A dynamic optimization method and system for interactive learning activities

Abstract

The present application relates to the technical field of learning activity optimization, and discloses a method and system for dynamic optimization of interactive learning activities, including: obtaining user learning data, extracting user interest tags and learning ability tags through a preset user behavior analysis model, and generating user status tags; based on the user status tags, in combination with a preset course knowledge base, generating a corresponding activity generation path; according to the activity generation path, matching corresponding learning activities to obtain an initial learning activity; in combination with the user's real-time learning status, optimizing the activity generation path by adding path branches, pruning path branches, or replacing path nodes to obtain an activity optimization path; optimizing the initial learning activity through the activity optimization path to obtain an optimized learning activity, so as to dynamically optimize the interactive learning activity; the present application can dynamically optimize the learning activity and respond in real time to the status changes during the user's learning process.

Description

Technical Field

[0001] The present application relates to the technical field of learning activity optimization, and in particular to a method and system for dynamic optimization of interactive learning activities. Background Art

[0002] Currently, interactive learning activities have become an important learning model. However, existing technologies have many shortcomings in the design and optimization of learning activities. They cannot adjust and optimize learning activities in real time according to different users, and the designed learning activities are less interesting, resulting in poor learning effects.

[0003] The existing technology has the following problems: it adopts fixed course content and learning paths, and does not fully consider the differences in interests and learning abilities among individual users, resulting in poor learning effects; the learning activity optimization mechanism is relatively simple and cannot dynamically optimize learning activities in real time and comprehensively based on the user's real-time learning status; there is a lack of comprehensive consideration of knowledge relevance and activity difficulty, resulting in a lack of coherence and progressiveness in the learning process; in order to solve at least one of the above problems, the present invention proposes a method and system for dynamic optimization of interactive learning activities. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the main purpose of the present invention is to provide a method and system for dynamically optimizing interactive learning activities, which can effectively solve the problems in the background art. The specific technical solutions of the present invention are as follows:

[0005] A method for dynamic optimization of interactive learning activities, comprising:

[0006] Obtain the user's learning data, extract the user's interest tags and learning ability tags through the preset user behavior analysis model, and generate user status tags;

[0007] Based on the user status tag and in combination with a preset course knowledge base, a corresponding activity generation path is generated;

[0008] Generate a path based on the activity, match the corresponding learning activity, and obtain the initial learning activity;

[0009] In combination with the user's real-time learning status, the activity generation path is optimized by adding path branches, pruning path branches, or replacing path nodes to obtain an activity optimization path;

[0010] The initial learning activity is optimized through the activity optimization path to obtain an optimized learning activity, so as to dynamically optimize the interactive learning activity.

[0011] Specifically, the method of acquiring the user's learning data, extracting the user's interest tags and learning ability tags through a preset user behavior analysis model, and generating the user status tags includes:

[0012] Obtain user learning data by collecting user behavior logs, learning time, and interaction records;

[0013] Based on the learning data, extract the user's interest tags and learning ability tags through a preset user behavior analysis model;

[0014] The interest tag and the learning ability tag are integrated to generate a user status tag.

[0015] Specifically, the generating of a corresponding activity generation path based on the user status tag and in combination with a preset course knowledge base includes:

[0016] Extract the user's interest characteristics and ability characteristics based on the user status label and construct the user status vector;

[0017] Based on the user state vector, a match is performed in a preset course knowledge base to generate an activity generation path.

[0018] Specifically, the matching is performed in a preset course knowledge base based on the user state vector to generate an activity generation path, including:

[0019] Based on the preset course knowledge base, knowledge activities are used as activity nodes, and associations are established between activity nodes based on knowledge relevance and activity difficulty to construct a knowledge graph;

[0020] According to the user state vector, corresponding learning activities are matched in the knowledge graph to obtain a set of candidate paths;

[0021] By calculating the matching value between each path in the candidate path set and the user state vector, the path with the highest matching value is selected from the candidate path set as the activity generation path.

[0022] Specifically, generating a path based on the activity, matching corresponding learning activities, and obtaining initial learning activities include:

[0023] According to the activity generation path, a corresponding first learning activity is matched for each activity node in a preset activity template library;

[0024] Optimizing the first learning activity based on the user state vector to obtain a second learning activity;

[0025] The second learning activities are combined according to the order of activity nodes in the activity generation path to obtain an initial learning activity.

