A knowledge space-based teaching optimization method, system, device and medium

Abstract

This invention discloses a teaching optimization method, system, device, and medium based on a knowledge space. The method specifically includes: using the National Industrial Internet Identifier Resolution Secondary Node Platform, parsing official textbooks to generate knowledge point entities with identifier codes and establishing a relationship network; based on the relationship network, collecting problem-solving steps from multi-platform question banks, authenticating them using an identifier code matching mechanism, and synchronizing the authenticated steps to the basic knowledge space and marking them as extended knowledge point databases, while simultaneously calculating the difficulty threshold of knowledge point association combinations based on error statistics; based on the relationship network, capturing learners' cross-platform problem-solving steps in real time, authenticating them using an identifier code matching mechanism, and marking learners as not having mastered the preceding knowledge points if they fail authentication, generating a learner's problem-solving knowledge space with status markings. This invention achieves comprehensive analysis of learners' knowledge point mastery, difficulty identification, and generation of personalized teaching optimization suggestions, improving teaching quality and learning outcomes.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a teaching optimization method, system, device, and medium based on knowledge space. Background Technology

[0002] In today's education field, instructional quality analysis is crucial for improving teaching effectiveness, optimizing teaching strategies, and helping students better master knowledge. Currently, most instructional quality analyses rely primarily on learner behavior data and problem-solving data. However, this analytical approach has many limitations, leading to inaccurate learner analysis and consequently affecting the accuracy of recommended learning plans and test question types.

[0003] On the one hand, there is an issue of incompleteness in acquiring learners' learning behavior data. Learners may complete the learning of a certain knowledge point offline or through other channels or methods, and this learning behavior data cannot be obtained by existing analysis systems. For example, students may participate in tutoring classes, read relevant books, or engage in group discussions outside of class; the knowledge and experience accumulated through these learning behaviors cannot be incorporated into the existing teaching quality analysis system. This results in an incomplete analysis of learners' basic learning-related aspects, making it impossible to accurately grasp the learners' overall learning situation. Ultimately, this leads to biased analysis results that fail to truly reflect learners' knowledge mastery and ability levels.

[0004] On the other hand, many current technologies judge learners based on the accuracy of their answers to a specific question type, thereby analyzing their mastery of related knowledge points. This analytical approach is too general and cannot accurately pinpoint learners' deficiencies in knowledge mastery and application. For example, if a learner answers question A correctly but question B incorrectly, judging solely by question type accuracy cannot determine which specific knowledge point the learner struggled with—it could be in text comprehension, application, or other related factors. Because these deficiencies cannot be precisely identified, recommended learning plans or test types are also imprecise, failing to meet learners' personalized learning needs and hindering their ability to improve knowledge and skills.

[0005] Furthermore, many current technologies analyze students' learning paths based on their learning records on platforms. However, each platform stores its data independently, providing only extracted knowledge points and word segmentation information. This data silo phenomenon makes it impossible to comprehensively and deeply analyze learners' learning processes and knowledge mastery. For example, different platforms may have differences in the division and expression of knowledge points, leading to difficulties in data integration and hindering the formation of a unified and complete learner knowledge profile. Simultaneously, providing only extracted knowledge points and word segmentation information, lacking detailed analysis of problem-solving steps, fails to provide in-depth understanding of learners' thought processes and knowledge application during problem-solving, further limiting the accuracy and effectiveness of teaching quality analysis. Summary of the Invention

[0006] The purpose of this invention is to provide a teaching optimization method, system, device, and medium based on knowledge space. By constructing a knowledge space system, it realizes a comprehensive analysis of learners' mastery of knowledge points, identification of difficulties, and generation of personalized teaching optimization suggestions, thereby improving teaching quality and learning outcomes and solving at least one of the aforementioned problems in the prior art.

[0007] Firstly, the present invention provides a teaching optimization method based on knowledge space, the method specifically including:

[0008] Based on the national industrial internet identifier resolution secondary node platform, the system parses official textbooks to generate knowledge point entities with identifier codes and establishes a relational network that includes sequential, parallel, and comprehensive types.

[0009] Based on the relationship network, the problem-solving steps of question banks from multiple platforms are collected and authenticated by the identification code matching mechanism. After authentication, the steps are synchronized to the basic knowledge space and marked as the extended knowledge point library. At the same time, the difficulty threshold of knowledge point association combination is calculated by combining the error statistics.

[0010] Based on the relationship network, learners' cross-platform problem-solving steps are captured in real time, and authentication is performed by combining the identification code matching mechanism. If the authentication fails, it is marked as a state of not mastering the previous knowledge points, and a problem-solving knowledge space with status marking is generated for the learner.

[0011] For problem-solving questions, a path comparison device maps learners' paths to the standard answer, and a multi-dimensional analysis model is used to calculate feature state values ​​and statistically analyze the knowledge point mastery rate.

[0012] For multiple-choice and true / false questions, a question type correlation tool is used to match answer questions of the same type, and then the error type is deduced based on the characteristic status of the related answer questions.

[0013] Based on clustering of highly overlapping question types using information from the beginning and end of the problem-solving path, class data is projected onto a stage-based knowledge space. Multi-dimensional heatmaps are rendered to visualize the distribution of knowledge deficiencies according to the mastery rate of knowledge points. The Apriori algorithm is used to mine the weight relationship between the mastery of knowledge points and problem-solving features, and suggestions for optimizing the mastery path of knowledge points are generated.

[0014] Secondly, the present invention provides a teaching optimization system based on knowledge space, the system specifically comprising:

[0015] The first teaching optimization module is used to parse the knowledge point entities with identification codes generated from the official textbook based on the National Industrial Internet Identifier Resolution Secondary Node Platform, and to establish a relational network that includes sequential, parallel and comprehensive types.

[0016] The second teaching optimization module is used to collect problem-solving steps from multiple platform question banks based on relationship networks, and to authenticate them using an identifier code matching mechanism. After authentication, the steps are synchronized to the basic knowledge space and marked as extended knowledge point library. At the same time, it combines error statistics to calculate the difficulty threshold of knowledge point association combinations.

[0017] The third teaching optimization module is used to capture learners' cross-platform problem-solving steps in real time based on the relationship network, and to authenticate them by combining the identification code matching mechanism. If the authentication fails, it is marked as a state of not mastering the previous knowledge points, and a problem-solving knowledge space with status marking is generated for the learner.

[0018] The fourth teaching optimization module is used to map learners' paths to standard answers through a path comparison device for problem-solving questions, calculate feature state values ​​based on a multi-dimensional analysis model, and statistically analyze the knowledge point mastery rate.

[0019] The fifth teaching optimization module is used to match multiple-choice and true / false questions with answer questions of the same type through a question type association tool, and then infer the error type based on the characteristic status of the associated answer questions.

[0020] The sixth teaching optimization module is used to cluster highly overlapping question types based on the information at the beginning and end of the problem-solving path, project class data into the stage-based knowledge space, render multi-dimensional heat maps to visualize the distribution of knowledge deficiencies according to the mastery rate of knowledge points, and use the Apriori algorithm to mine the weight relationship between the mastery of knowledge points and problem-solving features to generate optimization suggestions for the mastery path of knowledge points.

[0021] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, and a computer program stored in the memory, wherein when the computer program is executed on the processor, it implements the knowledge space-based teaching optimization method as described in any of the above methods.

[0022] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the knowledge space-based teaching optimization method as described in any of the above methods.

[0023] Compared with the prior art, the present invention has at least one of the following technical effects:

[0024] 1. This invention, by constructing a knowledge space system, enables a comprehensive analysis of learners' mastery of knowledge points, identification of difficulties, and generation of personalized teaching optimization suggestions, thereby improving teaching quality and learning outcomes.

[0025] 2. This invention utilizes the National Industrial Internet Identifier Resolution Secondary Node Platform to accurately parse textbook content to generate knowledge point entities with identifier codes and construct multi-type relationship networks, ensuring the integrity and scalability of the knowledge system and providing a solid foundation for subsequent teaching optimization.

[0026] 3. This invention collects and authenticates the problem-solving steps from multiple platform question banks and synchronizes them to the basic knowledge space. Combined with the statistical calculation of incorrect questions, it calculates the threshold of knowledge point association and difficulty, effectively identifying teaching difficulties and providing data support for targeted teaching.

[0027] 4. This invention captures and authenticates learners' cross-platform problem-solving steps in real time, marks the status of unmastered prior knowledge points, generates a knowledge space with status markers, accurately locates learners' weak knowledge areas, and facilitates timely intervention and tutoring.

[0028] 5. This invention targets problem-solving questions, calculates feature state values ​​and statistically analyzes knowledge point mastery rates through path comparison and multi-dimensional analysis models, comprehensively assesses learners' mastery of knowledge points for problem-solving questions, and provides a basis for personalized learning.

[0029] 6. This invention targets multiple-choice and true / false questions, matching them with similar problem-solving questions and deducing the type of error to accurately pinpoint the reasons for learners' mistakes in multiple-choice questions, generating targeted learning remedial instructions, and improving learning efficiency.

[0030] 7. This invention clusters question types based on the information at the beginning and end of the problem-solving path, projects class data into a staged knowledge space, and generates a multi-dimensional heat map to visualize the distribution of knowledge deficiencies. It also explores the relationship between the mastery of knowledge points and the weight of problem-solving features, providing intuitive and scientific suggestions for optimizing class teaching.

[0031] 8. Unlike existing technologies that broadly judge knowledge mastery based on the accuracy of answering questions, this invention can analyze a learner's mastery of each knowledge point. Through detailed analysis of problem-solving steps, it can accurately pinpoint which knowledge points the learner has deficiencies in mastering and applying, such as text comprehension problems or unfamiliarity with knowledge application. This allows for more targeted learning recommendations and test question types to help learners better improve their knowledge and abilities.

[0032] 9. This invention establishes a basic knowledge space, a knowledge application space, and a learner's problem-solving knowledge space. It projects the phased knowledge space generated by the school based on teaching progress and examination syllabus requirements onto each student's problem-solving knowledge space. This integrates data from multiple platforms to form a more complete learner knowledge profile. Simultaneously, the analysis of detailed information such as problem-solving steps helps to gain a deeper understanding of learners' thought processes and knowledge application, improving the accuracy and effectiveness of teaching quality analysis. This allows school teachers to more clearly understand the overall knowledge mastery, various abilities, and knowledge gaps of students, providing strong support for teaching decisions.

[0033] 10. This invention, based on the precise identification of learners' deficiencies in knowledge mastery and application, can generate personalized learning plans and precise test question types for learners. By recommending learning content and test questions in a targeted manner, it can meet learners' individual learning needs, improve learning efficiency and effectiveness, and help students better master knowledge and enhance their abilities. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 This is a flowchart illustrating a knowledge space-based teaching optimization method according to an embodiment of the present invention.

