Learning difficulty intelligent adjustment device and method

Abstract

The application relates to the computer technical field and discloses a learning difficulty intelligent adjustment device and method, which comprises the following steps: extracting key knowledge elements from pre-acquired user learning data to construct a learning graph; analyzing the knowledge mastery degree and learning ability of a user according to the learning graph to obtain a user ability evaluation result; combining the user ability evaluation result and a node correlation relationship in the learning graph to generate corresponding learning content; analyzing the update situation of knowledge elements and the change situation of the user learning ability in the user learning process, dynamically updating the learning graph to obtain an updated learning graph; and adjusting the difficulty of the learning content in real time according to the updated learning graph to obtain updated learning content, so that the learning difficulty is intelligently adjusted; the learning content difficulty is intelligently adjusted, the needs of different users in different learning stages are met, and the learning effect and the knowledge mastery degree are effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a learning difficulty intelligent adjustment device and method. BACKGROUND

[0002] The existing learning system provides unified learning content and difficulty, without considering the difference between the knowledge mastery and learning ability of different students; in the learning process of the user, the learning content and difficulty cannot be adjusted in time according to the updating of the knowledge elements and the change of the learning ability in the learning process of the students; the correlation between different subjects and different knowledge points is poor, and it is difficult to build a comprehensive and accurate knowledge system, resulting in that the arrangement of the learning content lacks systematicness and logicalness, and the students are difficult to form a complete knowledge network.

[0003] To solve at least one of the above problems, the present application provides a learning difficulty intelligent adjustment device and method. SUMMARY

[0004] In view of the deficiencies of the prior art, the main purpose of the present application is to provide a learning difficulty intelligent adjustment device and method, which can effectively solve the problems in the background art. The specific technical scheme of the present application is as follows:

[0005] A learning difficulty intelligent adjustment method, comprising:

[0006] extracting key knowledge elements from pre-acquired user learning data to construct a learning graph;

[0007] analyzing the knowledge mastery and learning ability of the user according to the learning graph to obtain a user ability evaluation result;

[0008] combining the user ability evaluation result and the node association relationship in the learning graph to generate corresponding learning content, wherein the nodes include knowledge nodes and ability nodes;

[0009] analyzing the updating of the knowledge elements and the change of the learning ability of the user in the learning process of the user, dynamically updating the learning graph to obtain an updated learning graph;

[0010] adjusting the difficulty of the learning content in real time according to the updated learning graph to obtain updated learning content, so as to realize learning difficulty intelligent adjustment.

[0011] Specifically, the extracting key knowledge elements from pre-acquired user learning data to construct a learning graph comprises:

[0012] extracting corresponding knowledge points from the pre-acquired user learning data to obtain key knowledge elements;

[0013] The key knowledge elements are taken as knowledge nodes, the corresponding knowledge nodes are connected by analyzing the correlation between the key knowledge elements, and an initial learning graph is obtained;

[0014] The user learning data is fused and mapped into the initial learning graph, and a corresponding learning graph is obtained.

[0015] Specifically, according to the learning graph, the knowledge mastery and learning ability of the user are analyzed, and a user ability evaluation result is obtained, including:

[0016] According to the correlation between the knowledge nodes in the learning graph, the answering situation of the user on each knowledge node is analyzed, and the knowledge mastery of the user is obtained;

[0017] The knowledge nodes reflecting different subject knowledge points in the learning graph are aggregated, and multiple types of subject nodes are obtained;

[0018] For each type of subject node, the knowledge mastery of the user is evaluated respectively, and the knowledge mastery of each type of subject is obtained.

[0019] According to the knowledge mastery of each type of subject, a corresponding ability node is added to each type of subject node and connected;

[0020] In the user learning process, according to the change of the knowledge mastery of each type of subject of the user, the attributes of the ability node are dynamically updated, and an updated ability node is obtained;

[0021] Combined with the updated ability node corresponding to each type of subject, the learning ability of the user is evaluated, and a user ability evaluation result is obtained.

[0022] Specifically, combined with the user ability evaluation result and the node correlation in the learning graph, corresponding learning content is generated, including:

[0023] According to the user ability evaluation result, a target template library of corresponding difficulty is matched in a preset question template library;

[0024] By analyzing the node correlation in the learning graph, a corresponding weight is assigned to each node;

[0025] Combined with the target template library and the weight, a corresponding learning content is generated through a preset content generation model.

[0026] Specifically, in the user learning process, the updating situation of the knowledge elements and the change situation of the user learning ability are analyzed, the learning graph is dynamically updated, and an updated learning graph is obtained, including:

[0027] The updating situation of knowledge elements and the change situation of user learning ability are analyzed in the user learning process to obtain an updating learning state;

[0028] According to the updating learning state, nodes and node connection relationships in the learning graph are updated to obtain a first updating learning graph;

[0029] The structure of the first updating learning graph is updated by analyzing learning ability levels of different users to obtain an updating learning graph.

[0030] Specifically, according to the updating learning state, nodes and node connection relationships in the learning graph are updated to obtain a first updating learning graph, including:

[0031] According to the updating learning state, when a new knowledge point is added, a corresponding knowledge node is added in the learning graph, and a connection relationship between the new knowledge point and an existing knowledge node is established by analysis to obtain a first updating node;

[0032] When the user's knowledge mastery degree is improved, the corresponding knowledge node attribute and connection relationship are updated to obtain a second updating node;

[0033] According to the second updating node, the corresponding ability node attribute and connection relationship are updated to obtain a third updating node;

[0034] The nodes and node connection relationships in the learning graph are updated in combination with the first updating node, the second updating node and the third updating node to obtain a first updating learning graph.

[0035] Specifically, the structure of the first updating learning graph is updated by analyzing learning ability levels of different users to obtain an updating learning graph, including:

[0036] According to the user's learning ability, the user is classified by a preset learning ability evaluation model to obtain multi-level users, wherein the multi-level users include primary users, intermediate users and advanced users;

[0037] For the primary users, the structure of the first updating learning graph is simplified to obtain an updating learning graph of the primary users;

[0038] For the intermediate users, the node association relationship between the knowledge nodes in the first updating learning graph is improved to obtain an updating learning graph of the intermediate users;

[0039] For the advanced users, the knowledge nodes in the first updating learning graph are expanded to obtain an updating learning graph of the advanced users.

