A teaching method and platform based on a multi-modal knowledge base

Abstract

The present application relates to the technical field of knowledge teaching, and particularly relates to a teaching method and platform based on a multi-modal knowledge base, comprising: associating multi-modal resources of the multi-modal knowledge base with course knowledge points to form knowledge point cards, creating a challenge test, and determining a first knowledge point mastery degree and weak knowledge points; according to the knowledge point cards of the weak knowledge points, using an AI tutor to teach, and obtaining a second knowledge point mastery degree; based on the second knowledge point mastery degree, generating a learning situation analysis result, and adjusting teaching content. The present application integrates text, image, audio, video and test question resources through the multi-modal knowledge base to form knowledge point cards, and realizes structured management and accurate tracking of knowledge points. In combination with the AI tutor, weak knowledge points are guided and retested, the knowledge mastery degree is dynamically updated, the learning situation analysis is generated and the teaching content is adjusted, and closed-loop learning is realized.

Description

Technical Field

[0001] This invention relates to the field of knowledge teaching technology, specifically to a teaching method and platform based on a multimodal knowledge base. Background Technology

[0002] With the development of educational informatization and intelligent teaching, more and more schools and online education platforms are trying to support teaching through digital resources.

[0003] However, existing teaching methods and systems still have many shortcomings in personalized teaching, learning process tracking, and knowledge mastery analysis: Teaching resources are often limited in form; current teaching systems rely heavily on single-format materials such as text or PowerPoint presentations, lacking the integration of multimodal resources like images, videos, audio, and test questions. This makes it difficult for students to gain a comprehensive and diverse understanding of knowledge. Knowledge point connections are insufficient; traditional teaching platforms typically present course content by chapter or unit, lacking structured management and multimodal resource connections for specific knowledge points. This makes it difficult for teachers to accurately identify students' weak knowledge areas and provide targeted teaching content. A closed-loop learning mechanism is lacking; current teaching models are typically a one-way "teaching-testing-assessment" process. Students' learning processes lack continuous tracking and dynamic adjustment, making it difficult to achieve cyclical optimization of teaching based on knowledge mastery and fully realize the value of educational data. Summary of the Invention

[0004] (a) Purpose of the invention The purpose of this invention is to provide a teaching method and platform based on a multimodal knowledge base. This method integrates text, images, audio, video, and test question resources through the multimodal knowledge base to form knowledge point cards, enabling structured management and precise tracking of knowledge points. Combined with AI teaching assistants, it provides tutoring and retesting for weak knowledge points, dynamically updates knowledge mastery, generates learning analysis, and adjusts teaching content to achieve closed-loop learning. This method improves the relevance and efficiency of student learning while providing teachers with a scientific basis for teaching decisions, thereby enhancing overall teaching quality.

[0005] (II) Technical Solution To address the above problems, this invention provides a teaching method based on a multimodal knowledge base, comprising: Identify the course knowledge points corresponding to the course class; The multimodal resources of the multimodal knowledge base are associated with the course knowledge points to form knowledge point cards; Based on the multimodal knowledge base and the knowledge point cards, a challenge test is created; Obtain the results of the challenge test, and based on the results, determine the mastery level of the first knowledge point and the weak knowledge points; Based on the knowledge point cards for the identified weak knowledge areas, teaching is conducted using AI teaching assistants; Based on the retest results of the weak knowledge points, update the mastery level of the first knowledge point to obtain the mastery level of the second knowledge point; Based on the mastery of the second knowledge point, a learning analysis result is generated, and the teaching content is adjusted accordingly.

[0006] In another aspect of the present invention, preferably, determining the course knowledge points corresponding to the course class includes: Obtain the course syllabus corresponding to the class and divide the course syllabus into knowledge units; Extract the features of the knowledge units and connect them hierarchically to obtain a knowledge point structure tree; The knowledge point structure tree consists of course knowledge points corresponding to the course class.

[0007] In another aspect of the present invention, preferably, the step of associating the multimodal resources of the multimodal knowledge base with the course knowledge points to form knowledge point cards includes: Obtain multimodal resources from the multimodal knowledge base, extract features from the multimodal resources, and obtain resource features; Calculate the similarity between the resource features and the knowledge point structure tree to obtain a similarity value; Based on the similarity value and the preset similarity threshold, the association between the multimodal resources and the corresponding course knowledge points is established; Based on the preset knowledge point card template, the associated multimodal resources are structurally encapsulated to form knowledge point cards.

