Method for constructing a knowledge progression map

Abstract

The application is suitable for the technical field of knowledge map, and provides a construction method of knowledge progression map, which comprises the following steps: obtaining a target knowledge point; determining relevant courses and the correlation between the relevant courses from a pre-constructed course knowledge graph according to the target knowledge point; constructing a knowledge base map according to the relevant courses and the correlation; and dividing the knowledge base map into learning stages according to at least one preset course attribute of the relevant courses to obtain a knowledge progression map containing multiple learning stages. The application can quickly and accurately construct a knowledge map suitable for course learning.

Description

Technical Field

[0001] This invention belongs to the field of knowledge map technology, and in particular relates to a method for constructing a knowledge progression map. Background Technology

[0002] Knowledge maps, as an effective knowledge management tool, play a vital role in course learning. They visually represent the overall knowledge structure of a course and the distribution of various knowledge points, providing learners with a visual knowledge framework. They clearly present the connections and structure between knowledge points, helping students clarify learning goals and paths, build knowledge systems, and improve learning efficiency.

[0003] In related technologies, constructing knowledge maps typically requires manually collecting a large amount of data, filtering out useful information, organizing and classifying it, and then manually designing graphics, connecting nodes, and drawing the knowledge map. This traditional construction method is usually suitable for fixed knowledge systems. For course learning, different learners, or the same learner at different stages, have different learning needs, thus requiring the construction of targeted knowledge maps. Traditional construction methods are cumbersome, involve a huge workload in drawing and modifying, are time-consuming, and are prone to errors. Furthermore, traditional knowledge maps cannot provide learners with a systematic learning navigation solution.

[0004] Therefore, there is an urgent need for a new way to construct knowledge maps to support course learning. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a method for constructing a knowledge progression map, so as to quickly and accurately construct a knowledge map suitable for course learning.

[0006] This invention provides a method for constructing a knowledge progression map, including:

[0007] Acquire the target knowledge points;

[0008] Based on the target knowledge points, relevant courses and the relationships between them are determined from a pre-constructed course knowledge graph.

[0009] Based on the relevant courses and the associated relationships, construct a knowledge base map;

[0010] Based on at least one preset course attribute of the relevant courses, the knowledge base map is divided into learning stages to obtain a knowledge progression map containing multiple learning stages.

[0011] As one possible implementation, the association includes: prior and successor;

[0012] Accordingly, constructing a knowledge base map based on the relevant courses and the associated relationships includes:

[0013] Based on the prerequisite and successor relationships among the related courses, arrows are used to connect the related courses that have an association, resulting in a tree-like learning path for the related courses;

[0014] Based on the tree-like learning path, a knowledge base map is constructed.

[0015] As one possible implementation, the preset course attributes include: learning difficulty;

[0016] Accordingly, the step of dividing the knowledge base map into stages based on at least one preset course attribute of the relevant courses includes:

[0017] Obtain at least one learning difficulty threshold and sort them from smallest to largest;

[0018] In the knowledge base map, the first relevant course with a learning difficulty greater than the first learning difficulty threshold is found on each branch from front to back and is taken as the first dividing point; all relevant courses before the first dividing point are divided into a learning stage.

[0019] Starting from each first dividing point, find the first relevant course on each branch with a learning difficulty greater than the next learning difficulty threshold, and use it as the second dividing point; divide all relevant courses before the second dividing point into a learning stage;

[0020] If there are no courses with a learning difficulty greater than the corresponding learning difficulty threshold on a branch, then all related courses on that branch and before it will be merged into the current learning stage.

[0021] Repeat the above steps until all learning difficulty thresholds are divided into learning stages.

[0022] As one possible implementation, the preset course attribute also includes: centrality; the centrality of the related course is: the number of related courses associated with the related course;

[0023] Accordingly, the step of dividing the knowledge base map into stages based on at least one preset course attribute of the relevant courses further includes:

[0024] After completing the learning stage division of all learning difficulty thresholds, determine whether the centrality of each relevant course is greater than the preset centrality threshold.

[0025] If the centrality of any related course is greater than the centrality threshold, then the related course and previous related courses in the current learning stage will be assigned to the previous learning stage.

[0026] As one possible implementation, the preset course attributes also include: time cost;

[0027] Accordingly, the step of dividing the knowledge base map into learning stages based on at least one preset course attribute of the relevant courses further includes:

[0028] Calculate the total time cost for each learning stage based on the time cost of each relevant course within each learning stage.