[0036] Figure 2 This is a schematic diagram of the structure of a knowledge space-based teaching optimization system provided in an embodiment of the present invention;

[0037] Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0038] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0039] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0040] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0041] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0042] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0043] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0044] In this application embodiment, the entity executing the process includes a terminal device. This terminal device includes, but is not limited to, devices capable of executing the methods disclosed in this application, such as servers, computers, smartphones, and tablets. Figure 1 A flowchart illustrating a knowledge space-based teaching optimization method according to an embodiment of the present invention is shown below, in detail:

[0045] S101, based on the National Industrial Internet Identifier Resolution Secondary Node Platform, parses official textbooks to generate knowledge point entities with identifier codes and establishes a relational network that includes sequential, parallel, and comprehensive types.

[0046] In this embodiment, the construction of the real-time knowledge space includes the construction of a basic knowledge space, a knowledge application space, and a learner's problem-solving knowledge space. The construction of the corresponding knowledge spaces is uploaded to the National Industrial Internet Identifier Resolution Secondary Node Platform in real time for information retrieval and application by the cross-platform teaching platform.

[0047] For a particular learner's problem-solving knowledge space, this learner's knowledge space can be accessed on different learning platforms to analyze and push the knowledge resources and question banks of that platform.

[0048] With proper certification and authorization, schools or teachers can also access students' knowledge space information to conduct comprehensive analysis of the students.

[0049] The construction of the basic knowledge space is based on national textbooks and related official teaching courses, which are relatively authoritative basic knowledge spaces. A unique identifier is generated for each knowledge point under a knowledge domain of a certain discipline, and the relationships between knowledge points are obtained, including sequential relationships (including causal relationships, progressive relationships, and hierarchical relationships), parallel relationships (knowledge points exist in parallel, are independent of each other, and have no necessary connection with each other), and comprehensive relationships (knowledge points are interconnected and applied in a comprehensive manner).

[0050] Based on sequential relations, a "knowledge point-relation-knowledge point" triple is established, generating a unique identifier for each "entity-relation-entity" triple.

[0051] Based on comprehensive relationships, a unique identifier is generated by combining "multi-entity-multi-relationship-multi-entity" relationships.

[0052] Match the teaching materials and course content with the entity identifiers, sequential relation combination identifiers, and comprehensive relation combination identifiers corresponding to the knowledge points, and mark the knowledge points as the basic knowledge point base.

[0053] S102, based on a relational network, collects problem-solving steps from question banks across multiple platforms, and performs authentication using an identifier code matching mechanism. After authentication, it is synchronized to the basic knowledge space and marked as an extended knowledge point library. At the same time, it combines error statistics to calculate the difficulty threshold of knowledge point association combinations.

[0054] In this embodiment, the question bank of the school or various teaching platforms and the corresponding problem-solving steps are obtained. The knowledge points and knowledge point reasoning relationship combinations applied in each problem-solving step are broken down. The knowledge points and knowledge point reasoning relationship combinations applied in the problem-solving are matched with the entity identifier code and knowledge point relationship combination generated in step S101.

[0055] If the knowledge points or combinations of knowledge point relationships used in solving the problem do not match the identifier code, the knowledge points or combinations of knowledge point relationships are pushed to the expert review system at the second-level node for authentication. If the review is passed, the experts define the subject and classification, and generate a unique identifier code for each knowledge point or combination of knowledge point relationships. The problem-solving steps are then associated with the knowledge points and combinations of knowledge point relationships. Simultaneously, the corresponding knowledge points and combinations of knowledge point relationships are also synchronized to the public basic knowledge space database and marked as an extended first knowledge point database. If the review fails, the validity of the current problem-solving steps is marked as pending determination.

[0056] If the knowledge points or combinations of knowledge point relationships used in solving the problem can be matched with the identifier code, the problem-solving steps are associated with the knowledge points and combinations of knowledge point relationships.

[0057] As shown above, the solution knowledge points are linked together for each question. If a question has multiple solution steps, multiple solution knowledge points are linked together using the above method.

[0058] The process involves obtaining the steps learners took to answer each problem incorrectly, identifying the knowledge point association paths in their answers, comparing the knowledge points and associations at each step of the problem-solving process with those in the learners' answers, and counting the number of learner answers that did not match (Gn) and the number that matched (Hn). The accuracy rate for each step is calculated as Ln = Hn / (Gn + Hn). When Ln < a threshold, the corresponding knowledge point and association for that step are recorded as a difficult point. This process identifies the difficult points and associations for each problem.

[0059] S103, based on a relational network, captures learners' cross-platform problem-solving steps in real time and uses an identifier code matching mechanism for authentication. If authentication fails, it is marked as a state of not mastering the prior knowledge points, and a problem-solving knowledge space with status markers is generated for the learner.

[0060] In this embodiment, the steps of a learner's problem-solving questions in a certain school or teaching platform are obtained, and the knowledge points and knowledge point reasoning relationships applied in each step of the problem-solving process are broken down. The knowledge points and knowledge point reasoning relationships applied in the problem-solving process are matched with the entity identifier code corresponding to the knowledge point and the knowledge point relationship combination generated in step S101.

[0061] If the knowledge points or combinations of knowledge point relationships used in problem-solving cannot be matched with an identifier code, the knowledge points or combinations of knowledge point relationships are pushed to the expert review system at the second-level node for authentication. If the review is passed, the experts define the subject and classification, and generate unique identifier codes for the knowledge points or combinations of knowledge point relationships. Problem-solving steps are then associated with the knowledge points and combinations of knowledge point relationships. Simultaneously, the corresponding knowledge points and combinations of knowledge point relationships are also synchronized to the public basic knowledge space library and marked as extended second-level knowledge point libraries. If the review fails, the learner is marked as having taken an incorrect problem-solving step, and the learner is marked as not having mastered the prerequisite knowledge point corresponding to the current problem-solving step.

[0062] If the knowledge points or combinations of knowledge point relationships used in solving the problem can be matched with the identifier code, the problem-solving steps are associated with the knowledge points and combinations of knowledge point relationships.

[0063] As shown above, this ultimately forms a path connecting the knowledge points for solving all the questions the learner answers.

[0064] As a brief summary of the above steps S101 to S103, for example, the first step (1) sample data acquisition: acquire partial problem-solving and problem-solving steps data of each chapter as sample problem-solving data.

[0065] Step 2 (2) Sample data splitting and labeling: The problem-solving steps in the sample problem-solving data are split and labeled as the smallest problem-solving units. The knowledge relationships between each problem-solving unit in the problem-solving steps are defined as unknown (e.g., theorems, rules, etc.). The tools and methods applied by each problem-solving unit in the problem-solving steps are defined as unknown (e.g., completing the square method, Dijkstra's algorithm, etc.). The reasoning paths of each problem-solving unit in the problem-solving steps are defined as unknown (e.g., forward thinking, reverse thinking, divide and conquer, etc.).

[0066] Step 3 (3) Sample data labeling: The knowledge expert management system selects knowledge points in the basic knowledge space as the knowledge points corresponding to the smallest problem-solving units in the sample problem-solving steps in (2); selects methods such as completing the square and algorithms in the basic knowledge space as the tool methods for the smallest problem-solving units in the sample problem-solving steps in (2); selects knowledge points such as theorems and rules in the basic knowledge space as the reasoning connections between the smallest problem-solving units in the sample problem-solving steps in (2); selects thinking strategies (forward thinking, reverse thinking, divide and conquer, etc.) in the basic knowledge space as the reasoning paths between or between multiple smallest problem-solving units in the sample problem-solving steps in (2).

[0067] Step 4 (4) Establish a word segmentation library: For example, establish logical words used in the reasoning between each problem-solving unit in the problem-solving process, such as the problem-solving steps "because...therefore...", "or", "and", etc.

[0068] Step 5 (5) Model building: Using the labeled sample data as the training set, validation set and test set, build an extraction model for each dimension of the problem-solving steps, including the knowledge points corresponding to each problem-solving unit and the predicted score of the knowledge points corresponding to each problem-solving unit, the tools and methods corresponding to each problem-solving unit and the predicted score of the tools and methods corresponding to each problem-solving unit, the reasoning association between problem-solving units and the predicted score of the reasoning association, the reasoning path between multiple problem-solving units and the predicted score of the reasoning path.

[0069] Step 6 (6) Knowledge Application Space Library / Learner Problem-Solving Knowledge Space: The system automatically splits and marks each smallest problem-solving unit in the problem-solving step data based on the word segmentation library for the standard answers or learner problem-solving data corresponding to the questions on each platform. At the same time, it automatically marks the corresponding logical words and imports the system-marked problem-solving step data into the model to extract the corresponding problem-solving step information data, including the knowledge points corresponding to each problem-solving unit and the predicted score of the knowledge points corresponding to the problem-solving unit, the tool methods corresponding to each problem-solving unit and the predicted score of the tool methods corresponding to the problem-solving unit, the reasoning association between problem-solving units and the predicted score of the reasoning association, and the reasoning path between multiple problem-solving units and the predicted score of the reasoning path. For the predicted score that reaches the threshold, it is defined as valid information. For the predicted score that does not reach the threshold, it is defined as information to be reviewed. At the same time, one or more problem-solving units and logical words that do not mention knowledge points, tool methods, reasoning associations, or reasoning paths are also defined as information to be reviewed. After the information awaiting review is reviewed or corrected by the expert review system, for knowledge points, tools, reasoning connections, and reasoning paths that exceed the basic knowledge space, experts can define them according to the specific circumstances during the review process and synchronize them to the basic knowledge space to expand the relevant knowledge in the basic knowledge space. If this review information comes from the solution steps of the platform's standard answer, then this expanded knowledge point, tool, reasoning connection, and reasoning path are marked as extending the first knowledge point base; if this review information comes from the solution steps of the platform's learners, then this expanded knowledge point, tool, reasoning connection, and reasoning path are marked as extending the second knowledge point base. As above, the information awaiting review is transformed into knowledge points, tools, reasoning connections, and reasoning paths corresponding to the approved solution units. In this way, the standard answers of the question bank and the solution steps of learners are mapped one by one to the basic knowledge space. The solution step data is standardized, and standard answers and solution path data for each question are generated.

[0070] S104, for problem-solving questions, uses a path comparison device to map learners' paths to standard answers, calculates feature state values ​​based on a multi-dimensional analysis model, and statistically analyzes the knowledge point mastery rate.