[0008] In another aspect of the present invention, preferably, the step of creating a challenge test based on the multimodal knowledge base and the knowledge point cards includes: Obtain a set of test question resources related to the course knowledge points from a multimodal knowledge base; The test question selection mode includes manual selection mode and automatic selection mode based on knowledge points; In the manual selection mode, the test question resources are filtered according to the knowledge point cards to determine the test question set; In the automatic selection mode, questions are automatically extracted from the question resources associated with the corresponding knowledge points based on the knowledge point cards according to preset extraction rules to form a question set; Based on the set of questions, a challenge test is generated, and a correspondence is established between the challenge test and the course knowledge points.

[0009] In another aspect of the present invention, preferably, obtaining the results of a challenge test, and determining the mastery level of a first knowledge point and weak knowledge points based on the results of the challenge test, includes: Obtain students' answer data in the challenge test; Based on the correspondence between each question in the challenge test and the course knowledge points, the answer data is mapped to the corresponding course knowledge points, and the accuracy and score rate of each course knowledge point are statistically analyzed. Calculate the mastery level of the first knowledge point of each course based on the accuracy and score rate of each knowledge point. Based on the mastery level of the first knowledge point and the preset mastery level threshold, knowledge points are screened to identify weak knowledge points.

[0010] In another aspect of the present invention, preferably, the step of using AI teaching assistants to teach based on knowledge point cards for the weak knowledge points includes: Obtain the knowledge point card corresponding to the weak knowledge point; Based on the knowledge point definitions, associated multimodal resources, and test question information in the knowledge point cards, teaching prompt information is constructed; Input the teaching prompts into the AI ​​teaching assistant model to generate explanations for weak knowledge points; Based on the content of the explanation, corresponding knowledge point cards and related multimodal resources are pushed to students, and interactive Q&A tutoring is provided. Record interaction data and learning behavior data with the AI ​​teaching assistant to create learning records targeting weak knowledge points.

[0011] In another aspect of the present invention, preferably, updating the mastery level of the first knowledge point based on the retest results of the weak knowledge points to obtain the mastery level of the second knowledge point includes: Obtain retest answer data for the weak knowledge points, the retest answer data including question identifiers, answer results and score information; Based on the correspondence between retest questions and weak knowledge points, the retest answer data is mapped to the corresponding knowledge points, and the retest score rate of each weak knowledge point is calculated. Based on the retest score rate and the mastery of the first knowledge point, the mastery of the corresponding knowledge point is corrected according to the preset update rules to obtain the mastery of the second knowledge point.

[0012] In another aspect of the present invention, preferably, the retest result is obtained through the following steps: Retrieve test question resources associated with the weak knowledge points from the multimodal knowledge base; According to the preset random sampling rules, test questions are randomly selected from the test question resources; The retest results are generated based on the answers to the randomly selected test questions.

[0013] In another aspect of the present invention, preferably, the step of generating learning analysis results based on the mastery level of the second knowledge point and adjusting the teaching content accordingly includes: Obtain the mastery level of the second knowledge point for all students in the course class; Statistical analysis was performed on the mastery data to calculate the average mastery of each knowledge point, the distribution of weak knowledge points, and the fluctuation of mastery. A learning progress analysis report is generated based on the statistical analysis results. Based on the learning analysis report, a teaching content adjustment plan is generated to address the weak knowledge points.

[0014] In another aspect, preferably, a teaching platform based on a multimodal knowledge base includes: The first module: Determine the course knowledge points corresponding to the course class; Association module: Associates the multimodal resources of the multimodal knowledge base with the course knowledge points to form knowledge point cards; Create a module: Based on the multimodal knowledge base and the knowledge point cards, create a challenge test; The second determination module: obtains the test results and, based on the test results, determines the mastery level of the first knowledge point and the weak knowledge points; Teaching module: Based on the knowledge point cards for the identified weak knowledge points, teaching is conducted using AI teaching assistants; Update module: Based on the retest results of the weak knowledge points, update the mastery level of the first knowledge point and obtain the mastery level of the second knowledge point; Adjustment module: Based on the mastery of the second knowledge point, generate learning analysis results and adjust the teaching content.