[0026] Specifically, the activity generation path is optimized by adding path branches, pruning path branches, or replacing path nodes in combination with the user's real-time learning status to obtain the activity optimization path, including:

[0027] Based on the user's real-time learning status, the path optimization area is identified by analyzing the activity node density and dependency relationships in the activity generation path;

[0028] Optimizing the path optimization area by adding path branches, pruning path branches, or replacing path nodes to obtain a first optimized path;

[0029] The activity generation path is optimized according to the first optimization path to obtain an activity optimization path.

[0030] Specifically, optimizing the path optimization area by adding path branches, pruning path branches, or replacing path nodes to obtain a first optimized path includes:

[0031] Analyze users' real-time interests and learning activity completion rates based on their real-time learning status;

[0032] According to the real-time interest, a path branch corresponding to the activity is added to the path optimization area to obtain a first optimization area;

[0033] When the completion rate of the learning activity is lower than a preset completion rate threshold, the path branch of the corresponding activity is pruned to obtain a second optimized area;

[0034] By analyzing the matching degree between the real-time learning state and the knowledge graph, the path nodes whose matching degree is less than a preset matching degree threshold are replaced to obtain a third optimized region;

[0035] The first optimization region, the second optimization region, and the third optimization region are combined to obtain a first optimization path.

[0036] Specifically, the initial learning activity is optimized through the activity optimization path to obtain an optimized learning activity, so as to dynamically optimize the interactive learning activity, including:

[0037] According to the activity optimization path, the initial learning activity is optimized in layers to obtain the corresponding first optimized activity;

[0038] The first optimization activities are combined to obtain optimized learning activities, so as to dynamically optimize the interactive learning activities.

[0039] Specifically, the initial learning activity is hierarchically optimized according to the activity optimization path to obtain the corresponding first optimization activity, including:

[0040] According to the optimized activity nodes in the activity optimization path, the corresponding learning activities are matched in the preset activity template library to obtain the optimized activities;

[0041] The optimized activity replaces the corresponding learning activity in the initial learning activity, and the order and duration of the learning activities are adjusted to obtain a first optimized activity.

[0042] A dynamic optimization system for interactive learning activities, used to implement the dynamic optimization method for interactive learning activities, comprising:

[0043] The user status analysis module obtains the user's learning data, extracts the user's interest tags and learning ability tags through the preset user behavior analysis model, and generates user status tags;

[0044] An activity generation path formulation module generates a corresponding activity generation path based on the user status tag and in combination with a preset course knowledge base;

[0045] A learning activity formulation module generates a path based on the activity, matches the corresponding learning activity, and obtains the initial learning activity;

[0046] An activity generation path optimization module optimizes the activity generation path by adding path branches, pruning path branches, or replacing path nodes in combination with the user's real-time learning status to obtain an activity optimization path;

[0047] The learning activity optimization module optimizes the initial learning activity through the activity optimization path to obtain an optimized learning activity, so as to dynamically optimize the interactive learning activity.

[0048] Compared with the prior art, this application has the following beneficial effects:

[0049] This application analyzes the user status based on the user's multi-dimensional learning data, matches the corresponding activity generation path in combination with the knowledge graph, optimizes the activity generation path according to the user's real-time learning status, and performs hierarchical optimization of the initial learning activities based on the optimized path to achieve dynamic adjustment of learning activities, which can respond to changes in the user's learning process in real time, ensure that the optimized learning activities are more in line with user needs and learning rhythm, and improve the user's learning efficiency and learning effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a workflow diagram of a dynamic optimization method for interactive learning activities in Example 1 of the present invention;

[0051] Figure 2 A schematic diagram of the activity generation path matching process in Example 1 of the present invention;

[0052] Figure 3Schematic diagram of the path optimization area optimization process in Example 1 of the present invention;

[0053] Figure 4 This is a structural diagram of a dynamic optimization system for interactive learning activities in Example 2 of the present invention. DETAILED DESCRIPTION

[0054] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0055] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0056] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0057] Example 1:

[0058] This embodiment provides a method for dynamic optimization of interactive learning activities, such as Figure 1 As shown, the method for dynamic optimization of interactive learning activities includes:

[0059] S101: Obtain user learning data, extract user interest tags and learning ability tags through a preset user behavior analysis model, and generate user status tags;

[0060] S102: Based on the user status tag and in combination with a preset course knowledge base, generate a corresponding activity generation path;

[0061] S103, generating a path based on the activity, matching corresponding learning activities, and obtaining an initial learning activity;

[0062] S104. Optimizing the activity generation path by adding path branches, pruning path branches, or replacing path nodes based on the user's real-time learning status to obtain an activity optimization path;

[0063] S105: Optimize the initial learning activity through the activity optimization path to obtain an optimized learning activity, so as to dynamically optimize the interactive learning activity.

[0064] This embodiment first analyzes the user's learning status based on the user's learning data, combines the user's learning status and the preset course knowledge base, generates a corresponding activity generation path, matches the corresponding learning activities, and formulates learning activities suitable for the user's status. During the user's learning process, the activity generation path and learning activities are optimized accordingly according to the changes in the user's real-time learning status. By dynamically optimizing the learning activities, the learning content and form are always matched with the user's real-time learning status. Compared with the traditional fixed learning activity model, this application can avoid users wasting time on learning content that is not suitable for them, improve learning efficiency, and help users master knowledge and skills faster.

[0065] In this embodiment, first, the learning data generated by the user during the learning process is obtained, and the learning data is analyzed by a preset user behavior analysis model to extract the user's interest tags and learning ability tags, and generate user status tags; the preset user behavior analysis model can be a machine learning model or a statistical model, and a large amount of historical data is used to train the user behavior analysis model to obtain a pre-trained user behavior analysis model. The user behavior analysis model analyzes the user's preferences for different courses and learning resources, extracts the knowledge areas that the user is interested in, and obtains corresponding interest tags; the user's learning ability is evaluated based on data such as the user's learning progress and knowledge mastery, and a learning ability tag is generated; the interest tag and the learning ability tag are integrated into a user status tag, which can reflect the user's learning characteristics and current status, provide a basis for formulating learning activities that meet user needs, and improve the matching degree between learning activities and users.

[0066] Specifically, based on the user status tags obtained through analysis, matching is performed in the preset course knowledge base to screen out courses, knowledge points and learning resources that match the user's interests and learning abilities, and combining them in a corresponding logical order to obtain an activity generation path; generating activity generation paths based on user status tags can customize corresponding learning routes for users, so that learning activities are in line with the user's interests and abilities, helping users to learn knowledge more efficiently, avoid learning content that is too difficult or too easy, improve the pertinence and effectiveness of learning, and better meet the user's personalized learning needs.

[0067] For example, for user A, his user status label is: interests include Python programming, data structure and algorithm; learning ability is strong programming learning ability; in the course knowledge base, he is matched with resources such as Python advanced programming course, data structure and algorithm advanced course, Python data analysis project case, etc.; according to the user's learning ability and the logical relationship between knowledge, the generated activity generation path is: Python advanced programming course → data structure and algorithm advanced course → Python data analysis project case.

[0068] Specifically, according to the generated activity generation path, the corresponding learning activities are matched to obtain the initial learning activities; based on the learning content and sequence determined by the activity generation path, specific learning activity forms are searched in the learning resource library, including video courses, online tests, group discussions, project practices, etc., and the matched learning activities are combined to obtain the corresponding initial learning activities; by matching the corresponding learning activities through the activity generation path to obtain the initial learning activities, the learning habits and needs of different users can be met, the fun and participation of learning can be improved, and at the same time, it can ensure that users can fully and deeply master knowledge and skills.

[0069] During the user's learning process, the user's learning status is analyzed in real time, and the activity generation path is optimized according to the user's real-time learning status. New learning resources or courses are introduced by adding path branches to supplement knowledge gaps, path branches are pruned to remove unnecessary or inappropriate learning content, and path nodes are replaced with more appropriate learning resources to replace the original resources. The activity generation path is optimized to obtain an activity optimization path. By optimizing the activity generation path, it can be considered that the user's learning status will change during the learning process due to difficulties in understanding a certain knowledge point, accelerated or slowed learning progress, etc., resulting in the original activity generation path no longer being suitable. By optimizing the activity generation path, it can respond to changes in the user's learning process in a timely manner, so that learning activities always match the user's actual needs and learning situation, avoid learning burnout or learning difficulties due to inappropriate learning content, improve learning efficiency and learning effects, and realize dynamic adaptive adjustment of learning activities.