[0029] If the total time cost corresponding to any learning stage is greater than the preset time cost threshold, then the learning stage will be further divided into two learning stages.

[0030] As one possible implementation, the learning phase is further divided into two learning phases, including:

[0031] A target learning difficulty threshold is determined based on the learning difficulty threshold corresponding to the current learning stage and the previous learning stage; wherein, the target learning difficulty threshold is located between the learning difficulty thresholds corresponding to the current learning stage and the previous learning stage.

[0032] During this learning phase, the first relevant course with a learning difficulty greater than the target learning difficulty threshold is found on each branch from front to back, and is used as the target dividing point;

[0033] Based on the target dividing point, this learning phase is further divided into two learning phases.

[0034] As one possible implementation, obtaining at least one learning difficulty threshold includes:

[0035] Obtain learners' historical learning data;

[0036] Based on the historical learning data, evaluate the learner's learning ability time series;

[0037] Based on the learning ability time series, within a preset learning difficulty threshold range, the learning difficulty threshold is determined by a pre-trained neural network model.

[0038] As one possible implementation, after obtaining the knowledge progression map containing multiple learning stages, the method further includes:

[0039] Set advanced tests and advancement requirements for each learning stage;

[0040] After a learner completes any learning stage, the learner is tested using an advanced test for that stage, and the test results are obtained.

[0041] Based on the exam results, determine whether the advancement conditions are met;

[0042] If the advancement conditions are met, the next learning stage of the knowledge advancement map is opened; if the advancement conditions are not met, the next learning stage of the knowledge advancement map is not opened.

[0043] As one possible implementation, the method further includes:

[0044] Access to multimodal course resources;

[0045] Multimodal course resources are transformed into structured text data, and entity extraction is performed on the text data to obtain the corresponding knowledge points.

[0046] The knowledge points are analyzed to identify course associations, and a course knowledge graph is constructed.

[0047] As one possible implementation, obtaining multimodal course resources includes: obtaining multimodal course resources from a website through web crawling technology.

[0048] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows:

[0049] This invention pre-constructs a course knowledge graph. Based on the target knowledge points, it can identify relevant courses from the course knowledge graph and automatically construct a basic knowledge map based on the relationships between the relevant courses. This enables the rapid and accurate construction of a knowledge map suitable for course learning, helping learners clarify their learning goals and paths. Furthermore, based on at least one course attribute of the relevant courses, the basic knowledge map is divided into learning stages to obtain a knowledge progression map containing multiple learning stages. This provides learners with a systematic and phased learning navigation solution, improving learning efficiency. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 This is a schematic diagram illustrating the construction process of the course knowledge graph provided in an embodiment of the present invention;

[0052] Figure 2 This is a flowchart illustrating the method for constructing a knowledge progression map provided in an embodiment of the present invention;

[0053] Figure 3 This is a schematic diagram of the learning stage division provided in the embodiments of the present invention. Figure 1 ;

[0054] Figure 4 This is a schematic diagram of the learning stage division provided in the embodiments of the present invention. Figure 2 . Detailed Implementation

[0055] 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 the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0056] To illustrate the technical solution described in this invention, specific embodiments are described below.

[0057] A knowledge map is a knowledge management tool that visually presents concepts, knowledge points, and their relationships within a knowledge domain. Its general structure includes: nodes, representing knowledge elements such as concepts, knowledge points, and topics; and lines, indicating relationships between nodes, such as causal, progressive, and parallel relationships. In education, knowledge maps help students clarify learning goals and paths, build systematic knowledge frameworks, and improve learning efficiency and effectiveness.

[0058] A course knowledge graph is a structured form of knowledge representation that presents course-related information in the form of a graph. Nodes represent entities such as courses and knowledge points, while edges represent the various relationships between these entities, such as the knowledge points included in a course, prerequisite relationships between knowledge points, and sequential relationships between courses. By constructing a course knowledge graph, the connections between various elements in the course system can be clearly displayed, providing a foundation for subsequent operations.

[0059] See Figure 1 As shown, the process of constructing the course knowledge graph in this embodiment is as follows:

[0060] (1) Acquisition of course resource data.

[0061] Multimodal course resources, such as images, text, and videos, can be obtained from websites and other channels through web scraping technology, manual methods, and other means.

[0062] (2) Data preprocessing and storage.

[0063] For different types of data, this embodiment transforms unstructured multimodal teaching data into structured text data, and then extracts the corresponding knowledge points from the text data through entity extraction.