[0015] (III) Beneficial Effects The above-described technical solution of the present invention has the following beneficial technical effects: This invention utilizes a multimodal knowledge base to link various types of teaching resources, such as text, images, audio, video, and test questions, with course knowledge points, forming structured knowledge point cards. This enables unified management and efficient utilization of teaching resources. Through knowledge point-based challenge tests and a knowledge mastery calculation mechanism, it allows for refined analysis of student learning progress and accurate identification of weak knowledge points. AI teaching assistants provide targeted explanations and learning guidance for these weak points, and a retesting mechanism dynamically evaluates student learning outcomes, continuously updating knowledge mastery levels. This provides teachers with visualized teaching feedback and decision-making support, enabling dynamic adjustment and optimization of teaching content. Ultimately, this significantly improves the relevance, personalization, and overall quality of teaching, while simultaneously enhancing students' learning efficiency and knowledge mastery. Attached Figure Description

[0016] Figure 1 This is an overall flowchart of one embodiment of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and the accompanying drawings. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of the invention. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0018] Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0019] In the description of this invention, it should be noted that the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0020] Furthermore, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0021] The invention will now be described in more detail with reference to the accompanying drawings. In the various drawings, the same elements are indicated by similar reference numerals. For clarity, the various parts in the drawings are not drawn to scale.

[0022] Example 1 A teaching method based on a multimodal knowledge base. Figure 1 An overall flowchart of one embodiment of the present invention is shown, as follows: Figure 1 As shown, it includes: Identifying the course knowledge points corresponding to the course class; a course class refers to a teaching organization unit established within a teaching platform for a specific course, used to carry the course's teaching activities, student member information, and learning process data. A course class includes: a set of students enrolled in or participating in the course, a corresponding teacher or teaching administrator, the course's teaching progress, testing tasks, and learning records. In this embodiment, identifying the course knowledge points corresponding to the course class includes: Obtain the course syllabus corresponding to the class, and divide the syllabus into knowledge units. The syllabus can be obtained from the course management module of the teaching platform, course documents uploaded by teachers, or the school's teaching management system. The course syllabus may include information such as course objectives, chapter arrangement, description of teaching content, knowledge scope, and ability requirements. Divide the course syllabus into knowledge units, based on the chapter structure, topic titles, or content semantics in the syllabus, and divide the teaching content into multiple relatively independent knowledge units.

[0023] Features of the knowledge units are extracted and connected hierarchically to obtain a knowledge point structure tree; the knowledge point structure tree represents the course knowledge points corresponding to the course class. Content parsing is performed on each knowledge unit to extract feature information describing the knowledge unit. This feature information may include the knowledge unit name, keywords, topic semantic representation, chapter position, dependencies, and knowledge point type. The parent-child hierarchical relationship of the knowledge units is determined based on the chapter structure in the course syllabus; for example, a higher-level chapter corresponds to a parent node, and a lower-level knowledge unit corresponds to a child node. For knowledge units without explicit hierarchical identifiers, their connection relationships can be determined based on semantic similarity or dependency rules; for example, knowledge units with high semantic relevance and later teaching order can be used as child nodes or associated nodes of preceding knowledge units. Using the overall course knowledge unit as the root node, each knowledge unit is connected hierarchically according to parent-child relationships to form a tree-like hierarchical structure. Each node corresponds to a knowledge unit and stores the feature information of that knowledge unit and its connection relationships with other nodes.

[0024] The multimodal resources of the multimodal knowledge base are associated with the course knowledge points to form knowledge point cards, including: The process involves acquiring multimodal resources from the multimodal knowledge base, extracting features from these resources, and obtaining resource features. Multimodal resources can include at least one type, such as text materials, teaching materials, images, audio, video, and test questions. For text resources, semantic features can be obtained through word segmentation, keyword extraction, and semantic vector representation. For image or video resources, object information or topic features can be obtained through visual feature extraction methods. For audio resources, semantic features can be extracted based on speech recognition results. For test question resources, knowledge point tags and question stem semantic features can be extracted. Through the above processing, different types of resources are converted into a unified feature representation for subsequent matching processing.

[0025] The similarity between the resource features and the knowledge point structure tree is calculated to obtain a similarity value. The semantic feature representations corresponding to each knowledge point in the knowledge point structure tree are obtained, and the resource features and knowledge point features are matched to determine the degree of association between the resource and the knowledge point. The similarity value is used to characterize the semantic relevance between the multimodal resource and the target knowledge point. Cosine similarity can be used for calculation.