[0071] In this embodiment, for problem-solving questions, if a learner is marked as not having mastered the prerequisite knowledge point corresponding to the current problem-solving step, the learner is marked as not having mastered this prerequisite knowledge point. The learner's characteristic state for this knowledge point is defined as M0 (not mastered), and relevant knowledge point courses or reminders for practice are pushed to them.

[0072] For questions where unmarked learners have not mastered the prerequisite knowledge point corresponding to the current problem-solving step:

[0073] (1) Match the knowledge point association paths corresponding to the learner's problem-solving steps with the knowledge point association paths of the question's answer (if multiple knowledge point association paths exist, compare multiple paths), map the knowledge point association paths of the question's answer to the knowledge point association paths corresponding to the learner's problem-solving steps, and then perform analysis:

[0074] a) If a certain proportion (a pre-set threshold) of the knowledge points and knowledge associations used in the answer to the question can be found in the knowledge points and knowledge associations in the knowledge point association path corresponding to the learner's problem-solving steps in sequence, the found knowledge points or knowledge associations are marked as "mastered" by the learner for this question, and the learner's analysis result for this question in dimension a is marked as 1; otherwise, it is -1.

[0075] b. If the knowledge points used in the last step of the connection path between two knowledge points are the same, the learner is marked as "mastered" based on the knowledge points or knowledge connection combination used in the last step, and the learner's analysis result for this question in dimension b is marked as 2; otherwise, it is -2.

[0076] c. Determine whether the learner's final solution is accurate. If it is accurate, mark the learner's analysis result for this question in dimension c as 3; otherwise, mark it as -3.

[0077] d. Whether the knowledge points and knowledge associations used in the learner's problem-solving knowledge point association path include the extended second knowledge point library in the public basic knowledge space library. If the knowledge points and knowledge associations used in the extended second knowledge point library in the public basic knowledge space library are included, the learner is marked as "mastered" for this question. If so, the learner's analysis result for this question in dimension d is marked as 4; otherwise, it is -4.

[0078] (2) Analyze the learner's characteristics based on the judgment results in (1):

[0079] 1) If a=1, b=2, and c=3, then analyze the learner's understanding of the knowledge points and knowledge connections involved in the answer to the question, define the characteristic state as M1, and no further related knowledge point courses or questions will be pushed;

[0080] 2) If a = 1, b = 2, and c = -3, then the learner's understanding of the knowledge points and connections involved in the answer to the question is analyzed, but careless calculation is made. The characteristic state is defined as M2, and a large number of reminders on pure calculation of related knowledge points can be pushed to improve calculation ability.

[0081] 3) If a = -1, c = 3, and d = 4, then the learner has a good grasp of the knowledge points and knowledge connections involved in solving the problem, and has a strong ability to expand knowledge. The characteristic state is defined as M3. Other problem-solving questions of the same type are pushed to test whether the knowledge points are firmly grasped.

[0082] As described above, we obtain whether the learner has mastered the knowledge points for each question. We count the number of times each knowledge point has been marked as mastered by the learner (E1) across all the learner's answers (including the number of knowledge points in the extended second knowledge point library in the public basic knowledge space), and the number of times each knowledge point has been applied across all the learner's answers (E2). We then calculate the learner's mastery level for each knowledge point as E = E1 / E2.

[0083] S105, for multiple choice and true / false questions, match the same type of question-solving questions through the question type association tool, and then infer the error type based on the characteristic status of the associated question-solving questions.

[0084] In this embodiment, for multiple-choice questions, 1) the learner's answer is compared with the correct answer. For questions answered correctly, there may be two possibilities: lucky guess or actual correct answer. For questions answered incorrectly, there may be two possibilities: careless calculation error or actual incorrect answer. 2) The knowledge point association paths corresponding to the solution process of the multiple-choice question currently being answered by the learner are obtained from the school or various teaching platforms. The last knowledge point and knowledge association combination in the knowledge point association path of the open-ended questions are obtained from the open-ended questions that have the same last knowledge point and knowledge association combination as the last knowledge point and knowledge association combination in the knowledge point association path of the current multiple-choice question. The knowledge point association paths of these open-ended questions are mapped to the knowledge point association paths corresponding to the multiple-choice questions. If a certain proportion (a pre-set threshold) of the knowledge points and knowledge association combinations applied in the open-ended questions can be found in the knowledge points and knowledge association combinations in the knowledge point association path corresponding to the solution steps of the current multiple-choice question in sequence, then the current open-ended question and the current multiple-choice question are determined to be the same type of question.

[0085] The study analyzes learners' mastery of problem-solving questions of the same type as multiple choice and true / false questions, and identifies the proportion of a certain number (which can be predefined) of the most recently solved problems in this question type, focusing on the features with the highest proportions (M0, M1, M2, M3).

[0086] If the feature obtained is M0, but the learner answers the multiple-choice questions correctly, it is determined that the learner guessed correctly on this multiple-choice question by chance. The learner has not mastered the knowledge points for answering the question and needs to practice with similar question types to test their knowledge and further explore the learner's knowledge gaps.

[0087] If the feature obtained is M1 or M2, but the learner answers the multiple-choice question incorrectly, it is determined that the learner made a careless calculation error in this multiple-choice question. In this case, a large number of reminders on pure calculation of related knowledge points can be pushed to improve the learner's calculation ability.

[0088] If the feature obtained is M3, but the learner answers the multiple-choice questions incorrectly, it is determined that the learner's knowledge of this type of question is not solid. It is necessary to push relevant knowledge point courses and practice questions of the same type to test the learner's knowledge gaps and further explore the gaps in the learner's knowledge.

[0089] S106 clusters highly overlapping question types based on the information at the beginning and end of the problem-solving path, projects class data into a stage-based knowledge space, renders a multi-dimensional heatmap to visualize the distribution of knowledge deficiencies according to the mastery rate of knowledge points, and mines the weight relationship between the mastery of knowledge points and problem-solving features through the Apriori algorithm to generate optimization suggestions for the mastery path of knowledge points.

[0090] In this embodiment, the knowledge point association paths of all question answers are obtained. The last knowledge point and knowledge association combination of the answer application are called tail information, and the first knowledge point and knowledge association combination of the answer application are called head information. Based on the tail information and head information, the questions are grouped, and the knowledge point association paths of each question are deduced backward. Questions with high overlap in sequential knowledge points and knowledge association combinations (with a pre-set overlap threshold) are grouped into the same question type.

[0091] For a learner, if we analyze their problem-solving characteristics for the same set of questions and, excluding the characteristic M2, there are other characteristics such as M0, M1, and M3, then we can analyze that this learner has weak text comprehension or scenario application ability, which we define as M4 = "weak comprehension ability based on this question type", and recommend learning resources to improve comprehension ability.

[0092] For each question type, we analyze and study its associated core knowledge points: based on the knowledge point mastery data and question-solving characteristics data of learners at the same learning stage, such as the characteristic data of learners in the same grade of a school. We obtain the mastery data E of each knowledge point of each question for all questions in a certain set of questions for each learner, and at the same time, we statistically analyze the characteristic data M (such as M0, M1, M2, M3, M4) reflected by learners based on each question. We align the knowledge points of all questions (for example: Question 1 involves knowledge points a, b, c, Question 2 involves knowledge points b, c, d. Here, "involved" includes knowledge points that learners have applied to the extended second knowledge point library in the common basic knowledge space library and marked as mastered. After alignment, the knowledge points are a, b, c, d, and an alignment coefficient is matched to them. The alignment coefficient for Question 1 is (). Given (1,1,1,0), and the alignment coefficient of question 2 is (0,1,1,1), we remove data with the same alignment coefficient but different feature data reflected by learners based on each question type. We also remove duplicate data with the same alignment coefficient but the same feature data reflected by learners based on each question type. This forms a knowledge point association model with the identifier code corresponding to the knowledge point as the column. For each row of data, the learner's mastery data for each knowledge point E is multiplied by the alignment coefficient, which equals the feature data M reflected by the learner based on this question. We then construct a knowledge point association model and use the Apriori algorithm to mine the association rules of knowledge points. By using correlation analysis, we can obtain the influence weight of the mastery level of each knowledge point on the characteristic value reflected by this question type. For example, a higher or lower mastery level of knowledge point n has the greatest impact on the final response M of this question type. Furthermore, we can analyze the correlation between the mastery levels of knowledge points. For instance, a higher or lower mastery level of knowledge point m will lead to a higher or lower mastery level of knowledge points m1, m2, m3, etc. Similarly, a higher or lower mastery level of knowledge points x1, x2, x3, etc. will lead to a higher or lower mastery level of knowledge point x. This helps learners quickly master core knowledge points and shifts their response characteristics when answering a certain question type towards a better M1 or M3, ultimately making it easier to master the answering of each question type. Simultaneously, it helps learners understand the promoting relationships between knowledge points and quickly establish knowledge mastery paths.

[0093] School or class teachers can obtain the mastery score E of all students in the entire grade or class for each knowledge point. This score is then numerically categorized, for example, E0 = 0, E1 = (0, 0.2), E2 = (0.2, 0.4), E3 = (0.4, 0.6), E4 = (0.6, 0.8), and E5 = (0.8, 1). This statistically generates the overall mastery of students across all knowledge points or related knowledge combinations. Based on the current grade or class's teaching progress, a shared basic knowledge space database generated according to this progress is obtained. This database then incorporates the learners' problem-solving knowledge space and problem-solving characteristic data. By mapping this part of the public basic knowledge space, we can obtain the overall characteristics of each knowledge point and knowledge association combination. For example, for each knowledge point or knowledge association combination, how many students have characteristics E0, E1, E2, E3, E4, and E5? Different colors and corresponding shades of color are used to represent the student situation in different dimensions. Schools and teachers can have a clearer understanding of students' overall mastery and application of knowledge points, and can strengthen teaching and practice in a targeted manner. Similarly, by analyzing students' answers to a certain type of question, and based on the proportion of students answering each type of question and reflecting characteristics M0, M1, M2, M3, and M4, we can conduct targeted teaching or testing.

[0094] Based on steps S101 to S106 above, in some embodiments, a knowledge point application library is generated by acquiring the solution steps (including the standard answer and the solution steps approved by the learner) for each question type in the question bank, as well as the knowledge points and reasoning process information of the associated knowledge space. For each question type, extended knowledge points and links to related knowledge point paths are added to the knowledge point application library for that question type, thus generating a knowledge point application library for each question type. Based on the information in the knowledge point application library for each question type, the number of source knowledge points and reasoning relationships for each step of reasoning for each knowledge point is counted, and the total number of times each knowledge point is applied in the question type is counted as a high-frequency applied knowledge point. Based on this high-frequency applied knowledge point, the reasoning proportion of each preceding knowledge point is calculated backwards. Based on this high-frequency applied knowledge point, the reasoning application frequency of the preceding one or more knowledge points is counted, thus identifying the high-frequency knowledge points used for problem-solving.