[0040] Specifically, according to the updated learning graph, the difficulty of the learning content is adjusted in real time to obtain updated learning content, so as to realize intelligent adjustment of learning difficulty, including:

[0041] According to the updated learning graph, the knowledge mastery degree of the user for each knowledge node is analyzed to obtain an updated knowledge mastery degree;

[0042] Based on the updated knowledge mastery degree, matching is performed in the updated learning graph to obtain an updated node path;

[0043] According to the updated node path, matching is performed in a preset question template library to obtain an updated template;

[0044] Based on the updated template, a corresponding updated learning content is generated through a preset content updating model to realize intelligent adjustment of learning difficulty.

[0045] Specifically, based on the updated knowledge mastery degree, matching is performed in the updated learning graph to obtain an updated node path, including:

[0046] Based on the updated knowledge mastery degree, nodes with a knowledge mastery degree change greater than a preset threshold are extracted in the updated learning graph to obtain a plurality of to-be-updated knowledge nodes;

[0047] By analyzing the association relationship between knowledge nodes, the plurality of to-be-updated knowledge nodes are connected to obtain an updated node path.

[0048] A learning difficulty intelligent adjustment device for realizing the learning difficulty intelligent adjustment method, including:

[0049] A learning graph construction module extracts key knowledge elements from pre-acquired user learning data to construct a learning graph;

[0050] A user evaluation module analyzes the knowledge mastery degree and learning ability of the user according to the learning graph to obtain a user ability evaluation result;

[0051] A learning content generation module generates corresponding learning content in combination with the user ability evaluation result and the node association relationship in the learning graph, wherein the node includes a knowledge node and an ability node;

[0052] A learning graph updating module analyzes the update situation of knowledge elements and the change situation of user learning ability during user learning, dynamically updates the learning graph, and obtains an updated learning graph;

[0053] The learning difficulty adjustment module adjusts the difficulty of the learning content in real time according to the updated learning graph, obtains updated learning content, and realizes intelligent adjustment of learning difficulty.

[0054] Compared with the prior art, the present application has the following beneficial effects:

[0055] The present application extracts key knowledge elements from user learning data, constructs a learning graph, dynamically analyzes and evaluates the knowledge mastery and learning ability of the user according to the learning graph, and updates the ability evaluation result in real time, dynamically updates the learning graph in combination with the change of the learning ability of the user, and adjusts the difficulty of the learning content in real time according to the updated learning graph, realizes intelligent adjustment of learning difficulty, meets the needs of different users in different learning stages, and improves the knowledge mastery and learning ability of the user. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 A working flow chart of a learning difficulty intelligent adjustment method in the embodiment 1 of the present application;

[0057] Figure 2 A schematic diagram of learning graph node updating in the embodiment 1 of the present application;

[0058] Figure 3 A schematic diagram of primary user updating learning graph in the embodiment 1 of the present application;

[0059] Figure 4 A schematic diagram of intermediate user updating learning graph in the embodiment 1 of the present application;

[0060] Figure 5 A schematic diagram of advanced user updating learning graph in the embodiment 1 of the present application;

[0061] Figure 6 A structural schematic diagram of a learning difficulty intelligent adjustment device in the embodiment 2 of the present application. DETAILED DESCRIPTION

[0062] In order to make the above objectives, characteristics and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0063] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be practiced in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the concept of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0064] Secondly, the "one embodiment" or "an embodiment" referred to herein means a specific feature, structure, characteristic, or combination of features and / or characteristics described herein that can be included in at least one implementation of the present application. The various appearances of "in one embodiment" or "in an embodiment" in the specification do not all refer to the same embodiment nor do they necessarily refer to a single implementation of the present application, nor are they mutually exclusive.

[0065] Embodiment 1

[0066] The embodiment provides a learning difficulty intelligent adjustment method, as shown in the method, the method comprises the following steps of: Figure 1

[0067] S101, extracting key knowledge elements from pre-acquired user learning data, and constructing a learning graph;

[0068] S102, analyzing the knowledge mastery and learning ability of the user according to the learning graph, and obtaining a user ability evaluation result;

[0069] S103, combining the user ability evaluation result and the node association relationship in the learning graph, and generating corresponding learning content, wherein the node comprises a knowledge node and an ability node;

[0070] S104, analyzing the update situation of the knowledge elements and the change situation of the user learning ability in the user learning process, dynamically updating the learning graph, and obtaining an updated learning graph;

[0071] S105, adjusting the difficulty of the learning content in real time according to the updated learning graph, obtaining updated learning content, and realizing learning difficulty intelligent adjustment.

[0072] The embodiment constructs a learning graph according to user learning data, and in the user learning process, the knowledge mastery and learning ability of the user are evaluated in real time, the difficulty of the learning content is adjusted in real time, the learning content is always matched with the ability of the user, learning difficulties or low learning efficiency caused by unsuitable difficulty are avoided, and the learning effect is improved; the problem that in the traditional learning system, single learning content causes the user to be unable to adapt to the learning difficulty, and the learning effect is poor is solved.

[0073] ​In the embodiment, first, in the user learning process, the learning data of the user is collected, including the viewing record of the online course, the homework completion, the test score, etc., the collected data is data cleaned, and the redundant or wrong data is removed; second, the key knowledge elements are extracted from the preprocessed data, the key knowledge elements can reflect the core knowledge points in the learning process, the logical relationship between the knowledge elements is analyzed based on the key knowledge elements, and the corresponding learning graph is constructed; by constructing the learning graph, it is helpful for the user to establish a systematic knowledge system, according to the connection between the knowledge points, the user's mastery of knowledge is analyzed, and the learning content is dynamically adjusted.