[0026] Based on the similarity value and a preset similarity threshold, an association is established between the multimodal resource and the corresponding course knowledge point. When the similarity value between a resource and a knowledge point is greater than or equal to the preset threshold, it is determined that the resource and the knowledge point are associated, and a mapping relationship between the resource identifier and the knowledge point identifier is established. If a resource and multiple knowledge points meet the association conditions, a many-to-many association relationship can be established.

[0027] Following a pre-defined knowledge point card template, associated multimodal resources are structurally encapsulated to form knowledge point cards. These cards uniformly present knowledge point information and its associated resources. Card content may include a knowledge point identifier, name, description, a list of associated multimodal resources, and relevant test question information. The generated knowledge point cards are stored in a multimodal knowledge base and linked to course classes and corresponding knowledge points to support subsequent teaching, testing, and learning analysis.

[0028] Based on the multimodal knowledge base and the knowledge point cards, a challenge test is created, including: The system retrieves a set of test question resources associated with course knowledge points from a multimodal knowledge base; it reads the set of course knowledge points corresponding to the course class, and retrieves test question resources associated with each knowledge point from the multimodal knowledge base based on the knowledge point identifier. The test question resources may include multiple-choice, true / false, fill-in-the-blank, and short-answer questions. The retrieved test question resources are then aggregated to form a candidate test question resource set, retaining data such as test question identifiers, test question content, question type information, and their corresponding knowledge point identifiers.

[0029] The test can be selected based on a choice of modes, including manual selection and automatic selection based on knowledge points; the mode can be set by the teacher when creating the challenge test.

[0030] In the manual selection mode, test question resources are filtered based on the knowledge point cards to determine the test question set. Teachers can filter candidate test question resources according to course chapters, knowledge point identifiers, or knowledge point names, and select target test questions from the filtering results. After receiving the teacher's test question selection instruction, the system summarizes the selected test questions to form a test question set and records the mapping relationship between each test question and its corresponding knowledge point.

[0031] In the automatic selection mode, questions are automatically extracted from the question resources associated with the corresponding knowledge points based on the knowledge point cards, according to preset extraction rules, to form a question set. The preset extraction rules may include at least one of the following: question quantity requirements, question type ratio requirements, and difficulty distribution requirements. The system selects questions that meet the conditions from the candidate questions according to the extraction rules to form a question set.

[0032] Based on the aforementioned question set, a challenge test is generated. The question set is arranged according to a preset test paper structure, and a challenge test paper is generated. A correspondence is established between the challenge test and the course knowledge points. The challenge test can be set up with a test structure containing multiple levels, each level corresponding to one or more knowledge points. The system simultaneously establishes the correspondence between the challenge test and the course knowledge points and stores this correspondence in the system to support subsequent analysis of answer results and calculation of knowledge point mastery.

[0033] Obtain the test results, and based on these results, determine the mastery level of the first knowledge point and the weak knowledge points, including: Acquire student response data during the challenge test; after students complete the challenge test, record their response process and result data, which may include question identifiers, student answers, correct / incorrect answers, scores, response time, and level information. Summarize the response data and establish a correlation with the corresponding challenge test papers.

[0034] Based on the correspondence between each question in the challenge test and the course knowledge points, the answer data is mapped to the corresponding course knowledge points, and the accuracy and score rate of each course knowledge point are statistically analyzed. The mapping relationship between questions and knowledge points is read, and the course knowledge point identifier corresponding to each question is determined according to the question identifier. Students' answers to each question are then aggregated under the corresponding knowledge point. For multiple questions corresponding to the same knowledge point, the answer results are summarized and statistically analyzed.

[0035] Based on the accuracy and score rates of each course's knowledge points, the mastery level of the first knowledge point is calculated. The accuracy rate of each knowledge point is obtained by calculating the ratio of the number of correctly answered questions to the total number of questions based on the collected answer results. Simultaneously, the score rate of each knowledge point is calculated by comparing the student's score on the corresponding question to the full score on that question. The accuracy and score rates reflect the student's mastery of the corresponding knowledge points. The mastery level of the first knowledge point can be calculated using a weighted average according to corresponding weights.

[0036] Based on the mastery level of the first knowledge point and a preset mastery threshold, knowledge points are screened to identify weak knowledge points. When the mastery level of the first knowledge point of a certain knowledge point is lower than the preset mastery threshold, the system determines that the knowledge point is a weak knowledge point.