[0095] In some embodiments, step S101 above, which involves parsing the knowledge point entities with identifier codes generated from official textbooks based on the National Industrial Internet Identifier Resolution Secondary Node Platform and establishing a relational network including sequential, parallel, and comprehensive types, specifically includes:

[0096] The semantic segmentation module of the National Industrial Internet Identifier Resolution Secondary Node Platform is used to parse the official teaching materials to generate independent knowledge units, and each knowledge unit is assigned a globally unique identifier code.

[0097] Based on the logical structure of the textbook, a causal chain analyzer is used to identify sequential relationships, while a semantic similarity matcher is used to extract parallel relationships. Then, a cross-chapter association mining tool is called to generate comprehensive relationships, forming a relationship network.

[0098] Create triplet identifiers containing entity-relation-entity for sequential and parallel relations, and generate multi-entity and multi-relation combination identifiers for composite relations;

[0099] When new teaching resources are added, the incremental parser is triggered to extract new knowledge points or relationships, which are then submitted to the expert review system of the secondary node platform. After authentication, an identification code is assigned and the relationship network is updated.

[0100] In this embodiment, official textbooks are an important carrier of knowledge dissemination in the field of education. However, the knowledge points in textbooks are scattered and lack systematic association identifiers, which is not conducive to teachers' teaching and students' self-directed learning. This embodiment aims to utilize the National Industrial Internet Identifier Resolution Secondary Node Platform to parse official textbooks, generate knowledge point entities with identifier codes, and establish a relational network including sequential, parallel, and comprehensive types, providing a foundation for teaching optimization.

[0101] Specifically, ensure the normal operation of the National Industrial Internet Identifier Resolution Secondary Node Platform. The platform should possess a semantic segmentation module capable of semantic understanding and analysis of text. Official textbooks should be input into the platform as electronic documents in various formats, such as PDF and Word. The platform should have the corresponding document parsing capabilities to convert the document content into processable text data. Utilize the platform's semantic segmentation module to perform semantic segmentation on the textbook text. This module uses natural language processing technology to identify semantic boundaries in the text, dividing the textbook content into independent knowledge units. For example, in a mathematics textbook, a theorem, formula, or concept can be considered an independent knowledge unit. Assign a globally unique identifier to each independent knowledge unit. The identifier can be in the form of numbers, letters, or combinations thereof, ensuring uniqueness within the platform's knowledge system. Identifier assignment can be achieved through the platform's built-in identifier generation algorithm, which generates unique identifiers based on the characteristics and generation order of the knowledge units.

[0102] A causal chain analyzer is used to identify sequential relationships in textbooks. By analyzing logical statements in the textbook, such as "because...therefore..." and "first...then...", the causal chain analyzer determines the sequence and causal relationships between knowledge points. For example, in explaining the proof of a mathematical theorem, there is a strict sequential order between the steps; the causal chain analyzer can identify the sequential relationships between these steps. For the identified sequential relationships, a triplet identifier is created containing entity-relationship-entity. Here, the entity is the identifier of the knowledge unit, and the relation is "sequence" or a similar descriptive term. The triplet identifier clearly represents the sequential relationship between knowledge points.

[0103] Parallel relationships in textbooks are extracted using a semantic similarity matcher. The semantic similarity matcher calculates the semantic similarity between different knowledge units; when the similarity exceeds a certain threshold, these knowledge units are considered to have a parallel relationship. For example, in a Chinese language textbook, multiple words or sentences describing the same thing may have a parallel relationship. For the extracted parallel relationships, a triplet identifier containing entity-relationship-entity is created, with the relationship description being "parallel".

[0104] The cross-chapter association mining tool is invoked to generate comprehensive relationships. This tool analyzes the connections between knowledge points across different chapters in the textbook, identifying relationships that span multiple chapters and integrate multiple knowledge points. For example, in a physics textbook, mechanics and electricity concepts may be interconnected in certain comprehensive application problems. For the generated comprehensive relationships, a multi-entity, multi-relationship combination identifier is created to accurately describe the complex relationships between multiple knowledge points.

[0105] The identified sequential, parallel, and comprehensive relationships are integrated to form a complete knowledge point relationship network. This network is stored in a graph structure, where knowledge units are nodes, relationships are edges, and the edge attributes represent the relationship type (sequential, parallel, or comprehensive). The knowledge point relationship network is stored in the database of the National Industrial Internet Identifier Resolution Secondary Node Platform. The database should possess efficient data storage and retrieval capabilities to facilitate subsequent access and updates to the relationship network.

[0106] When new teaching resources (such as new textbook chapters, teaching aids, etc.) are added, the incremental parser is triggered. The incremental parser automatically detects knowledge points and relationships in the new resources. It parses the new teaching resources, extracting new knowledge points or relationships. For new knowledge points, an identifier is assigned according to the method in step one; for new relationships, a corresponding identifier is created based on their type (sequential, parallel, or combined). The extracted new knowledge points or relationships are submitted to the expert review system of the secondary node platform. The expert review system consists of experts in relevant fields who review and certify the new content. After expert certification, the new knowledge points and relationships are assigned identifiers and updated in the knowledge point relationship network. The update process includes adding new nodes and edges to the graph structure to ensure the integrity and accuracy of the relationship network.

[0107] In this embodiment, official teaching materials can be effectively parsed to generate knowledge point entities with identifiable codes, and a relational network including sequential, parallel, and comprehensive types can be established. This relational network provides a clear knowledge structure framework for teaching, helping teachers better organize teaching content and enabling students to more systematically grasp the connections between knowledge points, thereby improving learning outcomes. Simultaneously, the incremental parsing and updating mechanism ensures that the relational network can promptly reflect the content of newly added teaching resources, maintaining its timeliness and accuracy.

[0108] In some embodiments, in step S102 above, the step of collecting problem-solving steps from multi-platform question banks based on a relationship network, performing authentication using an identifier code matching mechanism, synchronizing the results to the basic knowledge space after authentication and marking them as extended knowledge points, and simultaneously calculating the difficulty threshold of knowledge point association combinations based on error statistics, specifically includes:

[0109] Collect problem-solving steps from question banks across multiple platforms, call the step semantic parser to break them down into atomic-level operation units and extract knowledge point elements;

[0110] Based on the relationship network, knowledge point elements are input into the identifier matching engine for matching. If the matching fails, the expert review request process is triggered. After expert certification, an identifier code is assigned to the new knowledge point, which is synchronously updated to the basic knowledge space and marked as the extended knowledge point library.

[0111] Based on the authentication results, atomic operation units are reorganized to generate knowledge point association paths, and the learner's wrong question database is accessed to count the accuracy of each step. When the accuracy rate is lower than the preset accuracy rate threshold, the corresponding knowledge point combination is automatically marked as a difficult point.

[0112] In this embodiment, the multi-platform question bank contains rich teaching resources, but the problem-solving steps are scattered and lack unified knowledge point identification, making it difficult to directly use for teaching optimization. This embodiment aims to collect problem-solving steps from the multi-platform question bank based on the constructed knowledge point relationship network, authenticate them through an identification code matching mechanism, synchronize them to the basic knowledge space and mark them as extended knowledge point databases, and combine error statistics to calculate the difficulty threshold of knowledge point association combinations, providing accurate difficulty analysis and targeted teaching suggestions for teaching.

[0113] Specifically, identify the multi-platform question banks from which to collect solution steps. These platforms can be common online education platforms, learning apps, etc. Establish technical interfaces with each platform to obtain access to their question banks and data interfaces for solution steps. For example, this can be done by signing data sharing agreements with the platforms to obtain text or structured data of the solution steps. Use data collection tools or scripts to periodically collect solution steps from the question banks of each platform. The collection frequency can be set according to actual needs, such as daily or weekly. The collected data should include question information, a detailed description of the solution steps, solution time, and other relevant information. Store the collected data in a temporary database for subsequent processing.

[0114] Develop or select a suitable step semantic parser that has the ability to semantically understand the problem-solving steps described in natural language. For example, the parser should be able to identify semantic elements such as actions, objects, and conditions in the steps.

[0115] The collected problem-solving steps are input into a step semantic parser, which breaks down the problem-solving steps into atomic-level operation units according to semantic logic. For example, the problem-solving steps of a math problem, "first calculate the expression inside the parentheses, then perform the multiplication operation," can be broken down into two atomic-level operation units: "calculate the expression inside the parentheses" and "perform the multiplication operation."

[0116] While breaking down operations into atomic-level units, the parser extracts key knowledge elements from each unit. These elements can be specific concepts, formulas, theorems, etc. For example, in the unit "calculate the expression within parentheses," the extracted key knowledge elements might be "rules of arithmetic operations," etc.

[0117] A tag matching engine is built using the existing knowledge point relationship network. This engine can search for matching knowledge point tags in the relationship network based on the characteristics of knowledge point elements. The extracted knowledge point elements are input into the tag matching engine for matching. The engine searches for the corresponding knowledge point tag code in the relationship network based on the semantic information and contextual relationships of the knowledge point elements. For example, if the extracted knowledge point element is "Pythagorean theorem," the engine searches for the tag code corresponding to this theorem in the relationship network.

[0118] If a match fails, it indicates that the knowledge point element may be a new knowledge point or one not recorded in the existing relationship network. In this case, the expert review request process is triggered. The knowledge point element that failed to match, along with the relevant problem-solving steps, is sent to the expert review system. The expert review system consists of teachers and education experts in the relevant subject area, who will review and evaluate the submitted content.

[0119] After expert verification, if a knowledge point is confirmed as a new knowledge point, the experts will assign it a globally unique identifier. Simultaneously, the new knowledge point and its identifier will be updated to the basic knowledge space and marked as an extended knowledge point in the extended knowledge point repository. The extended knowledge point repository stores important knowledge points that are not found in textbooks but appear in actual problem-solving processes, providing a basis for expanding teaching content.

[0120] Based on the identifier matching results, atomic-level operation units are recombined according to the problem-solving order to generate knowledge point association paths. For example, the solution process of a physics problem involves multiple knowledge points. By matching identifiers, the knowledge point identifier codes corresponding to each operation unit are connected to form a complete knowledge point association path.

[0121] The generated knowledge point association paths are integrated with the learner's error database. The learner's error database records the error information that students made during the problem-solving process, including question information, solution steps, and reasons for the error.