[0074] Specifically, according to the learning graph, the learning data of the user at each knowledge node is analyzed, the user's mastery of each knowledge point is calculated, the user's learning ability is analyzed based on the user's mastery of each knowledge point, and the user's ability evaluation result is obtained; by obtaining the accurate user ability evaluation result, the user's learning situation can be accurately understood, and the basis for generating learning content is provided; according to the user ability evaluation result, the appropriate learning content difficulty level is determined for the user, for example, the primary user is suitable for the basic difficulty content, the intermediate user can select the medium difficulty content, and the advanced user can challenge the high difficulty content; according to the corresponding learning content difficulty level, the learning content suitable for the current learning ability of the user is generated; by generating learning content with corresponding difficulty level for users with different learning abilities, the learning needs of different users can be met, and the pertinence and effectiveness of learning are improved.

[0075] In the user learning process, the learning ability and knowledge point mastery of the user are updated in real time, according to the change of the user's learning ability, the learning of new knowledge and the mastery of old knowledge, the node attributes and connection relationship of the learning graph are dynamically updated, so that the learning graph can always reflect the real-time learning state of the user, timely adapt to the update of knowledge and the development of user learning ability, and ensure the continuous effectiveness of learning; according to the updated learning graph, the difficulty of the learning content is adjusted in real time, so that the learning content always matches the user's ability; when the user's ability improves, the difficulty of the learning content is increased; when the user encounters difficulties in some knowledge nodes, the difficulty of the related content is reduced; the corresponding learning content is updated to obtain updated learning content, and the order of the learning content is rearranged to adapt to the change of difficulty; by updating the difficulty of the learning content in real time, the difficulty of the learning content can always match the user's ability, the efficiency and effect of learning are improved, the user's interest in learning is avoided due to the difficulty of the learning content, and the learning motivation and enthusiasm are enhanced.

[0076] The application extracts key knowledge elements from user learning data, constructs a learning graph, dynamically analyzes and evaluates the knowledge mastery and learning ability of the user according to the learning graph, and updates the ability evaluation result in real time, dynamically updates the learning graph in combination with the change of the learning ability of the user, and adjusts the difficulty of the learning content in real time according to the updated learning graph, realizes intelligent adjustment of the learning difficulty, meets the needs of different users in different learning stages, and improves the knowledge mastery and learning ability of the user.

[0077] Further, the extracting of the key knowledge elements from the pre-acquired user learning data and the construction of the learning graph comprise:

[0078] S201, extracting corresponding knowledge points from pre-acquired user learning data to obtain key knowledge elements;

[0079] S202, taking the key knowledge elements as knowledge nodes, connecting the corresponding knowledge nodes by analyzing the association relationship between the key knowledge elements to obtain an initial learning graph;

[0080] S203, fusing the user learning data and mapping the user learning data into the initial learning graph to obtain a corresponding learning graph.

[0081] In the embodiment, after the user learning data is collected and data cleaning is performed on the user learning data, the text type learning data is analyzed by using a natural language processing technology to identify the knowledge points therein; for the structured data, the relevant knowledge points are extracted according to the fields and meanings of the data; for example, in the mathematical examination score data, the knowledge points of function, geometry and probability are extracted according to the chapters and types to which the questions belong; the multiple knowledge points are extracted as key knowledge elements, the core knowledge content can be obtained from the massive learning data by extracting the key knowledge elements, and the pertinence of learning and the quality of the learning graph are improved.

[0082] Specifically, after extracting the key knowledge elements, the correlation between the knowledge elements is analyzed, the key knowledge elements are taken as knowledge nodes, the connection is established by analyzing the correlation between the knowledge elements, and the initial learning graph is constructed; according to the characteristics of the subject knowledge and the law of learning, the correlation type is set; for example, in the mathematics discipline, there are “preposition relationship” (such as learning function before mastering basic algebraic operation), “application relationship” (such as the application of geometric knowledge in actual measurement), etc.; in Chinese learning, there are “theme correlation” (such as different articles unfolding around the same theme), “syntax correlation” (such as the correlation between words and sentence structure), etc.; by analyzing the context information, the frequency and order of knowledge use in the learning data, the correlation between the knowledge elements is analyzed, the key knowledge elements are taken as knowledge nodes, the connection is established between the nodes according to the correlation, and the initial learning graph is constructed; by constructing the initial learning graph, the correlation between the knowledge points is quickly established, which provides a structural basis for analyzing the knowledge mastery of the user and generating the personalized learning path.

[0083] Specifically, the preprocessed user learning data is matched with the knowledge nodes in the initial learning graph; for example, the data such as the correct rate of the user's answering at a knowledge point and the learning time are associated with the corresponding knowledge nodes in the learning graph; the user learning data is added as an attribute for each knowledge node, for example, the “mastering degree score” attribute is added in the knowledge node, and the score is calculated according to the user's answering at the knowledge point; according to the data mapping result, the initial learning graph is updated to obtain the learning graph containing the user learning information; by adding the personalized features in the learning graph, the learning status of each user can be accurately reflected, which provides a basis for generating corresponding learning content.

[0084] Further, according to the learning graph, the knowledge mastery and learning ability of the user are analyzed to obtain the user ability evaluation result, including:

[0085] S301, according to the correlation between the knowledge nodes in the learning graph, the answering of the user at each knowledge node is analyzed to obtain the knowledge mastery of the user;

[0086] S302, the knowledge nodes reflecting different subject knowledge points in the learning graph are aggregated to obtain multi-class subject nodes;

[0087] S303, for each class of subject nodes, the knowledge mastery of the user is evaluated respectively to obtain the knowledge mastery of each class of subject;

[0088] S304, according to the knowledge mastery of each class of subject, the corresponding ability nodes are added to each class of subject node and connected;

[0089] S305, in the user learning process, according to the change of the user's mastery of each type of subject knowledge, the attribute of the ability node is dynamically updated, and the updated ability node is obtained;

[0090] S306, in combination with the updated ability node corresponding to each type of subject, the user learning ability is evaluated, and the user ability evaluation result is obtained.