[0064] (3) Knowledge graph construction.

[0065] For preprocessed courses or knowledge points, association rule mining methods are used to mine course associations and construct a professional knowledge graph that integrates multiple course resources. Knowledge elements are extracted from text data through technologies such as entity recognition and relationship extraction, and then organized into a knowledge graph.

[0066] Knowledge graphs can be stored using graph databases. Graph databases prioritize efficient graph querying and searching, storing data using nodes and edges, and can also store data containing attribute values. Therefore, to facilitate queries related to target knowledge points, a graph database approach is used to store the knowledge graph for the data structure course. This approach better supports the storage and retrieval of relationships between knowledge points and their attribute information.

[0067] Furthermore, this embodiment finds relevant courses based on the course knowledge graph and generates a knowledge map according to the relationship between courses. It can quickly and accurately construct a knowledge map suitable for course learning, providing learners with an intuitive and clear learning navigation, helping them to better understand the course system and knowledge structure, rationally arrange the learning sequence, and significantly improve learning efficiency.

[0068] Figure 2 This is a schematic diagram illustrating the implementation process of the knowledge progression map construction method provided in this embodiment of the invention. See [link / reference]. Figure 2 As shown, the method includes:

[0069] Step S201: Obtain the target knowledge points.

[0070] The target knowledge point here can be the knowledge keywords that the learner wants to learn, such as "programming" or more specifically "sorting algorithm", etc. This embodiment does not limit it.

[0071] Step S202: Based on the target knowledge points, determine the relevant courses and the relationships between them from the pre-built course knowledge graph.

[0072] This embodiment pre-constructs a domain-specific knowledge graph. A domain-specific knowledge graph is built for a specific domain or industry, targeting a specific task or user group. Because it is domain-specific, a domain-specific knowledge graph can perform more refined filtering and verification of data, ensuring the accuracy and reliability of knowledge. For example, a knowledge graph specifically built for the programming domain covers knowledge of programming language syntax, semantics, data structures, algorithms, programming frameworks, development tools, and many other aspects. The integration, representation, and association of various knowledge points can help developers learn, develop, and solve problems more efficiently.

[0073] Based on a given target knowledge point, the course knowledge graph is used to find courses that are directly or indirectly related to it. For example, if the target knowledge point is "convolutional neural network," it might be directly related to the "deep learning" course in the knowledge graph, as this course mainly explains various deep learning models, including convolutional neural networks. Simultaneously, it might be indirectly related to the "machine learning" course, because machine learning is the foundation of deep learning, and some algorithms and theories in the "machine learning" course are prerequisites for understanding "convolutional neural networks." After finding relevant courses, the connection methods and relationship types of these courses in the knowledge graph are further analyzed to clarify their specific connections. For example, both "advanced mathematics" and "probability theory and mathematical statistics" are related to the "machine learning" course. The calculus and other knowledge in "advanced mathematics" are the foundation for some algorithm derivations in "machine learning," so they have a foundational support relationship. Meanwhile, "probability theory and mathematical statistics" provides the theoretical basis for data modeling and analysis in "machine learning," so they have a theoretical application relationship. By identifying these connections, the logical links between courses can be better understood, providing strong support for learning path planning.

[0074] Step S203: Construct a knowledge base map based on relevant courses and relationships.

[0075] In this embodiment, constructing a knowledge base map based on relevant courses and relationships is a process of visually presenting the knowledge covered by the courses and their relationships.

[0076] As one possible implementation, the relationships can include prerequisites and successors. The process of constructing a knowledge base map can be as follows: based on the prerequisite and successor relationships between related courses, connect related courses with arrows to obtain a tree-like learning path for the related courses; and construct a knowledge base map based on the tree-like learning path.

[0077] Prerequisite relationships between courses refer to the fact that prior knowledge and skills from other courses are required to learn a particular course; otherwise, it is difficult to understand and master the content of the first course. Successive relationships between courses refer to the fact that after learning a particular course, one can further learn knowledge and skills from other courses. These courses are usually in-depth extensions and expansions of the first course.

[0078] Based on the established prerequisites and successors, arrows connect related courses. For example, starting with "Advanced Mathematics," arrows point to courses that depend on it, such as "Probability Theory and Mathematical Statistics" and "Algorithm Analysis"; after completing "Data Structures," arrows point to its subsequent courses, such as "Database Principles," and so on. Through these arrow connections, the sequence and dependencies between courses are clearly presented, forming a tree-like learning path. This tree-like learning path visually demonstrates the logical order between courses, allowing learners to clearly see what to learn first and what to learn next.