[0122] For each step in the knowledge point association path, the learner's accuracy rate at that step is calculated. The accuracy rate can be obtained by calculating the ratio of the number of students who solved the problem correctly to the total number of students who solved the problem. When the accuracy rate of a knowledge point combination is lower than a preset accuracy threshold, that knowledge point combination is automatically marked as a difficult point. For example, if the preset accuracy threshold is 70%, and the accuracy rate of a knowledge point combination in solving multiple questions is lower than 70%, then that knowledge point combination is considered a teaching difficulty.

[0123] This embodiment effectively collects problem-solving steps from question banks across multiple platforms, breaks them down, extracts knowledge point elements, and matches them with identifiers. New knowledge points are incorporated into the basic knowledge space and marked as extended knowledge points. Simultaneously, by combining the learner's error database with statistics on the accuracy of knowledge point association combinations, teaching difficulties are accurately identified. This provides teachers with abundant teaching resources and analytical support, helping them to conduct targeted teaching on challenging points and improve teaching quality; it also helps students better understand their knowledge weaknesses and engage in targeted learning.

[0124] In some embodiments, in step S103 above, the step of capturing learners' cross-platform problem-solving steps in real time based on a relationship network, performing authentication using an identifier code matching mechanism, and marking learners as not having mastered the prerequisite knowledge points if authentication fails, and generating a learner problem-solving knowledge space with status markers, specifically includes:

[0125] It captures learners' cross-platform problem-solving steps in real time and calls a step standardization processor to convert them into a structured operation sequence;

[0126] Based on the relationship network, the structured operation sequence is input into the identifier matching engine for matching. If the matching fails, the expert system is triggered for rapid authentication. When the authentication fails, the prerequisite knowledge points are automatically located and marked as not mastered.

[0127] By integrating matching results and status markers, knowledge point association paths are constructed according to the problem-solving sequence and encapsulated into traceable knowledge space units with status markers.

[0128] In this embodiment, in a digital education environment, learners often practice problem-solving on multiple online learning platforms. However, this cross-platform problem-solving data is scattered and lacks unified management, making it difficult to directly reflect the learner's knowledge mastery. This embodiment aims to capture learners' cross-platform problem-solving steps in real time based on a constructed knowledge point relationship network, authenticate them through an identifier code matching mechanism, mark the state of unmastered prior knowledge points, and generate a learner's problem-solving knowledge space with status markers, providing strong support for personalized teaching and learning diagnosis.

[0129] Specifically, we have established data integration mechanisms with multiple mainstream online learning platforms. Through negotiations with these platforms, we have obtained access to their problem-solving step data interfaces. For example, we have partnered with online math practice platforms and physics simulation experiment platforms to ensure real-time access to learners' problem-solving records on these platforms.

[0130] Building upon data access, a dedicated data capture tool should be deployed. This tool can be customized software based on web crawling technology, capable of scraping learners' problem-solving steps from interfaces of various platforms according to preset rules and frequencies. The captured data should include key information such as learner ID, problem information, solution timestamp, and detailed descriptions of the problem-solving operations.

[0131] The captured problem-solving steps are transmitted in real time to an intermediate data processing server. The server employs an efficient data transmission protocol to ensure data integrity and timeliness. On the server, a dedicated data storage area is established for each learner, storing the captured problem-solving steps in chronological order for subsequent processing.

[0132] A standardized processor for problem-solving steps has been developed, which has the ability to uniformly process problem-solving steps across different platforms and formats. The processor has a set of preset rules and templates for identifying and transforming various elements in the problem-solving steps.

[0133] The real-time captured solution steps are input into a step standardization processor. The processor first cleanses the solution steps, removing irrelevant characters and formatting information. Then, based on preset rules and templates, the solution steps are broken down into a series of standardized operation units and arranged into a structured operation sequence according to the solution order. For example, for a geometry problem, the solution steps "first draw a line segment AB, then draw a circle with A as the center and AB as the radius" are converted into two structured operation units: "draw line segment AB" and "draw a circle with A as the center and AB as the radius".

[0134] A tag matching engine is configured using the constructed knowledge point relationship network. The engine stores the tag codes of all knowledge points and the relationships between them, enabling it to quickly find the corresponding knowledge point tag code based on the input operation unit. The standardized structured operation sequence is input into the tag matching engine for matching. The engine analyzes each operation unit sequentially, searching for a matching knowledge point tag code in the knowledge point relationship network based on the semantic information and contextual relationships of the operation unit.

[0135] If a match for an operation unit fails, it indicates that the knowledge points involved in the operation are not explicitly identified in the existing relationship network, or that the learner's operation involves special circumstances. In this case, the expert system's rapid authentication process is triggered. The failed-match operation unit, along with relevant question information and solution context, is sent to the expert system. The expert system, composed of teachers and educational experts in relevant subjects, can quickly review and evaluate the submitted content.

[0136] After expert system verification, if it is confirmed that the learner has not mastered the knowledge point corresponding to the operation unit, the system will automatically locate the prerequisite knowledge points that the knowledge point depends on. The location of prerequisite knowledge points can be achieved by analyzing the logical relationships and dependencies in the knowledge point relationship network. For example, when learning the concept of derivative in advanced mathematics, if the learner has not mastered the concept of limit, the system will mark the concept of limit as a prerequisite knowledge point that has not been mastered.

[0137] Based on the identifier matching results and status marker information, a knowledge point association path is constructed according to the problem-solving sequence. For each structured operation unit, if the match is successful, the corresponding knowledge point identifier code is recorded; if the match fails and the unit is marked as not mastering the preceding knowledge point, the identifier code and status information of that preceding knowledge point are also recorded. The knowledge point information corresponding to all operation units is connected according to the problem-solving sequence to form a complete knowledge point association path. For example, in solving a chemistry experiment problem, multiple knowledge points are involved. By constructing a knowledge point association path, the knowledge points applied by the learner in sequence during the problem-solving process and any potential knowledge gaps can be clearly seen.

[0138] The constructed knowledge point association paths are encapsulated into traceable knowledge space units with status tags. Each knowledge space unit not only contains the knowledge point association paths but also records detailed information such as the learner's problem-solving time, problem source, and solution steps, facilitating subsequent tracing and analysis. For example, teachers can view these knowledge space units to understand a learner's learning journey and identified difficulties with a particular knowledge point, providing a basis for personalized tutoring. Simultaneously, the encapsulated knowledge space units are stored in the learner's personal knowledge space database for easy access and updates.

[0139] In this embodiment, learners' problem-solving steps across platforms can be captured in real time and standardized and matched with identifiers. For operations that fail authentication, the system can accurately mark the lack of prior knowledge, construct knowledge point association paths with status tags, and encapsulate them into traceable knowledge space units. This provides teachers with a comprehensive and detailed analysis of learners' knowledge mastery, helping them implement personalized teaching strategies and improve teaching quality; it also helps learners better understand their learning status and target their knowledge gaps and learning improvements accordingly.

[0140] In some embodiments, step S104 above, specifically for answer questions, involves mapping the learner's path to the standard answer using a path comparator, calculating feature state values ​​based on a multi-dimensional analysis model, and statistically analyzing the knowledge point mastery rate. This includes:

[0141] For problem-solving questions, a path comparison device aligns the learner's solution path with the standard answer path by logical segments, and a sliding window algorithm is used to calculate the overlap ratio of knowledge point identification codes within the segments.

[0142] Determine if the overlap ratio meets the standard and mark it as dimension A; check the consistency of the path endpoint identifier code and mark it as dimension B; verify the correctness of the final answer and mark it as dimension C; scan the extended knowledge point base identifier code and mark it as dimension D.

[0143] Based on the A-dimensional state, B-dimensional state, C-dimensional state, and D-dimensional state, historical question data is aggregated to construct a knowledge point-application matrix;

[0144] By traversing the knowledge point-application matrix, the frequency of each knowledge point being marked as mastered and the total frequency of application are counted, and the knowledge point mastery rate is calculated.

[0145] In this embodiment, problem-solving questions are an important question type for assessing learners' ability to comprehensively apply knowledge during the teaching and learning process. However, traditional teaching assessment methods often struggle to comprehensively and accurately analyze learners' mastery of knowledge points in problem-solving questions. This embodiment aims to address problem-solving questions by using a path comparison device to map learners' problem-solving paths to the standard answer paths, employing a multi-dimensional analysis model to calculate feature state values ​​and statistically analyze knowledge point mastery rates, thereby providing teachers with more precise teaching feedback and helping learners better understand their knowledge weaknesses.

[0146] Specifically, online learning platforms or examination systems record learners' steps in solving problems in real time. When a learner begins to solve a problem, the system records each step in chronological order, such as the text entered, the diagram drawn, and the selected options. These steps are then organized logically to form the learner's problem-solving path. For example, in a mathematical geometry proof problem, the learner's solution path might include steps such as "drawing auxiliary lines," "applying theorems," and "performing reasoning."

[0147] Subject matter experts pre-define standard answer paths based on the question requirements and the relevant knowledge system. These standard answer paths detail the standard steps and knowledge points required to solve the problem. For example, using a mathematical geometry proof, the standard answer path clearly indicates the theorems, formulas, and reasoning processes needed for each step.

[0148] Develop or select a path matching tool that has the ability to logically segment and align two paths. Logical segments can be divided according to the semantics and function of the problem-solving steps. For example, in a math problem, the proof process can be divided into logical segments such as "introducing conditions," "performing reasoning," and "drawing a conclusion."

[0149] The learner's problem-solving path and the standard answer path are input into a path comparison device. The device divides the two paths into corresponding logical segments according to a pre-defined logical segmentation rule. Then, a sliding window algorithm is used to compare the knowledge point identifiers within each logical segment. The sliding window algorithm slides across the logical segments of the two paths, calculating the overlap ratio of the knowledge point identifiers within the window. For example, in a certain logical segment, if the standard answer path involves knowledge point identifiers A, B, and C, and the learner's problem-solving path involves knowledge point identifiers A, B, and D, then the overlap ratio of the knowledge point identifiers within that logical segment is 2 / 3.

[0150] Based on the characteristics of the subject and the difficulty of the questions, a standard for the overlap ratio of knowledge point identification codes is pre-set. For example, for simple problem-solving questions, the standard can be set to an overlap ratio of no less than 80%; for complex problem-solving questions, the standard can be set to no less than 60%. The calculated overlap ratio of knowledge point identification codes within each logical segment is compared with the standard. If the overlap ratio reaches or exceeds the standard, the A-dimensional status of that logical segment is marked as "compliant"; otherwise, it is marked as "non-compliant".