[0091] In this embodiment, the answer data of the user on each knowledge node is obtained from the learning system, including the correct rate of answering, the answering time, the error type and the like, according to the association relationship of the knowledge nodes in the learning graph, the importance of each knowledge node is analyzed, for example, when a knowledge node is the prerequisite knowledge of a plurality of other knowledge nodes, the importance of the knowledge node is relatively high; in combination with the answering data and the association information of the knowledge nodes, the knowledge mastery degree score of each knowledge node is calculated, the weighted average method is adopted, the importance of the knowledge node is higher, and the weight of the answering condition in the calculation is larger; for example, the correct rate of answering of a knowledge node is 80%, and the weight of the node is 0.6, so the knowledge mastery degree score of the knowledge node is 80% x 0.6 = 0.48 points; through the calculation of the knowledge mastery degree score, the mastery degree of the user to each knowledge node can be accurately evaluated.

[0092] Specifically, according to the subject attribute of the knowledge point, the knowledge nodes in the learning graph are classified, for example, the knowledge nodes of mathematics, physics, chemistry and the like are classified respectively; a subject node is created for each subject, and the knowledge nodes belonging to the subject are associated with the subject node; for each type of subject node, the knowledge mastery degree scores of all the knowledge nodes under each type of subject node are summarized, and the weighted average is performed according to the importance of the knowledge nodes, so as to calculate the knowledge mastery degree of each type of subject; through the analysis of the user's learning level of the subject, the corresponding learning plan can be formulated for different subjects, and the corresponding learning content is generated.

[0093] Specifically, as Figure 2As shown, according to the calculated knowledge mastery degree of each type of subject, the learning ability of the user for each type of subject is evaluated, a corresponding ability node is added to each type of subject node, and a connection is established between the subject node and the ability node, and the ability node is valued according to the knowledge mastery degree; during the user's learning process, the user's knowledge mastery degree change in each subject is continuously monitored, new answer data and learning records are regularly collected, and the user's learning ability in each subject is re-evaluated according to the change of the knowledge mastery degree, for example, when the user's knowledge mastery degree in a subject significantly improves, it indicates that the user's learning ability in that subject has been enhanced, the attribute value of the corresponding ability node is increased, and the updated ability node is obtained; the updated ability node can reflect the change of the user's learning ability in real time, ensuring the accuracy and timeliness of the evaluation result. The attribute data of the updated ability node corresponding to each type of subject is combined, and the user's learning ability evaluation result is calculated by weighted summation, which can comprehensively and objectively evaluate the user's learning ability and provide data support for formulating personalized learning plans and development plans.

[0094] Further, in combination with the user ability evaluation result and the node association relationship in the learning graph, corresponding learning content is generated, including:

[0095] S401, according to the user ability evaluation result, matching the target template library of corresponding difficulty in the preset question template library;

[0096] S402, by analyzing the node association relationship in the learning graph, assigning a corresponding weight to each node;

[0097] S403, in combination with the target template library and the weight, generating corresponding learning content through a preset content generation model.

[0098] In this embodiment, according to the user ability evaluation result, the user is analyzed, including primary, intermediate, and advanced, each level corresponding to different learning ability; the questions in the preset question template library are marked with corresponding difficulty, and the difficulty level is determined by factors such as the number of knowledge points involved in the question, the complexity, and the thinking ability required for solving the problem, including easy, medium, and difficult; according to the level corresponding to the user ability evaluation result, the question templates of the corresponding difficulty level are selected from the preset question template library to form a target template library; for example, when the user's ability is evaluated as intermediate, the question templates of medium difficulty are selected and put into the target template library; by selecting the target template library of corresponding difficulty, the difficulty of the learning content can be adapted to the user's ability, improving the pertinence and effectiveness of learning, and avoiding wasting time on inappropriate difficult questions.

[0099] Specifically, according to the corresponding node association relationship in the learning graph, the importance of each node is analyzed, and the node is assigned a weight; by analyzing the association relationship between the nodes in the learning graph, including the order, the cause and effect relationship, the containing relationship and the like, a corresponding weight distribution principle is formulated, for example, the node in the core position of the knowledge system and having an important influence on subsequent learning has a higher weight; and the node with weak auxiliary and correlation has a lower weight; in this embodiment, the analytic hierarchy process is specifically used to assign a weight to each node according to the weight distribution principle; by assigning the weight, the generated learning content is more consistent with the structure and key points of the knowledge system, the key knowledge is more embodied and strengthened in the learning content, and it is helpful for the user to better master the key knowledge points.

[0100] Specifically, according to the calculated node weight and the corresponding target template library, the question templates related to each node are selected from the target template library, the number of question templates corresponding to the knowledge node with a higher weight can be relatively more, so as to ensure that the key knowledge is fully practiced, the selected question templates are combined according to the association relationship and logical order of the nodes in the learning graph, for example, the questions corresponding to the pre-knowledge nodes are arranged first, and then the questions of the subsequent related knowledge nodes are arranged, so that the learning content has continuity; the combined question templates are further processed by using a preset content generation model, such as adding necessary instructions, prompts and the like, and finally a complete learning content is generated; the learning content generated according to the target template library and the corresponding knowledge node weight not only considers the ability level of the user, but also highlights the key points and logical structure of the knowledge system, and improves the efficiency and quality of the learning content generation.

[0101] Further, in the user learning process, the updating situation of the knowledge elements and the change situation of the user learning ability are analyzed, and the learning graph is dynamically updated to obtain an updated learning graph, including:

[0102] S501, in the user learning process, the updating situation of the knowledge elements and the change situation of the user learning ability are analyzed to obtain an updated learning state;

[0103] S502, according to the updated learning state, the nodes and the node connection relationship in the learning graph are updated to obtain a first updated learning graph;

[0104] S503, by analyzing the learning ability levels of different users, the structure of the first updated learning graph is updated to obtain an updated learning graph.

[0105] In the learning process of the user, the nodes and the connection relationship between the nodes in the learning graph are dynamically updated according to the change of the learning state of the user, and the structure of the learning graph is optimized according to the learning ability level of different users, so as to provide a learning path and a knowledge system that meet the characteristics of each user, enhance the pertinence and effectiveness of learning, and improve the learning experience of the user.