[0079] A tree-like learning path primarily reflects the sequential order of courses, while a knowledge map offers a more comprehensive and richer overview. When constructing a knowledge base map, in addition to the sequential relationships between courses, other relevant information can be added, such as the main content of each course, key points and difficulties, reference materials, and learning suggestions. Furthermore, the layout of course nodes and the style of arrows in the map can be adjusted according to the degree and closeness of the connections between courses, making the knowledge base map more intuitive and easier to understand. Through a knowledge base map, learners can not only understand the learning sequence of courses but also access more comprehensive learning resources and guidance.

[0080] Step S204: Based on at least one preset course attribute of the relevant courses, divide the knowledge base map into learning stages to obtain a knowledge progression map containing multiple learning stages.

[0081] The difficulty, depth, and knowledge type of a course can all be used as preset course attributes. Based on these attributes, the knowledge foundation map can be divided into multiple learning stages. These stages are then clearly presented in the knowledge progression map. Within each stage, the courses included are displayed, and the relationships between courses are indicated by lines or arrows. This provides learners with a logical learning path, helping them understand which courses to take at each stage and how to transition from one stage to the next, thus achieving gradual knowledge accumulation and skill enhancement.

[0082] For example, the knowledge advancement map can set advancement tests and advancement conditions for each learning stage; after a learner completes any learning stage, the learner is tested through the advancement test for that learning stage, and the test results are obtained; based on the test results, it is determined whether the advancement conditions are met; if the advancement conditions are met, the next learning stage of the knowledge advancement map is opened; if the advancement conditions are not met, the next learning stage of the knowledge advancement map is not opened.

[0083] This invention pre-constructs a course knowledge graph. Based on the target knowledge points, it can identify relevant courses from the course knowledge graph and automatically construct a basic knowledge map based on the relationships between the relevant courses. This enables the rapid and accurate construction of a knowledge map suitable for course learning, helping learners clarify their learning goals and paths. Furthermore, based on at least one course attribute of the relevant courses, the basic knowledge map is divided into learning stages to obtain a knowledge progression map containing multiple learning stages. This provides learners with a systematic and phased learning navigation solution, improving learning efficiency.

[0084] In one embodiment of the present invention, an algorithm for dividing the knowledge base map into stages is also designed to more rationally divide the learning stages. The following is in conjunction with... Figure 3 and Figure 4 The process of dividing the learning phase in this embodiment will be described in detail.

[0085] Figure 3 This is a knowledge base map generated for the knowledge point "sorting algorithms". K0-K9 represent: sorting algorithms, merge sort, insertion sort, selection sort, exchange sort, direct insertion sort, shell sort, heap sort, selection sort, and bubble sort, respectively.

[0086] (1) Obtain at least one learning difficulty threshold and sort them from smallest to largest; in the knowledge base map, find the first relevant course with a learning difficulty greater than the first learning difficulty threshold on each branch from front to back, and take it as the first dividing point; divide all relevant courses before the first dividing point into a learning stage; starting from each first dividing point, find the first relevant course with a learning difficulty greater than the next learning difficulty threshold on each branch from front to back, and take it as the second dividing point; divide all relevant courses before the second dividing point into a learning stage; if there is no relevant course with a learning difficulty greater than the corresponding learning difficulty threshold on the branch, then merge the branch and all relevant courses before it into the current learning stage; repeat the above steps until the learning stage division of all learning difficulty thresholds is completed.

[0087] See Figure 4 As shown, assume the first learning difficulty threshold is N1. Search for courses with a learning difficulty greater than N1 along each branch. If K1, K3, and K4 are all greater than N1, and K2 is not greater than K1, continue searching along K2. If K5 and K6 are both greater than N1, then all related courses before K1, K3, K4, K5, and K6 (i.e., K0 and K2) are classified into a learning stage 1.

[0088] Learning Phase 2 builds upon Learning Phase 1. Assume the second learning difficulty threshold is N2. Starting from the boundary between Learning Phase 1 and Learning Phase 2, search along each branch for courses with a learning difficulty greater than N2. Since K7, K8, and K9 are all greater than N2, all related courses preceding K7, K8, and K9 (i.e., K3 and K4) are grouped into one Learning Phase 2. For K1, K5, and K6, since they are not greater than N2 and have reached the end of the learning phase, they are directly incorporated into Learning Phase 2.