[0037] Based on the knowledge point cards for the identified weak areas, teaching is conducted using an AI teaching assistant, including: Obtain the knowledge point card corresponding to the weak knowledge point; Based on the knowledge point definitions, associated multimodal resources, and test question information in the knowledge point cards, teaching prompts are constructed. Core concepts, key terms, example content, and associated resource identifiers are extracted from the knowledge point cards, and teaching context information is generated by combining this with students' answers in the challenge test. The teaching prompts describe the students' current learning needs and teaching objectives for their weak knowledge points.

[0038] The teaching prompts are input into the AI ​​teaching assistant model to generate explanations targeting weak knowledge points. Specifically, the system uses the teaching prompts as model input, and the AI ​​teaching assistant model generates teaching content such as knowledge explanation text, learning step instructions, or example analyses. The explanations focus on the weak knowledge points to help students understand the relevant knowledge.

[0039] Based on the explained content, the system pushes corresponding knowledge point cards and related multimodal resources to students, and provides interactive Q&A tutoring. The teaching interface displays the explanation content generated by the AI ​​teaching assistant, along with the knowledge point cards and their associated multimodal resources. Students can interact with the AI ​​teaching assistant by entering questions or selecting interactive commands. The system receives student input and calls the AI ​​teaching assistant model to generate feedback information for targeted tutoring.

[0040] Record interaction data and learning behavior data with the AI ​​teaching assistant to create learning records targeting weak knowledge points. Learning behavior data can include information such as student resource viewing, question records, number of interactions, and learning duration.

[0041] Based on the retest results of the weak knowledge points, update the mastery level of the first knowledge point to obtain the mastery level of the second knowledge point, including: Obtain retest response data for the identified weak knowledge points. This retest response data includes question identifiers, response results, and score information. The retest results are obtained through the following steps: Retrieve test question resources associated with the weak knowledge points from the multimodal knowledge base; obtain the corresponding set of test question resources by reading the mapping relationship of the weak knowledge point identifiers in the multimodal knowledge base. The test question resources may include single-choice questions, multiple-choice questions, true / false questions, fill-in-the-blank questions, and short-answer questions to ensure coverage of the core content of the weak knowledge points.

[0042] According to the preset random selection rules, test questions are randomly selected from the test question resources. The preset random selection rules may include requirements such as the number of test questions, the proportion of question types, the distribution of difficulty, and the coverage of knowledge points, so as to ensure that the retest can comprehensively evaluate the students' mastery of weak knowledge points.

[0043] The retest results are generated based on the answers to randomly selected questions. After students complete the retest, their answers and scores for each question are recorded, forming a retest answer data set, which provides a basis for subsequent updates on knowledge mastery.

[0044] Based on the correspondence between retest questions and weak knowledge points, the retest answer data is mapped to the corresponding knowledge points, and the retest score rate of each weak knowledge point is calculated. The retest score rate reflects the degree to which students have mastered the weak knowledge points after the relearning process.

[0045] Based on the retest score rate and the mastery level of the first knowledge point, the mastery level of the corresponding knowledge point is corrected according to a preset update rule to obtain the mastery level of the second knowledge point. The preset update rule can directly cover or incrementally adjust, etc. The mastery level of the second knowledge point can more accurately reflect the student's knowledge mastery level after relearning and retesting.

[0046] Based on the mastery level of the second knowledge point, a learning analysis result is generated, and the teaching content is adjusted, including: Obtain the mastery level of the second knowledge point for all students in the course class to comprehensively reflect the students' mastery of the course knowledge points.

[0047] Statistical analysis was performed on the mastery data to calculate the average mastery of each knowledge point, the distribution of weak knowledge points, and the fluctuation of mastery. A learning analysis report is generated based on the statistical analysis results. This report may include an overview of the class's overall knowledge mastery level, weak knowledge points and their corresponding student distribution, trends in mastery levels, rankings of mastery levels, and analysis of dependencies between knowledge points. The report can be presented in charts or tables to provide teachers with a clear understanding of the class's learning situation. The report can simultaneously support multi-dimensional analysis by student, knowledge point, or chapter.

[0048] Based on the learning analysis report, a teaching content adjustment plan is generated for weak knowledge points. Combining the distribution of weak knowledge points, the class average mastery rate and the fluctuation of mastery rate, personalized or class-wide teaching adjustment plans are proposed, including increasing review time for weak knowledge points, adjusting the focus of classroom explanation, pushing multimodal learning resources for weak knowledge points, assigning targeted exercises or retest tasks, and arranging AI teaching assistants for tutoring.