[0106] In the embodiment, the learning data in the learning process of the user is collected and analyzed in real time, and new learned knowledge elements are identified, for example, the course chapters learned by the user, the material topics read by the user, etc. are analyzed to determine whether new knowledge points are introduced, and meanwhile, the learning time and the learning times of each knowledge element are recorded to obtain the frequency and degree of updating of the new knowledge points. The change of the learning ability of the user is evaluated from multiple aspects, and whether the mastery degree of knowledge is improved and whether the speed of mastering knowledge is accelerated is determined by comparing the change of the correct rate and the answering time of the same knowledge points of the user in different time periods. The improvement or change of the learning ability of the user is analyzed by comprehensively analyzing these factors. The updating of the knowledge elements and the change of the learning ability are integrated to obtain an updated learning state. By analyzing the updated learning state of the user, the learning progress and the change of the ability of the user can be understood in time, so that the learning graph is updated.

[0107] Specifically, according to the updated learning state of the user, the nodes and the connection relationship between the nodes in the learning graph are updated, specifically by adding new knowledge nodes, updating the node attributes and updating the connection relationship between the nodes, so as to realize the updating of the learning graph. By updating the nodes and the connection relationship, the newly learned knowledge elements and the new association between the knowledge can be reflected in time, so as to ensure that the learning graph is always consistent with the learning progress and the knowledge mastery of the user, and accurate learning navigation is provided.

[0108] Specifically, by analyzing the learning ability level of the user, the structure of the learning graph is further updated, the user is classified according to the learning ability of the user, and the structure of the learning graph is optimized for users of different levels, so that it is more consistent with the learning needs of users of different levels. For example, for users with strong learning ability, more challenging and deep knowledge structures can be provided. For users with weak learning ability, the graph structure can be simplified to highlight basic knowledge and key content to help them better understand and master knowledge. By optimizing the structure of the learning graph, the individualized needs of users with different learning abilities can be met, and the adaptability and effectiveness of learning can be improved.

[0109] Further, according to the updated learning state, the nodes and the connection relationship between the nodes in the learning graph are updated to obtain a first updated learning graph, including:

[0110] S601, according to the updated learning state, when a new knowledge point is added, a corresponding knowledge node is added in the learning graph, and a connection relationship between the new knowledge point and the existing knowledge node is established by analyzing the discipline to which the new knowledge point belongs, to obtain a first updated node;

[0111] S602, when the user's knowledge mastery degree is improved, the corresponding knowledge node attribute and connection relationship are updated to obtain a second updated node;

[0112] S603, according to the second updated node, the corresponding ability node attribute and connection relationship are updated to obtain a third updated node;

[0113] S604, in combination with the first updated node, the second updated node and the third updated node, the nodes and node connection relationships in the learning graph are updated to obtain a first updated learning graph.

[0114] In this embodiment, according to the updated learning state of the user, when the user learns a new knowledge point, a new knowledge node is created according to the new knowledge point, and the attributes of the new knowledge point are configured, including the knowledge point name, the discipline to which it belongs, and the like. By analyzing the discipline to which the new knowledge point belongs and the relationship with the existing knowledge nodes in the learning graph, a connection relationship is established between the new knowledge node and the related existing knowledge nodes. According to the new knowledge node, the learning graph is updated to obtain a first updated node. By updating the learning graph in a timely manner, the learning graph is kept consistent with the user's learning progress, so that the learning content is updated accordingly.

[0115] Specifically, according to the user's mastery of the knowledge point, the corresponding knowledge node attribute and connection relationship are updated to obtain a second updated node. According to the user's learning data, including the correct rate of answering questions, homework completion, test scores and the like, the user's mastery of each knowledge node is evaluated. For example, the user's correct rate in a series of tests on the "function" knowledge node has increased from 60% to 80%, indicating that the user's mastery of the "function" knowledge has improved. For the knowledge nodes whose mastery degree is improved, their attributes are updated and their mastery degree values are correspondingly improved. According to the association relationship between the knowledge nodes, considering that the mastery degree of a knowledge node is improved, the association strength with other related knowledge nodes is enhanced, and the node connection relationship is updated. For example, the improvement of the mastery degree of the "function" knowledge node makes the connection relationship between the "function" knowledge node and the "derivative" knowledge node closer, because function is the basis of derivative and high mastery of function knowledge helps to understand derivative. At this time, the weight of the connection edge between the two nodes can be increased to update the connection relationship. Thus, the change of the user's knowledge mastery degree is accurately reflected.

[0116] Specifically, according to the change of the user's mastery of the knowledge node in the second updated node, the corresponding ability node is updated to obtain a third updated node; first, the ability node related to the second updated node is determined; according to the update of the knowledge node, the attribute of the related ability node is updated, the attribute value of the corresponding ability node is improved, and the connection relationship between the ability node and the knowledge node is updated; by updating the ability node, the change of the user's learning can be reflected from the ability angle, and more accurate ability level basis is provided for the formulation of the personalized learning plan, so that the learning plan is more targeted.

[0117] Meanwhile, all new nodes and updated nodes in the first updated node, the second updated node and the third updated node are merged into the learning graph to update the learning graph to obtain a first updated learning graph, and the connection relationship between the updated nodes is checked and adjusted to ensure the accuracy and logic of the connection relationship, and to avoid isolated nodes or unreasonable connections; by updating the learning graph, the knowledge growth and ability improvement of the user in the learning process can be reflected in real time and dynamically, so as to formulate more targeted learning plan and learning content for the user and improve the learning effect.

[0118] Further, the structure of the first updated learning graph is updated by analyzing the learning ability levels of different users to obtain an updated learning graph, comprising:

[0119] S701, according to the learning ability of the user, the user is graded by a preset learning ability evaluation model to obtain multi-level users, wherein the multi-level users include primary users, intermediate users and advanced users;

[0120] S702, for the primary user, the structure of the first updated learning graph is simplified to obtain an updated learning graph of the primary user;

[0121] S703, for the intermediate user, the node association relationship between the knowledge nodes in the first updated learning graph is improved to obtain an updated learning graph of the intermediate user;

[0122] S704, for the advanced user, the knowledge node in the first updated learning graph is expanded to obtain an updated learning graph of the advanced user.