[0089] Following this pattern, the remaining K7, K8, and K9 are divided into a learning stage 3.

[0090] Since the learning difficulty thresholds are sorted from lowest to highest, the difficulty of each learning stage increases sequentially, allowing learners to progress gradually from easy to difficult, from simple to complex. Additionally, it's generally believed that earlier courses are less difficult than later ones. However, this may not be the case in some special circumstances, which could lead to the learning stages not always being appropriate. The defined learning stages primarily serve as a suggestion; learners can edit and modify the learning stages in the knowledge map based on their actual situation to create a learning plan tailored to their individual needs.

[0091] (2) After completing the learning stage division of all learning difficulty thresholds, determine whether the centrality of each related course is greater than the preset centrality threshold; if the centrality of any related course is greater than the centrality threshold, then divide the related course and the related courses before it in the current learning stage into the previous learning stage.

[0092] The centrality of a related course is the number of related courses associated with it. A higher centrality indicates a more important course. In this embodiment, related courses with a centrality greater than a centrality threshold, along with previous related courses, are placed in the previous learning stage to prioritize learning more important courses. For example, in... Figure 4 If the centrality of K4 is greater than the centrality threshold, then the courses K4 and earlier will be upgraded from learning stage 2 to learning stage 1.

[0093] (3) Calculate the total time cost corresponding to each learning stage based on the time cost of each related course within each learning stage; if the total time cost corresponding to any learning stage is greater than the preset time cost threshold, then the learning stage will be further divided into two learning stages.

[0094] This embodiment presets a time cost threshold, which is determined based on factors such as learning objectives, learning plans, learner's energy, and time constraints. For example, the time cost threshold is set at 120 hours. The total time cost of each learning stage is compared with the threshold. If the total time cost of a certain learning stage exceeds the threshold, then that learning stage needs to be further divided to avoid any single learning stage containing too much content and taking too long.

[0095] For example, a target learning difficulty threshold can be determined based on the learning difficulty thresholds corresponding to the current learning stage and the previous learning stage. The target learning difficulty threshold lies between the learning difficulty thresholds corresponding to the current and previous learning stages, such as the median of these thresholds. Within this learning stage, the first relevant course with a learning difficulty greater than the target learning difficulty threshold is found at each branch from front to back, serving as the target dividing point. Based on the target dividing point, this learning stage is further divided into two learning stages.

[0096] In some embodiments, the learning difficulty threshold can be determined in the following ways:

[0097] Obtain learners' historical learning data;

[0098] Based on historical learning data, assess learners' learning abilities over time.

[0099] Based on the learning ability time series, within the preset learning difficulty threshold range, the learning difficulty threshold is determined by a pre-trained neural network model.

[0100] Learners' historical learning data includes information from multiple aspects. From the perspective of the learning process, it includes the duration of each learning session, the time of learning, and the completion of assignments and tests; from the perspective of learning outcomes, it covers exam scores, assignment scores, and the degree of mastery of knowledge points. On the learning platform, the system can record learners' learning behaviors in different courses, the accuracy rate of completing exercises, and other data, which can comprehensively reflect the learner's learning journey and performance.

[0101] A learning ability time series refers to the changes in a learner's learning ability over time. It assesses learning ability by analyzing historical learning data. For example, based on data such as changes in test scores, the speed and quality of homework completion at different points in time, statistical methods or specific assessment models are used to quantify learning ability. For instance, if a learner's accuracy in completing homework and their test scores gradually increase during a period of time while studying mathematics, then their mathematical learning ability can be considered to be on an upward trend during this period. Arranging learning abilities at different points in time forms a learning ability time series. This time series helps to understand the dynamic changes in a learner's learning ability and to discover the characteristics and trends of their learning ability at different stages.

[0102] Based on the learning ability time series, learners' learning ability-related data are input into a pre-trained neural network model. The model, based on the input data and learned patterns and rules, determines a suitable learning difficulty threshold within a preset range. For example, if the learner's learning ability time series shows a gradual increase in ability, the neural network model might select a relatively high learning difficulty threshold within the range to recommend more challenging learning content; conversely, if learning ability shows a downward trend, the model might select a lower difficulty threshold to ensure the learner can gradually regain confidence and ability. This determined learning difficulty threshold better adapts to the learner's actual learning ability, improving learning outcomes.