[0049] This invention utilizes a multimodal knowledge base to link various types of teaching resources, such as text, images, audio, video, and test questions, with course knowledge points, forming structured knowledge point cards. This enables unified management and efficient utilization of teaching resources. Through knowledge point-based challenge tests and a knowledge mastery calculation mechanism, it allows for refined analysis of student learning progress and accurate identification of weak knowledge points. AI teaching assistants provide targeted explanations and learning guidance for these weak points, and a retesting mechanism dynamically evaluates student learning outcomes, continuously updating knowledge mastery levels. This provides teachers with visualized teaching feedback and decision-making support, enabling dynamic adjustment and optimization of teaching content. Ultimately, this significantly improves the relevance, personalization, and overall quality of teaching, while simultaneously enhancing students' learning efficiency and knowledge mastery.

[0050] Example 2 A teaching platform based on a multimodal knowledge base includes: The first module: Determine the course knowledge points corresponding to the course class; Association module: Associates the multimodal resources of the multimodal knowledge base with the course knowledge points to form knowledge point cards; Create a module: Based on the multimodal knowledge base and the knowledge point cards, create a challenge test; The second determination module: obtains the test results and, based on the test results, determines the mastery level of the first knowledge point and the weak knowledge points; Teaching module: Based on the knowledge point cards for the identified weak knowledge points, teaching is conducted using AI teaching assistants; Update module: Based on the retest results of the weak knowledge points, update the mastery level of the first knowledge point and obtain the mastery level of the second knowledge point; Adjustment module: Based on the mastery of the second knowledge point, generate learning analysis results and adjust the teaching content.

[0051] It should be understood that the specific embodiments described above are merely illustrative or explanatory of the principles of the invention and do not constitute a limitation thereof. Therefore, any modifications, equivalent substitutions, improvements, etc., made without departing from the spirit and scope of the invention should be included within the protection scope of the invention. Furthermore, the appended claims are intended to cover all variations and modifications falling within the scope and boundaries of the appended claims, or equivalent forms of such scope and boundaries.

[0052] The above description does not provide detailed explanations of the technical aspects of each layer's patterning and etching. However, those skilled in the art should understand that various methods existing in the prior art can be used to form layers and regions of the desired shape. Furthermore, to form the same structure, those skilled in the art can also design methods that are not entirely identical to those described above.

[0053] The present invention has been described above with reference to embodiments thereof. However, these embodiments are merely illustrative and not intended to limit the scope of the invention. The scope of the invention is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of the invention, and all such substitutions and modifications should fall within the scope of the invention.

[0054] Although embodiments of the present invention have been described in detail, it should be understood that various changes, substitutions, and modifications can be made to the embodiments of the present invention without departing from the spirit and scope of the invention.

[0055] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A teaching method based on a multimodal knowledge base, characterized in that, include: Identify the course knowledge points corresponding to the course class; The multimodal resources of the multimodal knowledge base are associated with the course knowledge points to form knowledge point cards; Based on the multimodal knowledge base and the knowledge point cards, a challenge test is created; Obtain the results of the challenge test, and based on the results, determine the mastery level of the first knowledge point and the weak knowledge points; Based on the knowledge point cards for the identified weak knowledge areas, teaching is conducted using AI teaching assistants; Based on the retest results of the weak knowledge points, update the mastery level of the first knowledge point to obtain the mastery level of the second knowledge point; Based on the mastery of the second knowledge point, a learning analysis result is generated, and the teaching content is adjusted accordingly.

2. The teaching method based on a multimodal knowledge base according to claim 1, characterized in that, The determination of the course knowledge points corresponding to the course class includes: Obtain the course syllabus corresponding to the class and divide the course syllabus into knowledge units; Extract the features of the knowledge units and connect them hierarchically to obtain a knowledge point structure tree; The knowledge point structure tree consists of course knowledge points corresponding to the course class.

3. The teaching method based on a multimodal knowledge base according to claim 2, characterized in that, The step of associating the multimodal resources of the multimodal knowledge base with the course knowledge points to form knowledge point cards includes: Obtain multimodal resources from the multimodal knowledge base, extract features from the multimodal resources, and obtain resource features; Calculate the similarity between the resource features and the knowledge point structure tree to obtain a similarity value; Based on the similarity value and the preset similarity threshold, the association between the multimodal resources and the corresponding course knowledge points is established; Based on the preset knowledge point card template, the associated multimodal resources are structurally encapsulated to form knowledge point cards.

