the consultation assessment module convert the scores to a percentage system using a conversion formula:

5. The intelligent teaching assistance system based on a large language model according to claim 1, characterized in that, The result review module specifically includes the following sub-modules: The score calculation submodule is responsible for calculating the match rate between the test score and the evaluation score. The matching judgment submodule is responsible for judging the authenticity of the student's learning situation reflected by the test results and the evaluation results based on the matching rate between the test scores and the evaluation scores. Manual review submodule: When the matching judgment submodule determines that the test score and the evaluation score do not match, a human teacher is invited to participate in the review process to review the test score and evaluation score and make necessary adjustments or corrections. The results analysis submodule is responsible for analyzing test scores and evaluation scores to extract detailed information about students' learning progress and output a comprehensive learning assessment report.

6. The intelligent teaching assistance system based on a large language model according to claim 5, characterized in that, The formula for calculating the matching rate of the score calculation submodule is as follows: Wherein, CR represents the matching rate, Score1 represents the test score output by the test module, and Score2 represents the evaluation score output by the consultation and assessment module.

7. The intelligent teaching assistance system based on a large language model according to claim 6, characterized in that, The matching judgment submodule has the judgment condition "CR≥95%?", and the corresponding instructions for the judgment result are as follows: if CR≥95%, the test score and the evaluation score are considered to match, the student's learning situation reflected is real, and the result analysis submodule is directly triggered; otherwise, if CR<95%, the test score and the evaluation score are considered to not match, the student's learning situation reflected is incorrect, and the manual review submodule is triggered.

8. The intelligent teaching assistance system based on a large language model according to claim 7, characterized in that, The workflow of the manual review submodule is as follows: receiving instructions and relevant data from the matching judgment submodule, assigning review tasks to appropriate human teachers, reviewing test scores and evaluation scores, providing review comments, and then updating the score records based on the review comments.

Images