[0151] Extract the knowledge point identifier codes at the endpoints of both the learner's problem-solving path and the standard answer path. These identifier codes typically represent the core knowledge points ultimately applied in the problem. Compare the learner's problem-solving path endpoint identifier codes with the standard answer path endpoint identifier codes. If they match, mark the B-dimensional state as "matched"; otherwise, mark it as "inconsistent".

[0152] The learner's final answer is evaluated based on the question requirements and the standard answer. Evaluation can be conducted manually or using an automated system (e.g., for multiple-choice, fill-in-the-blank questions with explicit answers). If the learner's final answer is correct, the C-dimensional status is marked as "Correct"; otherwise, it is marked as "Incorrect".

[0153] An extended knowledge point database is established, storing identification codes for knowledge points that learners may encounter during problem-solving but are not explicitly reflected in the standard answer path. These knowledge points may supplement or expand upon the standard answer, helping to more comprehensively assess learners' knowledge mastery. The database scans the knowledge point identification codes in the learner's problem-solving path, checking for matching codes in the extended knowledge point database. If a match is found, the D-dimensional status is marked as "involving extended knowledge points"; otherwise, it is marked as "not involving extended knowledge points".

[0154] Collect learners' problem-solving path data and corresponding multi-dimensional status markers (A, B, C, and D dimensions) for different questions at different times. This data can be stored in a database for subsequent analysis and processing. Aggregate historical question data according to knowledge point identifiers. Integrate all problem-solving path data and multi-dimensional status markers involving the same knowledge point to form a dataset based on knowledge point dimensions.

[0155] A knowledge point-application matrix is ​​constructed, where rows represent different knowledge points and columns represent different application scenarios (i.e., different questions). Elements in the matrix record the multi-dimensional status markers of each knowledge point within a given question. Based on aggregated historical question data, the multi-dimensional status markers of each knowledge point across different questions are populated into the knowledge point-application matrix. For example, for the solution status of knowledge point X in question Y, the corresponding A, B, C, and D dimension status markers are filled into the appropriate positions in the matrix.

[0156] The knowledge point-application matrix is ​​traversed, and the frequency of each knowledge point being marked as "mastered" is counted. In this embodiment, the criteria for judging "mastery" can be set according to the actual situation. For example, when the status of dimensions A, B, and C are all "meeting the standard," "consistent," and "correct," the knowledge point is considered to have been mastered in that question. The number of times each knowledge point is marked as "mastered" in all questions is counted. At the same time, the total application frequency of each knowledge point in all questions is counted, that is, the total number of times the knowledge point appears in the knowledge point-application matrix.

[0157] For each knowledge point, divide the frequency of its marked "mastery" by the total frequency of application to obtain the mastery rate of that knowledge point. For example, if knowledge point X is applied in 10 questions, and 6 of them are marked "mastery," then the mastery rate of knowledge point X is 6 / 10 = 60%.

[0158] In this embodiment, for open-ended questions, the learner's problem-solving path can be comprehensively and accurately analyzed. Through multi-dimensional status labeling and the construction of a knowledge point-application matrix, the mastery rate of each knowledge point can be calculated. This provides teachers with detailed and intuitive teaching feedback, helping them understand learners' mastery of different knowledge points and thus enabling targeted instruction and curriculum design. Simultaneously, learners can also identify their weaknesses based on the knowledge point mastery rate, allowing for targeted learning and review, thereby improving learning effectiveness.

[0159] In some embodiments, step S105 above, which involves matching multiple-choice questions with similar question-type answering questions using a question type association tool and then inferring the error type based on the characteristic states of the associated answering questions, specifically includes:

[0160] The tail identifier of the solution path of the choice and judgment questions is extracted by the question type association. Using the tail identifier as the key, all answer questions with the same tail identifier are matched in the question type fingerprint database.

[0161] Obtain the learner's latest feature state record on matching and answering questions, calculate the proportion of each feature state, and determine the dominant feature state;

[0162] By comparing the results of the choice-based judgment questions with the dominant feature state, when the two contradict each other, the error type is deduced based on preset rules, and targeted learning remedial instructions are generated.

[0163] In this embodiment, multiple-choice and true / false questions are common question types in educational assessment systems. Multiple-choice and true / false questions are typically used to quickly assess learners' memory and understanding of basic knowledge, while open-ended questions focus more on assessing learners' ability to synthesize knowledge and their thought processes. However, relying solely on the answers to multiple-choice and true / false questions often fails to provide a deeper understanding of the specific reasons for learners' errors. This embodiment aims to address multiple-choice and true / false questions by matching them with open-ended questions of the same type using a question type correlation tool, and then inferring the error type based on the characteristic states of the associated open-ended questions. This provides learners with more precise remedial learning suggestions, thereby improving learning outcomes.

[0164] Specifically, a question type association tool is developed, capable of extracting the tail identifier code of the solution path and matching it in a question type fingerprint database. The question type fingerprint database is a pre-built data storage system that stores a large amount of solution path information for open-ended questions and corresponding question feature identifiers, including the tail identifier code of the solution path. The meaning of the tail identifier code of the solution path is clearly defined. The tail identifier code can represent the identifier of the last key knowledge point or solution step involved in the solution path. For example, in a multiple-choice question about a historical event, the tail identifier code might correspond to the core knowledge points of that historical event, such as the time, place, and main figures of the event.

[0165] After a learner completes a multiple-choice or true / false question, the question type association tool extracts the tail identifier of the solution path. The solution path can be generated by recording the learner's operational steps and thought process during the answering process, such as recording the learner's chosen options and analytical approach to the question. Using the extracted tail identifier as the key, a search and matching process is performed in the question type fingerprint database. All open-ended questions containing the same tail identifier are identified. These open-ended questions and the multiple-choice or true / false questions have a certain correlation in terms of the key knowledge points or solution steps involved. For example, if the tail identifier of a multiple-choice or true / false question corresponds to a certain mathematical theorem, then the matched open-ended questions are likely to be application problems revolving around that theorem.

[0166] After learners complete the matched problem-solving questions, the system records their answers, including the time taken, steps taken, and final answer. This information constitutes a record of the learner's characteristic status on the problem-solving questions. This record is then organized and stored in a database. The database can be categorized and stored according to fields such as learner ID and question ID.

[0167] Based on the answers provided, the characteristic states can be categorized. For example, characteristic states can be divided into categories such as "clear problem-solving approach", "incorrect problem-solving steps", and "lack of proficiency in applying knowledge".

[0168] For each learner's characteristic state records on the matching problem-solving questions, the frequency of each characteristic state is counted, and its proportion in the total number of records is calculated. For example, if a learner exhibits the characteristic state of "incorrect solution steps" in 6 out of 10 matching problem-solving questions, then the proportion of this characteristic state is 60%.

[0169] A threshold for the percentage of each feature state is pre-defined, for example, 50%. The percentages of each feature state are compared to the threshold. If the percentage of a certain feature state exceeds the threshold, then that feature state is identified as the dominant feature state. For example, if the percentage of "incorrect solution steps" is 60%, exceeding the 50% threshold, then "incorrect solution steps" is the learner's dominant feature state.

[0170] After learners complete the multiple-choice questions, the system records their answers, indicating whether the selected option is correct.

[0171] Based on the characteristics of multiple-choice questions and the knowledge points they test, establish comparison rules between the results of multiple-choice questions and the dominant characteristic state. For example, if a multiple-choice question tests the memorization of a certain knowledge point, and the dominant characteristic state is "unfamiliarity with knowledge application," then when a learner makes a mistake on this multiple-choice question, it may indicate inaccurate memorization or a lack of in-depth understanding.

[0172] Compare the learner's answer to the true / false question with their dominant characteristic state. Determine if there is a contradiction. For example, if the correct answer to a true / false question requires reasoning based on a certain knowledge point, but the learner chose the wrong answer for that question, and their dominant characteristic state is "unclear problem-solving approach," then a contradiction can be considered.

[0173] Based on common answering errors, pre-defined rules are established. For example, when the answer to a true / false question is incorrect and the dominant characteristic is "incorrect solution steps," the error type may be "a deviation in understanding and mastering the solution steps"; when the answer to a true / false question is incorrect and the dominant characteristic is "lack of proficiency in applying knowledge," the error type may be "insufficient ability to apply knowledge points."

[0174] Based on the comparison results and preset rules, the types of errors learners make in selecting true / false questions can be deduced.

[0175] Establish a remedial strategy library that stores learning remedial strategies for different types of errors. For example, for the error type of "misunderstanding and mastering the problem-solving steps is flawed," remedial strategies may include providing detailed explanations of the problem-solving steps and conducting similar problem-solving step exercises; for the error type of "insufficient ability to apply knowledge points," remedial strategies may include providing practical application examples of knowledge points and organizing group discussions on the application of knowledge points.

[0176] Based on the type of error deduced, a corresponding remedial strategy is selected from the remedial strategy library to generate targeted learning remedial instructions. For example, if the error type is "a deviation in understanding and mastering the problem-solving steps," the generated learning remedial instruction could be "watch the video explaining the problem-solving steps for the knowledge point corresponding to this multiple-choice question, and complete 5 similar problem-solving step practice questions."

[0177] Choose an appropriate delivery method to convey remedial instructions to learners. For example, instructions can be delivered to learners through the messaging system of an online learning platform, email, or other means.

[0178] After receiving the remedial instruction, learners study and practice according to the instructions to make up for their knowledge gaps and insufficient problem-solving skills in multiple-choice and true / false questions.

[0179] In this embodiment, for multiple-choice and true / false questions, a question type correlation algorithm can be used to match similar open-ended questions. Based on the learner's characteristic state on open-ended questions, the error type can be deduced, and targeted remedial learning instructions can be generated. This helps teachers and learners gain a deeper understanding of the reasons for learners' errors on multiple-choice and true / false questions, providing learners with personalized learning support and improving their learning outcomes and grades. Simultaneously, it also provides a more scientific and precise basis for teaching evaluation and improvement.

[0180] In some embodiments, in step S106 above, the process of clustering highly overlapping question types based on the information at the beginning and end of the problem-solving path, projecting class data into a stage-based knowledge space, rendering a multi-dimensional heatmap to visualize the distribution of knowledge deficiencies according to the knowledge point mastery rate, and mining the weight relationship between the degree of knowledge point mastery and problem-solving features using the Apriori algorithm to generate optimization suggestions for knowledge point mastery paths specifically includes:

[0181] The path feature extractor extracts the knowledge point identification codes at the beginning and end of the questions. The density clustering algorithm is used to classify questions with an overlap of more than a preset overlap threshold into the same question type group, and a unique clustering identification code is generated for each question type group.