4. The teaching method based on a multimodal knowledge base according to claim 3, characterized in that, The creation of a challenge test based on the multimodal knowledge base and the knowledge point cards includes: Obtain a set of test question resources related to the course knowledge points from a multimodal knowledge base; The test question selection mode includes manual selection mode and automatic selection mode based on knowledge points; In the manual selection mode, the test question resources are filtered according to the knowledge point cards to determine the test question set; In the automatic selection mode, questions are automatically extracted from the question resources associated with the corresponding knowledge points based on the knowledge point cards according to preset extraction rules to form a question set; Based on the set of questions, a challenge test is generated, and a correspondence is established between the challenge test and the course knowledge points.

5. The teaching method based on a multimodal knowledge base according to claim 4, characterized in that, Obtain the test results, and based on these results, determine the mastery level of the first knowledge point and the weak knowledge points, including: Obtain students' answer data in the challenge test; Based on the correspondence between each question in the challenge test and the course knowledge points, the answer data is mapped to the corresponding course knowledge points, and the accuracy and score rate of each course knowledge point are statistically analyzed. Calculate the mastery level of the first knowledge point of each course based on the accuracy and score rate of each knowledge point. Based on the mastery level of the first knowledge point and the preset mastery threshold, knowledge points are screened to identify weak knowledge points.

6. The teaching method based on a multimodal knowledge base according to claim 5, characterized in that, The teaching method, which utilizes AI teaching assistants to teach based on knowledge point cards representing the weak knowledge points, includes: Obtain the knowledge point card corresponding to the weak knowledge point; Based on the knowledge point definitions, associated multimodal resources, and test question information in the knowledge point cards, teaching prompt information is constructed; Input the teaching prompts into the AI ​​teaching assistant model to generate explanations for weak knowledge points; Based on the content of the explanation, corresponding knowledge point cards and related multimodal resources are pushed to students, and interactive Q&A tutoring is provided. Record interaction data and learning behavior data with the AI ​​teaching assistant to create learning records targeting weak knowledge points.

7. The teaching method based on a multimodal knowledge base according to claim 6, characterized in that, The process of updating the mastery level of the first knowledge point based on the retest results of the weak knowledge points to obtain the mastery level of the second knowledge point includes: Obtain retest answer data for the weak knowledge points, the retest answer data including question identifiers, answer results and score information; Based on the correspondence between retest questions and weak knowledge points, the retest answer data is mapped to the corresponding knowledge points, and the retest score rate of each weak knowledge point is calculated. Based on the retest score rate and the mastery of the first knowledge point, the mastery of the corresponding knowledge point is corrected according to the preset update rules to obtain the mastery of the second knowledge point.

8. The teaching method based on a multimodal knowledge base according to claim 7, characterized in that, The retest results were obtained through the following steps: Retrieve test question resources associated with the weak knowledge points from the multimodal knowledge base; According to the preset random sampling rules, test questions are randomly selected from the test question resources; The retest results are generated based on the answers to the randomly selected test questions.

9. The teaching method based on a multimodal knowledge base according to claim 8, characterized in that, The process of generating learning analysis results based on the mastery of the second knowledge point, and adjusting teaching content accordingly, includes: Obtain the mastery level of the second knowledge point for all students in the course class; Statistical analysis was performed on the mastery data to calculate the average mastery of each knowledge point, the distribution of weak knowledge points, and the fluctuation of mastery. A learning progress analysis report is generated based on the statistical analysis results. Based on the learning analysis report, a teaching content adjustment plan is generated to address the weak knowledge points.

10. A teaching platform based on a multimodal knowledge base, characterized in that, include: The first module: Determine the course knowledge points corresponding to the course class; Association module: Associates the multimodal resources of the multimodal knowledge base with the course knowledge points to form knowledge point cards; Create a module: Based on the multimodal knowledge base and the knowledge point cards, create a challenge test; The second determination module: obtains the test results and, based on the test results, determines the mastery level of the first knowledge point and the weak knowledge points; Teaching module: Based on the knowledge point cards for the identified weak knowledge points, teaching is conducted using AI teaching assistants; Update module: Based on the retest results of the weak knowledge points, update the mastery level of the first knowledge point and obtain the mastery level of the second knowledge point; Adjustment module: Based on the mastery of the second knowledge point, generate learning analysis results and adjust the teaching content.

Images