[0123] In the embodiment, the user is classified according to the learning ability of the user, and factors such as knowledge mastery, learning speed, problem solving ability, and learning initiative are comprehensively considered. The learning ability of the user is evaluated by a preset learning ability evaluation model. The learning ability evaluation model can be a statistical analysis model or a machine learning model. In the embodiment, the learning ability evaluation model is a decision tree model. A large amount of historical data is used to train the decision tree model to obtain a pre-trained learning ability evaluation model. The learning ability evaluation factors of the user are input into the pre-trained learning ability evaluation model, and the model outputs the classification result of the user, including a primary user, an intermediate user, and an advanced user. According to the learning ability of the user, the learning graph of the user of different levels is optimized, and a corresponding personalized learning scheme is formulated. The learning graph of the user of different learning ability levels is updated, personalized learning support is provided for each user, and the needs of the user in different learning stages are met.

[0124] As shown in Figure 3 , for a primary user, the learning graph is updated by simplifying the structure of the first updated learning graph. According to the characteristics of the primary user, a corresponding simplified learning graph strategy is formulated, for example, the number of knowledge nodes is reduced, the core knowledge points are retained, and the connection relationship between the knowledge points is updated. By filtering out the core knowledge nodes, the connection relationship between the knowledge nodes is simplified, the most direct and important association is retained, and the updated learning graph of the primary user is obtained. By simplifying the structure, the learning difficulty can be reduced, and the primary user can more easily start learning. At the same time, it helps the primary user to concentrate on learning core knowledge.

[0125] As shown in Figure 4 , for an intermediate user, the learning graph is updated by improving the association relationship between the knowledge nodes of the first updated learning graph. This can help the intermediate user better understand the systematicness and logicalness of knowledge, and promote the mastery of knowledge. According to the learning needs and knowledge system of the intermediate user, the knowledge nodes whose association relationship needs to be enhanced are determined, including the knowledge nodes that need to be mastered in the intermediate learning stage and the knowledge nodes that are closely related to other important knowledge points. By shortening the length of the connection edge to improve the weight of the connection edge, the association relationship between the selected knowledge nodes is enhanced. After the association relationship is enhanced, the updated learning graph of the intermediate user is obtained. By enhancing the association relationship between the knowledge nodes, the intermediate user can deepen the understanding of knowledge and improve the application ability of knowledge, promote the perfection and systematization of the knowledge system of the intermediate user, and cultivate their logical thinking ability.

[0126] As shown in Figure 5As shown, for advanced users, by expanding knowledge nodes in the first updated learning graph, the learning graph is updated to meet the learning needs of advanced users and further improve their learning ability and knowledge level. According to the learning goals and subject fields of advanced users, the direction of knowledge node expansion is determined, for example, in the computer science learning graph, knowledge nodes are expanded to the frontier fields such as artificial intelligence, big data and cloud computing. In the first updated learning graph, knowledge nodes related to the expansion direction are added, including deeper professional knowledge, the latest research results, cross-disciplinary knowledge integration, etc. Each expanded knowledge node is assigned corresponding node attributes, including node name, field, difficulty level, etc. The newly added expanded knowledge nodes are connected with the related knowledge nodes in the original learning graph, the logical relationship between the knowledge nodes is analyzed, the corresponding connection relationship is determined and connected, and after the knowledge node expansion and connection relationship establishment, the updated learning graph of the advanced user is obtained. By expanding the knowledge nodes, the advanced users' demand for in-depth exploration of knowledge can be met, and the advanced users' knowledge can be broadened to cultivate their cross-disciplinary thinking ability and comprehensive quality.

[0127] Further, according to the updated learning graph, the difficulty of the learning content is adjusted in real time to obtain updated learning content, so as to realize intelligent adjustment of learning difficulty, including:

[0128] S801, according to the updated learning graph, analyzing the knowledge mastery degree of the user for each knowledge node to obtain an updated knowledge mastery degree;

[0129] S802, based on the updated knowledge mastery degree, matching in the updated learning graph to obtain an updated node path;

[0130] S803, according to the updated node path, matching in a preset question template library to obtain an updated template;

[0131] S804, based on the updated template, generating corresponding updated learning content through a preset content updating model to realize intelligent adjustment of learning difficulty.

[0132] In the embodiment, according to the updated learning graph, the knowledge mastery degree of the user is analyzed for each knowledge node, the learning state in the learning process of the user is updated, the updated knowledge mastery degree is obtained, the learning data corresponding to each knowledge node is extracted from the updated learning graph, the corresponding knowledge mastery degree score is calculated according to the learning data, and the updated knowledge mastery degree of each knowledge node is obtained; according to the calculated updated knowledge mastery degree, the knowledge nodes that need to be focused on or further learned are filtered out in the updated learning graph, and the learning path is determined according to the association relationship between the nodes, and the updated node path is obtained; by filtering out the updated node path, the user can be helped to learn weak knowledge in a targeted manner, gradually improve the knowledge system, and improve the learning effect.

[0133] Specifically, according to the updated node path, matching is performed in the preset question template library, the corresponding question template is filtered out in the preset question template library by analyzing the knowledge nodes in the updated node path and the learning order and emphasis of the knowledge, and the filtered question templates are combined according to the order of the updated node path to obtain an updated template; by filtering out the question combination suitable for the learning path of the user in the question template library, it can be ensured that the learning content is closely matched with the current learning needs and knowledge mastery degree of the user, and the learning is targeted and effective.