[0182] Based on the unique clustering identifier, a set of knowledge points for each stage is extracted according to the teaching progress, and a class student-knowledge point mastery rate matrix is ​​constructed. The matrix data of the class student-knowledge point mastery rate matrix is ​​then mapped to a three-dimensional knowledge space coordinate system.

[0183] Based on a three-dimensional knowledge space coordinate system, a mastery rate grading color mapping rule is set, and a transparency parameter is used to represent the density of people to generate an interactive heat map. The interactive heat map is used to display the distribution of knowledge gaps.

[0184] Based on interactive heatmaps, student problem-solving characteristics and knowledge point mastery rates are converted into transaction itemsets. By improving the Apriori algorithm, strong association rules are mined, and suggestions for optimizing knowledge point mastery paths are output.

[0185] In this embodiment, accurately grasping students' mastery of different knowledge points in a class and optimizing the knowledge point mastery path accordingly is crucial for improving teaching quality and student learning outcomes. Traditional knowledge point mastery assessment methods often lack in-depth analysis of question types and problem-solving characteristics, making it difficult to intuitively display the distribution of knowledge gaps and uncover the intrinsic relationship between knowledge point mastery and problem-solving characteristics. This embodiment aims to provide teachers and students with more accurate suggestions for optimizing knowledge point mastery paths by clustering highly overlapping question types based on the beginning and end information of the problem-solving path, combined with class data visualization and data mining techniques.

[0186] A path feature extractor was developed, capable of accurately extracting the identifiers of the beginning and end knowledge points from the solution path of a problem. The solution path can be generated by recording information such as the student's operational steps and thought process during the problem-solving process. For example, in a math word problem, the solution path might include steps such as analyzing the problem's conditions, constructing a solution strategy, and calculating the final answer. The identifiers of the beginning and end knowledge points correspond to the key knowledge points involved at the beginning and end of the solution, respectively.

[0187] Establish coding rules for knowledge point identifiers to ensure that each knowledge point has a unique and identifiable identifier. For example, identifiers can be generated by combining methods such as subject classification, chapter number, and knowledge point sequence number.

[0188] Density clustering algorithms (such as DBSCAN) are used to cluster the problems. Density clustering algorithms can discover clusters of arbitrary shapes and are robust to noisy data, making them suitable for handling complex distributions that may exist in the problem data.

[0189] Based on teaching needs and question characteristics, a pre-set overlap threshold is established. The overlap can be measured by comparing the similarity of the knowledge point identifiers at the beginning and end of the questions. For example, if two questions have more than a certain percentage (e.g., 70%) of overlapping knowledge point identifiers at the beginning and end, they are considered to have a high degree of overlap.

[0190] All questions are input into a path feature extractor to extract the knowledge point identifiers at the beginning and end. Then, a density clustering algorithm is used to cluster the questions, and questions with an overlap exceeding a preset threshold are grouped into the same question type group.

[0191] Generate a unique clustering identifier for each question type group. This identifier may include information such as the cluster number generated by the clustering algorithm and the clustering timestamp to ensure its uniqueness and traceability. Associate the generated clustering identifier with the corresponding question type group information and store them.

[0192] Based on the teaching plan and actual teaching progress, determine the set of knowledge points involved in the current stage. For example, in the course of a semester, divide the teaching into different stages according to the teaching weeks or chapters, and clarify the knowledge points that need to be mastered in each stage. From all the knowledge points, select the knowledge points that need to be focused on in the current stage to form a stage-specific knowledge point set.

[0193] Collect students' answer data for each knowledge point, including accuracy rate, answer time, and solution steps. Based on this data, calculate each student's mastery rate for each knowledge point. For example, mastery rate can be measured by the average score of students on related questions for that knowledge point. Construct a student-knowledge point mastery rate matrix, with students as rows and the set of knowledge points for each stage as columns. Each element in the matrix represents the corresponding student's mastery rate for the corresponding knowledge point.

[0194] Construct a three-dimensional knowledge space coordinate system, where each of the three dimensions represents different information. For example, one dimension can be set as the difficulty level of the knowledge point, another as the chapter to which the knowledge point belongs, and the third as the knowledge point mastery rate.

[0195] The data in the class student-knowledge point mastery rate matrix is ​​mapped onto a three-dimensional knowledge space coordinate system. The knowledge point mastery rate data corresponding to each student forms a data point in the coordinate system. Through the position and attribute information of the data point, the mastery of different knowledge points by students in the class can be intuitively displayed.

[0196] Based on teaching needs and actual circumstances, determine the grading standards for knowledge point mastery rates. For example, mastery rates can be divided into four levels: excellent (80%-100%), good (60%-79%), pass (40%-59%), and fail (0%-39%).

[0197] A different color is assigned to each mastery level, forming a color mapping rule. For example, excellent level corresponds to green, good level to blue, pass level to yellow, and fail level to red. Through color changes, the distribution of mastery rates for different knowledge points can be visually displayed.

[0198] In a three-dimensional knowledge space coordinate system, the number of data points in each region is counted, corresponding to the number of students, and the student density is calculated. The transparency parameter of the data points is then set based on the student density. Regions with higher student density have lower transparency and more vibrant colors; conversely, regions with lower student density have higher transparency and paler colors. By varying the transparency, the distribution and concentration of students across different knowledge points in the class can be visually displayed.

[0199] Choose a suitable visualization tool (such as D3.js, Tableau, etc.) to generate an interactive heatmap. Interactive heatmaps allow users to zoom, rotate, and pan using the mouse for more detailed data observation. Display the data points with pre-defined color mapping rules and transparency parameters in the visualization tool to generate the interactive heatmap. The heatmap can visually demonstrate students' mastery of different knowledge points and the distribution of knowledge gaps.

[0200] Extract problem-solving characteristics from students' answer data, such as solution time, number of steps, and whether auxiliary tools were used. Convert these characteristics and knowledge point mastery rates into transaction itemsets. Each student's answer record can be viewed as a transaction, containing the student's problem-solving characteristics for that question and the corresponding knowledge point mastery rate information.

[0201] Improvements can be made to the traditional Apriori algorithm to enhance mining efficiency and accuracy. For example, hash-based techniques can be used to reduce the number of candidate sets generated, or parallel computing methods can be employed to accelerate the algorithm's operation.

[0202] An improved Apriori algorithm is used to mine transaction itemsets and identify strong association rules. Strong association rules indicate a strong correlation between the degree of knowledge mastery and problem-solving characteristics. For example, it might be found that students who solve problems quickly and clearly demonstrate their solution steps have a higher mastery rate for a particular knowledge point.

[0203] Based on the strong association rules discovered, and combined with teaching objectives and students' actual situations, suggestions for optimizing the mastery path of knowledge points are generated. For example, if it is found that students have a low mastery rate on a certain knowledge point, and this is related to unclear problem-solving steps, then optimization suggestions may include strengthening the explanation and practice of the problem-solving steps for that knowledge point.

[0204] The generated suggestions for optimizing the learning path of knowledge points will be pushed to teachers and students so that teachers can adjust their teaching strategies and students can improve their learning methods.

[0205] In this embodiment, highly overlapping question types can be clustered based on the information at the beginning and end of the problem-solving path, projecting class data into a stage-based knowledge space and visually displaying the distribution of knowledge gaps through an interactive heatmap. Simultaneously, an improved Apriori algorithm is used to mine the weighted relationship between knowledge point mastery and problem-solving features, generating targeted suggestions for optimizing knowledge point mastery paths. This helps teachers more accurately understand students' knowledge mastery, adjust teaching strategies, and improve teaching quality; it also helps students identify their knowledge weaknesses, optimize learning methods, and improve learning outcomes.

[0206] Reference Figure 2 An embodiment of the present invention provides a teaching optimization system 2 based on knowledge space, wherein system 2 specifically includes:

[0207] The first teaching optimization module 201 is used to parse the knowledge point entities with identification codes generated from the official textbook based on the national industrial internet identifier resolution secondary node platform, and to establish a relational network that includes sequential, parallel and comprehensive types.

[0208] The second teaching optimization module 202 is used to collect problem-solving steps from multiple platform question banks based on relationship networks, and to authenticate them using an identifier code matching mechanism. After authentication, the steps are synchronized to the basic knowledge space and marked as extended knowledge point database. At the same time, the module calculates the difficulty threshold of knowledge point association combinations based on error statistics.

[0209] The third teaching optimization module 203 is used to capture learners' cross-platform problem-solving steps in real time based on the relationship network, and to authenticate them by combining the identification code matching mechanism. If the authentication fails, it is marked as a state of not mastering the previous knowledge points, and a problem-solving knowledge space with status marking is generated for the learner.

[0210] The fourth teaching optimization module 204 is used to map learners' paths to standard answers through a path comparison device for problem-solving questions, calculate feature state values ​​based on a multi-dimensional analysis model, and statistically analyze the knowledge point mastery rate.

[0211] The fifth teaching optimization module 205 is used to match multiple choice and true / false questions with answer questions of the same type through a question type association tool, and then infer the error type based on the characteristic status of the associated answer questions.

[0212] The sixth teaching optimization module 206 is used to cluster highly overlapping question types based on the information at the beginning and end of the problem-solving path, project class data into the stage-based knowledge space, render multi-dimensional heat maps to visualize the distribution of knowledge deficiencies according to the mastery rate of knowledge points, and use the Apriori algorithm to mine the weight relationship between the mastery of knowledge points and problem-solving features to generate optimization suggestions for the mastery path of knowledge points.

[0213] It is understandable that, such as Figure 1The content of the knowledge space-based teaching optimization method embodiments shown are all applicable to the knowledge space-based teaching optimization system embodiments. The specific functions implemented by the knowledge space-based teaching optimization system embodiments are the same as those shown below. Figure 1 The illustrated implementation of the knowledge space-based instructional optimization method is the same, and the beneficial effects achieved are the same as those shown. Figure 1 The beneficial effects achieved by the knowledge space-based teaching optimization method shown in the embodiment are also the same.

[0214] It should be noted that the information interaction and execution process between the above systems are based on the same concept as the method embodiments of the present invention. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0215] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0216] Reference Figure 3 The present invention also provides a computer device 3, including: a memory 302 and a processor 301, and a computer program 303 stored on the memory 302. When the computer program 303 is executed on the processor 301, it implements the knowledge space-based teaching optimization method as described in any of the above methods.

[0217] The computer device 3 may be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device 3 may include, but is not limited to, a processor 301 and a memory 302. Those skilled in the art will understand that... Figure 3 The computer device 3 is merely an example and does not constitute a limitation on the computer device 3. It may include more or fewer components than shown in the figure, or combine certain components, or different components, such as input / output devices, network access devices, etc.