[0103] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0104] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for constructing a knowledge progression map, characterized in that, include: Acquire the target knowledge points; Based on the target knowledge points, relevant courses and the relationships between them are determined from a pre-constructed course knowledge graph. Based on the relevant courses and the associated relationships, construct a knowledge base map; Based on at least one preset course attribute of the relevant courses, the knowledge base map is divided into learning stages to obtain a knowledge progression map containing multiple learning stages. The relationships include prerequisites and subsequent courses; correspondingly, constructing a knowledge base map based on the relevant courses and the relationships includes: Based on the prerequisite and successor relationships among the related courses, arrows are used to connect the related courses that have an association, resulting in a tree-like learning path for the related courses; Based on the aforementioned tree-like learning path, a knowledge base map is constructed; The preset course attributes include: learning difficulty; correspondingly, the step of dividing the knowledge base map into stages based on at least one preset course attribute of the relevant courses includes: Obtain at least one learning difficulty threshold and sort them from smallest to largest; In the knowledge base map, the first relevant course with a learning difficulty greater than the first learning difficulty threshold is found on each branch from front to back and is taken as the first dividing point; all relevant courses before the first dividing point are divided into a learning stage. Starting from each first dividing point, find the first relevant course on each branch with a learning difficulty greater than the next learning difficulty threshold, and use it as the second dividing point; divide all relevant courses before the second dividing point into a learning stage; If there are no courses with a learning difficulty greater than the corresponding learning difficulty threshold on a branch, then all related courses on that branch and before it will be merged into the current learning stage. Repeat the above steps until all learning difficulty thresholds are divided into learning stages; The preset course attributes also include: centrality; the centrality of the related course is: the number of related courses associated with the related course; correspondingly, the step of dividing the knowledge base map into stages based on at least one preset course attribute of the related courses also includes: After completing the learning stage division of all learning difficulty thresholds, determine whether the centrality of each relevant course is greater than the preset centrality threshold. If the centrality of any related course is greater than the centrality threshold, then the related course and previous related courses in the current learning stage will be assigned to the previous learning stage. The preset course attributes also include: time cost; correspondingly, the step of dividing the knowledge base map into learning stages based on at least one preset course attribute of the relevant courses also includes: Calculate the total time cost for each learning stage based on the time cost of each relevant course within each learning stage. If the total time cost corresponding to any learning stage is greater than the preset time cost threshold, then the learning stage will be further divided into two learning stages.

2. The method for constructing a knowledge progression map as described in claim 1, characterized in that, The learning phase is further divided into two learning phases, including: A target learning difficulty threshold is determined based on the learning difficulty threshold corresponding to the current learning stage and the previous learning stage; wherein, the target learning difficulty threshold is located between the learning difficulty thresholds corresponding to the current learning stage and the previous learning stage. During this learning phase, the first relevant course with a learning difficulty greater than the target learning difficulty threshold is found on each branch from front to back, and is used as the target dividing point; Based on the target dividing point, this learning phase is further divided into two learning phases.

3. The method for constructing a knowledge progression map as described in claim 1, characterized in that, The process of obtaining at least one learning difficulty threshold includes: Obtain learners' historical learning data; Based on the historical learning data, evaluate the learner's learning ability time series; Based on the learning ability time series, within a preset learning difficulty threshold range, the learning difficulty threshold is determined by a pre-trained neural network model.

4. The method for constructing a knowledge progression map as described in any one of claims 1 to 3, characterized in that, After obtaining the knowledge progression map containing multiple learning stages, the process also includes: Set advanced tests and advancement requirements for each learning stage; After a learner completes any learning stage, the learner is tested using an advanced test for that stage, and the test results are obtained. Based on the exam results, determine whether the advancement conditions are met; If the advancement conditions are met, the next learning stage of the knowledge advancement map is opened; if the advancement conditions are not met, the next learning stage of the knowledge advancement map is not opened.

5. The method for constructing a knowledge progression map as described in any one of claims 1 to 3, characterized in that, The method further includes: Access to multimodal course resources; Multimodal course resources are transformed into structured text data, and entity extraction is performed on the text data to obtain the corresponding knowledge points. The knowledge points are analyzed to identify course associations, and a course knowledge graph is constructed.

6. The method for constructing a knowledge progression map as described in claim 5, characterized in that, The acquisition of multimodal course resources includes: acquiring multimodal course resources from websites through web crawling technology.

Images