[0134] Specifically, according to the filtered updated template, the corresponding learning content is updated through a preset content updating model; according to the learning ability evaluation result and the updated knowledge mastery degree of the user, related parameters are set for the content updating model, the updated template is input into the preset content updating model, the model processes the question template according to the set parameters and built-in rules, including modifying, supplementing the question, adjusting the question order, adding problem solving ideas, knowledge point explanation and other auxiliary contents, after the content updating model processing, the complete updated learning content is output, the intelligent adjustment of the learning difficulty is realized, and the current learning needs of the user are met; by updating the learning content, highly personalized learning content can be generated, the learning ability and knowledge mastery degree of the user are accurately matched, the difficulty of the learning content is intelligently adjusted, the learning interest of the user can be stimulated, the learning motivation caused by improper difficulty can be avoided, and the learning effect can be effectively improved.

[0135] Further, based on the updated knowledge mastery degree, matching is performed in the updated learning graph to obtain an updated node path, including:

[0136] S901、based on the updated knowledge mastery degree, extracting nodes with a knowledge mastery degree change greater than a preset threshold from the updated learning graph to obtain a plurality of to-be-updated knowledge nodes;

[0137] S902, connecting the plurality of knowledge nodes to be updated by analyzing the association relationship between the knowledge nodes to obtain an updated node path.

[0138] In this embodiment, a threshold of a knowledge mastery degree change amount is set in advance according to the characteristics of the learning content and the learning goal, for example, the threshold is set to 15% for basic knowledge points, and the threshold is set to 10% for difficult knowledge points. The current updated knowledge mastery degree is compared with the previously recorded knowledge mastery degree, and the mastery degree change amount of each knowledge node is calculated, for example, the mastery degree of a certain knowledge node is 60% last time, and the updated mastery degree is 75% currently, and the change amount is 15%. All knowledge nodes in the updated learning graph are traversed, and the calculated mastery degree change amount is compared with the preset threshold. When the change amount is greater than the preset threshold, the knowledge node is extracted as a knowledge node to be updated. For example, the preset threshold is 10%, and the calculated mastery degree change amount of the knowledge node is 15%. Therefore, the knowledge node is included in the set of knowledge nodes to be updated. By quickly screening the knowledge nodes with significant changes in the user's learning process, the learning planning efficiency is improved.

[0139] Specifically, according to the screened knowledge nodes to be updated, for each knowledge node to be updated, the association relationship between the knowledge node and other knowledge nodes to be updated in the updated learning graph is found. According to the logical order of knowledge and the learning rule, the starting node of the path is determined from the knowledge nodes to be updated, and the basic knowledge node is selected as the starting node. According to the association relationship between the knowledge nodes, the other knowledge nodes to be updated are sequentially connected from the starting node to form an updated node path. In the connection process, the coherence and logicality of the path are ensured. When there are multiple knowledge nodes to be updated and the association between them is relatively complex, multiple paths can be constructed to meet different learning needs, for example, starting from the "quadratic equation" node, connecting to the "quadratic function" node, and then connecting to the "utilizing quadratic function to solve practical problems" node to form a complete updated node path. By connecting the knowledge nodes, the user can systematically learn the knowledge, avoid knowledge fragmentation, and improve the learning effect.

[0140] Embodiment 2

[0141] In this embodiment, as Figure 6 , a learning difficulty intelligent adjustment device is provided for implementing the learning difficulty intelligent adjustment method, comprising:

[0142] A learning graph construction module extracts key knowledge elements from pre-acquired user learning data to construct a learning graph.

[0143] a user evaluation module configured to analyze the knowledge mastery level and learning ability of the user according to the learning graph, and obtain a user ability evaluation result;

[0144] a learning content generation module configured to generate corresponding learning content in combination with the user ability evaluation result and the node association relationship in the learning graph, wherein the nodes include knowledge nodes and ability nodes;

[0145] a learning graph updating module configured to analyze the updating situation of the knowledge elements and the change situation of the user learning ability in the user learning process, and dynamically update the learning graph to obtain an updated learning graph;

[0146] a learning difficulty adjustment module configured to adjust the difficulty of the learning content in real time according to the updated learning graph to obtain updated learning content, so as to realize intelligent adjustment of the learning difficulty.

[0147] In this embodiment, the learning graph construction module is responsible for extracting key knowledge elements from the pre-acquired user learning data and constructing a learning graph. Through the mining and arrangement of the learning data, the knowledge is presented in the form of a graph, and the relationship between various knowledge elements is clearly displayed, which facilitates the system to comprehensively understand the knowledge system and the learning content of the user. The user evaluation module analyzes the knowledge mastery level and learning ability of the user based on the learning graph, and obtains a user ability evaluation result. Through the analysis of the performance of the user at each knowledge node and ability node in the learning graph, including the understanding, mastery and application ability of the knowledge points, and the comprehensive performance in different learning tasks, the level and ability characteristics of the user in the knowledge learning are evaluated.

[0148] Specifically, the learning content generation module generates learning content suitable for the user by combining the user's ability assessment results and the association relationship between the knowledge nodes and the ability nodes in the learning graph, selects the corresponding difficulty and correlation degree of the knowledge nodes from the learning graph according to the user's ability level, and organizes them into targeted learning materials, including exercises, explanation documents, video tutorials, etc., to ensure that the generated learning content not only meets the user's existing knowledge level, but also helps the user gradually improve their ability, effectively accumulate knowledge, and gradually improve their ability. The learning graph update module continuously monitors the update of knowledge elements and the changes in the user's learning ability during the user's learning process, and dynamically updates the learning graph. When the user learns new knowledge, masters new skills, or deepens their understanding of existing knowledge, the nodes and connection relationships in the learning graph are adjusted in a timely manner to adapt to the user's learning progress and changes in the knowledge structure. By continuously updating the learning graph, the system can always analyze and make decisions based on the user's latest state, ensuring the adaptability and effectiveness of the learning content. The learning difficulty adjustment module adjusts the difficulty of the learning content in real time according to the updated learning graph, generates updated learning content, and realizes intelligent adjustment of learning difficulty. By analyzing the user's mastery of each knowledge node in the updated learning graph, the user's learning progress and ability improvement are determined, appropriate question templates are selected from the preset question template library, and content that meets the user's current learning difficulty is generated through the content update model to realize intelligent adjustment of learning content difficulty and improve learning effectiveness and efficiency.