[0218] The processor 301 may be a Central Processing Unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0219] In some embodiments, the memory 302 may be an internal storage unit of the computer device 3, such as a hard disk or memory of the computer device 3. In other embodiments, the memory 302 may be an external storage device of the computer device 3, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 3. Furthermore, the memory 302 may include both internal and external storage units of the computer device 3. The memory 302 is used to store the operating system, applications, boot loader, data, and other programs, such as the program code of the computer program. The memory 302 can also be used to temporarily store data that has been output or will be output.

[0220] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the knowledge space-based teaching optimization method as described in any of the above methods.

[0221] In this embodiment, if the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a photographing device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0222] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0223] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0224] In the embodiments disclosed in this application, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling or direct coupling or communication connection may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0225] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

Claims

1. A teaching optimization method based on knowledge space, characterized in that, The method specifically includes: Based on the national industrial internet identifier resolution secondary node platform, the system parses official textbooks to generate knowledge point entities with identifier codes and establishes a relational network that includes sequential, parallel, and comprehensive types. Based on the relationship network, the problem-solving steps of question banks from multiple platforms are collected and authenticated by the identification code matching mechanism. After authentication, the steps are synchronized to the basic knowledge space and marked as the extended knowledge point library. At the same time, the difficulty threshold of knowledge point association combination is calculated by combining the error statistics. Based on the relationship network, learners' cross-platform problem-solving steps are captured in real time, and authentication is performed by combining the identification code matching mechanism. If the authentication fails, it is marked as a state of not mastering the previous knowledge points, and a problem-solving knowledge space with status marking is generated for the learner. For problem-solving questions, a path comparison device maps learners' paths to the standard answer, and a multi-dimensional analysis model is used to calculate feature state values ​​and statistically analyze the knowledge point mastery rate. For multiple-choice and true / false questions, a question type correlation tool is used to match answer questions of the same type, and then the error type is deduced based on the characteristic status of the related answer questions. Based on the information at the beginning and end of the problem-solving path, highly overlapping question types are clustered. Class data is projected into a stage-based knowledge space. Multi-dimensional heatmaps are rendered according to the knowledge point mastery rate to visualize the distribution of knowledge deficiencies. The Apriori algorithm is used to mine the weight relationship between the degree of knowledge point mastery and problem-solving features, and suggestions for optimizing the knowledge point mastery path are generated. Specifically, for multiple-choice questions, a question type correlation algorithm is used to match answer questions of the same type, and then the error type is deduced based on the characteristic states of the related answer questions. This includes: The tail identifier of the solution path of the choice and judgment questions is extracted by the question type association. Using the tail identifier as the key, all answer questions with the same tail identifier are matched in the question type fingerprint database. Obtain the learner's latest feature state record on matching and answering questions, calculate the proportion of each feature state, and determine the dominant feature state; By comparing the results of the choice-based judgment questions with the dominant feature state, when the two contradict each other, the error type is deduced based on preset rules, and targeted learning remedial instructions are generated.

2. The method according to claim 1, characterized in that, The aforementioned platform, based on the National Industrial Internet Identifier Resolution Secondary Node, parses official textbooks to generate knowledge point entities with identifier codes, and establishes a relational network including sequential, parallel, and comprehensive types, specifically including: The semantic segmentation module of the National Industrial Internet Identifier Resolution Secondary Node Platform is used to parse the official teaching materials to generate independent knowledge units, and each knowledge unit is assigned a globally unique identifier code. Based on the logical structure of the textbook, a causal chain analyzer is used to identify sequential relationships, while a semantic similarity matcher is used to extract parallel relationships. Then, a cross-chapter association mining tool is called to generate comprehensive relationships, forming a relationship network. Create triplet identifiers containing entity-relation-entity for sequential and parallel relations, and generate multi-entity and multi-relation combination identifiers for composite relations; When new teaching resources are added, the incremental parser is triggered to extract new knowledge points or relationships, which are then submitted to the expert review system of the secondary node platform. After authentication, an identification code is assigned and the relationship network is updated.

3. The method according to claim 1, characterized in that, The process involves collecting problem-solving steps from multiple platform question banks based on a relationship network, authenticating them using an identifier code matching mechanism, and then synchronizing them to the basic knowledge space and marking them as extended knowledge points. Simultaneously, it calculates the difficulty threshold for knowledge point association combinations based on error statistics, specifically including: Collect problem-solving steps from question banks across multiple platforms, call the step semantic parser to break them down into atomic-level operation units and extract knowledge point elements; Based on the relationship network, knowledge point elements are input into the identifier matching engine for matching. If the matching fails, the expert review request process is triggered. After expert certification, an identifier code is assigned to the new knowledge point, which is synchronously updated to the basic knowledge space and marked as the extended knowledge point library. Based on the authentication results, atomic operation units are reorganized to generate knowledge point association paths, and the learner's wrong question database is accessed to count the accuracy of each step. When the accuracy rate is lower than the preset accuracy rate threshold, the corresponding knowledge point combination is automatically marked as a difficult point.

4. The method according to claim 1, characterized in that, The method, based on a relational network, captures learners' cross-platform problem-solving steps in real time and uses an identifier code matching mechanism for authentication. If authentication fails, the learner is marked as not having mastered the prior knowledge points, generating a learner's problem-solving knowledge space with status markers, specifically including: It captures learners' cross-platform problem-solving steps in real time and calls a step standardization processor to convert them into a structured operation sequence; Based on the relationship network, the structured operation sequence is input into the identifier matching engine for matching. If the matching fails, the expert system is triggered for rapid authentication. When the authentication fails, the prerequisite knowledge points are automatically located and marked as not mastered. By integrating matching results and status markers, knowledge point association paths are constructed according to the problem-solving sequence and encapsulated into traceable knowledge space units with status markers.

5. The method according to claim 1, characterized in that, For the problem-solving section, a path comparison device maps the learner's path to the standard answer, calculates feature state values ​​based on a multi-dimensional analysis model, and statistically analyzes the knowledge point mastery rate. Specifically, this includes: For problem-solving questions, a path comparison device aligns the learner's solution path with the standard answer path by logical segments, and a sliding window algorithm is used to calculate the overlap ratio of knowledge point identification codes within the segments. Determine if the overlap ratio meets the standard and mark it as dimension A; check the consistency of the path endpoint identifier code and mark it as dimension B; verify the correctness of the final answer and mark it as dimension C; scan the extended knowledge point base identifier code and mark it as dimension D. Based on the A-dimensional state, B-dimensional state, C-dimensional state, and D-dimensional state, historical question data is aggregated to construct a knowledge point-application matrix; By traversing the knowledge point-application matrix, the frequency of each knowledge point being marked as mastered and the total frequency of application are counted, and the knowledge point mastery rate is calculated.

6. The method according to any one of claims 1 to 5, characterized in that, The method involves clustering highly overlapping question types based on the beginning and end information of the problem-solving path, projecting class data into a stage-based knowledge space, rendering multi-dimensional heatmaps to visualize the distribution of knowledge deficiencies according to the mastery rate of knowledge points, and using the Apriori algorithm to mine the weight relationship between the degree of mastery of knowledge points and problem-solving features to generate optimization suggestions for the mastery path of knowledge points, specifically including: The path feature extractor extracts the knowledge point identification codes at the beginning and end of the questions. The density clustering algorithm is used to classify questions with an overlap of more than a preset overlap threshold into the same question type group, and a unique clustering identification code is generated for each question type group. Based on the unique clustering identifier, a set of knowledge points for each stage is extracted according to the teaching progress, and a class student-knowledge point mastery rate matrix is ​​constructed. The matrix data of the class student-knowledge point mastery rate matrix is ​​then mapped to a three-dimensional knowledge space coordinate system. Based on a three-dimensional knowledge space coordinate system, a mastery rate grading color mapping rule is set, and a transparency parameter is used to represent the density of people to generate an interactive heat map. The interactive heat map is used to display the distribution of knowledge gaps. Based on interactive heatmaps, student problem-solving characteristics and knowledge point mastery rates are converted into transaction itemsets. By improving the Apriori algorithm, strong association rules are mined, and suggestions for optimizing knowledge point mastery paths are output.

7. A teaching optimization system based on knowledge space, characterized in that, The system specifically includes: The first teaching optimization module is used to parse the knowledge point entities with identification codes generated from the official textbook based on the National Industrial Internet Identifier Resolution Secondary Node Platform, and to establish a relational network that includes sequential, parallel and comprehensive types. The second teaching optimization module is used to collect problem-solving steps from multiple platform question banks based on relationship networks, and to authenticate them using an identifier code matching mechanism. After authentication, the steps are synchronized to the basic knowledge space and marked as extended knowledge point library. At the same time, it combines error statistics to calculate the difficulty threshold of knowledge point association combinations. The third teaching optimization module is used to capture learners' cross-platform problem-solving steps in real time based on the relationship network, and to authenticate them by combining the identification code matching mechanism. If the authentication fails, it is marked as a state of not mastering the previous knowledge points, and a problem-solving knowledge space with status marking is generated for the learner. The fourth teaching optimization module is used to map learners' paths to standard answers through a path comparison device for problem-solving questions, calculate feature state values ​​based on a multi-dimensional analysis model, and statistically analyze the knowledge point mastery rate. The fifth teaching optimization module is used to match multiple-choice and true / false questions with answer questions of the same type through a question type association tool, and then infer the error type based on the characteristic status of the associated answer questions. The sixth teaching optimization module is used to cluster highly overlapping question types based on the information at the beginning and end of the problem-solving path, project class data into the stage-based knowledge space, render multi-dimensional heat maps to visualize the distribution of knowledge deficiencies according to the mastery rate of knowledge points, and use the Apriori algorithm to mine the weight relationship between the mastery of knowledge points and problem-solving features to generate optimization suggestions for the mastery path of knowledge points. Specifically, for multiple-choice questions, a question type correlation algorithm is used to match answer questions of the same type, and then the error type is deduced based on the characteristic states of the related answer questions. This includes: The tail identifier of the solution path of the choice and judgment questions is extracted by the question type association. Using the tail identifier as the key, all answer questions with the same tail identifier are matched in the question type fingerprint database. Obtain the learner's latest feature state record on matching and answering questions, calculate the proportion of each feature state, and determine the dominant feature state; By comparing the results of the choice-based judgment questions with the dominant feature state, when the two contradict each other, the error type is deduced based on preset rules, and targeted learning remedial instructions are generated.

8. A computer device, characterized in that, include: A memory and a processor, and a computer program stored in the memory, which, when executed on the processor, implements the knowledge space-based teaching optimization method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, It stores a computer program, which, when executed by a processor, implements the knowledge space-based teaching optimization method as described in any one of claims 1 to 6.

Images