[0149] The basic principles and main features of the present application are shown and described above, and the advantages of the present application are also shown and described above. Those skilled in the art should understand that the present application is not limited by the above examples, and the above examples and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A learning difficulty intelligent adjustment method, characterized in that, The method comprises the following steps: extracting key knowledge elements from pre-acquired user learning data and constructing a learning graph; analyzing the knowledge mastery and learning ability of the user according to the learning graph to obtain a user ability evaluation result; generating corresponding learning content in combination with the user ability evaluation result and the node association relationship in the learning graph, wherein the nodes include knowledge nodes and ability nodes; analyzing the update situation of the knowledge elements and the change situation of the user learning ability in the user learning process to obtain an updated learning state; when a new knowledge point is added, adding a corresponding knowledge node in the learning graph, and establishing a connection relationship between the added knowledge point and the existing knowledge nodes by analyzing the discipline to which the added knowledge point belongs to obtain a first updated node; when the user's knowledge mastery improves, the association strength with related knowledge nodes is enhanced, the corresponding knowledge node attributes and connection relationships are updated to obtain a second updated node; according to the second updated node, the corresponding ability node attributes and connection relationships are updated to obtain a third updated node; combining the first updated node, the second updated node and the third updated node, updating the nodes and node connection relationships in the learning graph to obtain a first updated learning graph; according to the learning ability of the user, the user is classified through a preset learning ability evaluation model to obtain multi-level users, wherein the multi-level users include primary users, intermediate users and advanced users; for the primary users, the first updated learning graph is simplified in structure by reducing the number of knowledge nodes and retaining core knowledge points to obtain an updated learning graph for the primary users; for the intermediate users, the node association relationship between the knowledge nodes in the first updated learning graph is improved, the length of the connection edge between the knowledge nodes whose association relationship needs to be enhanced is shortened, and the weight of the connection edge is improved to obtain an updated learning graph for the intermediate users; for the advanced users, the knowledge nodes in the first updated learning graph are expanded to obtain an updated learning graph for the advanced users; according to the updated learning graph, the difficulty of the learning content is adjusted in real time to obtain updated learning content, so as to realize intelligent adjustment of learning difficulty.

2. The learning difficulty intelligent adjustment method according to claim 1, characterized in that, The method comprises the following steps: extracting corresponding knowledge points from the pre-acquired user learning data to obtain key knowledge elements; connecting the corresponding knowledge nodes by analyzing the association relationship between the key knowledge elements to obtain an initial learning graph; fusing the user learning data and mapping it into the initial learning graph to obtain a corresponding learning graph.

3. The learning difficulty intelligent adjustment method according to claim 1, characterized in that, According to the learning graph, the knowledge mastery and learning ability of the user are analyzed to obtain a user ability evaluation result, which comprises the following steps: analyzing the answering situation of the user on each knowledge node according to the association relationship between the knowledge nodes in the learning graph to obtain the knowledge mastery of the user; aggregating the knowledge nodes reflecting different discipline knowledge points in the learning graph to obtain multi-discipline nodes; For each type of subject node, the knowledge mastery of the user is evaluated respectively to obtain the knowledge mastery of each type of subject; According to the knowledge mastery of each type of subject, a corresponding ability node is added to each type of subject node and connected; During the user learning process, the attributes of the ability node are dynamically updated according to the change of the knowledge mastery of each type of subject of the user, to obtain an updated ability node; In combination with the corresponding updated ability node of each type of subject, the learning ability of the user is evaluated to obtain a user ability evaluation result.

4. The learning difficulty intelligent adjustment method according to claim 1, characterized in that, In combination with the user ability evaluation result and the node association relationship in the learning graph, corresponding learning content is generated, including: According to the user ability evaluation result, a target template library of corresponding difficulty is matched in a preset question template library; By analyzing the node association relationship in the learning graph, a corresponding weight is assigned to each node; In combination with the target template library and the weight, corresponding learning content is generated through a preset content generation model.

5. The learning difficulty intelligent adjustment method according to claim 1, characterized in that, According to the updated learning graph, the difficulty of the learning content is adjusted in real time to obtain updated learning content, so as to realize intelligent adjustment of learning difficulty, including: According to the updated learning graph, the knowledge mastery of the user for each knowledge node is analyzed to obtain an updated knowledge mastery; Based on the updated knowledge mastery, matching is performed in the updated learning graph to obtain an updated node path; According to the updated node path, matching is performed in a preset question template library to obtain an updated template; Based on the updated template, corresponding updated learning content is generated through a preset content update model, so as to realize intelligent adjustment of learning difficulty.

6. The learning difficulty intelligent adjustment method according to claim 5, characterized in that, Based on the updated knowledge mastery, matching is performed in the updated learning graph to obtain an updated node path, including: Based on the updated knowledge mastery, nodes with a knowledge mastery change greater than a preset threshold are extracted in the updated learning graph to obtain a plurality of to-be-updated knowledge nodes; By analyzing the association relationship between knowledge nodes, the plurality of to-be-updated knowledge nodes are connected to obtain an updated node path.

7. A learning difficulty intelligent adjustment device, characterized in that, For implementing the learning difficulty intelligent adjustment method according to any one of claims 1-6, including: A learning graph construction module extracts key knowledge elements from pre-acquired user learning data to construct a learning graph; A user evaluation module analyzes the knowledge mastery and learning ability of the user according to the learning graph to obtain a user ability evaluation result; A learning content generation module generates corresponding learning content in combination with the user ability evaluation result and the node association relationship in the learning graph, wherein the nodes include knowledge nodes and ability nodes; A learning graph update module analyzes the update of knowledge elements and the change of user learning ability during the user learning process, dynamically updates the learning graph to obtain an updated learning graph; A learning difficulty adjustment module adjusts the difficulty of the learning content in real time according to the updated learning graph to obtain updated learning content, so as to realize intelligent adjustment of learning difficulty.

Images