[0070] Specifically, according to the optimized activity optimization path, the initial learning activities are adjusted and optimized accordingly, inappropriate learning activities are replaced, new learning activities are added, unnecessary learning activities are removed, and optimized learning activities are obtained. By optimizing the learning activities, the learning activities can more accurately meet the real-time needs of users. The optimized learning activities can better help users overcome learning difficulties, consolidate knowledge, improve skills, and achieve more efficient learning.

[0071] This application analyzes the user status based on the user's multi-dimensional learning data, matches the corresponding activity generation path in combination with the knowledge graph, optimizes the activity generation path according to the user's real-time learning status, and performs hierarchical optimization of the initial learning activities based on the optimized path to achieve dynamic adjustment of learning activities, which can respond to changes in the user's learning process in real time, ensure that the optimized learning activities are more in line with user needs and learning rhythm, and improve the user's learning efficiency and learning effect.

[0072] Furthermore, the acquisition of user learning data, extraction of user interest tags and learning ability tags through a preset user behavior analysis model, and generation of user status tags include:

[0073] S201, obtaining user learning data by collecting user behavior logs, learning time and interaction records;

[0074] S202: Extracting user interest tags and learning ability tags based on the learning data using a preset user behavior analysis model;

[0075] S203: Fusion the interest tag and the learning ability tag to generate a user status tag.

[0076] In this embodiment, first, the user's operation data is collected to obtain the user's behavior log, learning time and interaction record. The behavior log records the user's access, selection and operation behavior of learning resources; the learning time reflects the time cost of the user's investment in learning; the interaction record reflects the user's interaction with the learning content, other users or platform functions, and the collected data is combined to obtain the user's learning data; by collecting multi-dimensional learning data of the user, the user's learning behavior can be understood from different angles, avoiding analysis deviations caused by single data, and providing comprehensive data for analyzing the user's learning status.

[0077] For example, on an online language learning platform, user B opened the "Basics of English Grammar" course page three times in one day, staying for 10 minutes, 15 minutes, and 20 minutes each time; watched the "Tense Explanation" video, which was 30 minutes long and actually watched 28 minutes; and posted a question about the usage of "Present Perfect Tense" in the course discussion area and received two replies.

[0078] Specifically, based on the collected learning data, the user's interests and learning abilities are analyzed, and the user's interest tags and learning ability tags are extracted through a preset user behavior analysis model; the collected learning data is preprocessed by cleaning, standardization, and other preprocessing to obtain preprocessed learning data. The preset user behavior analysis model can be a random forest model, a statistical model, or other models.

[0079] For example, the user behavior analysis model in this embodiment uses a random forest model. This model is trained using a large amount of historical learning data and corresponding user status labels to produce a pre-trained user behavior analysis model. The model learns the mapping relationship between data and labels. When new user learning data is input, the model can analyze the user's preference for different learning content based on the learned patterns and extract interest tags. Simultaneously, the model evaluates the user's performance in learning tasks, such as test scores and learning progress, to determine the user's learning ability and generate a learning ability tag. By extracting interest tags and learning ability tags, the user's potential interests and actual learning ability can be accurately explored, providing a basis for developing corresponding learning activities.

[0080] Specifically, interest tags and learning ability tags are combined to generate user status tags. Interest tags and learning ability tags are arranged and combined, and combined with the user's learning preferences and concerns as well as the user's strengths and weaknesses in learning. After integration, a comprehensive and integrated user status description is obtained, which provides more comprehensive user information and can better meet the user's personalized needs when formulating learning activities and recommending learning resources, thereby improving the user's learning experience.

[0081] For example, the user's interest tag "Business English" and learning ability tag "strong English interpretation learning ability, weak English listening learning ability" are unified and integrated into the generated user status tag {Interest: Business English; Learning ability: strong English interpretation learning ability, weak English listening learning ability}.

[0082] Furthermore, the corresponding activity generation path is generated based on the user status tag and combined with the preset course knowledge base, including:

[0083] S301, extracting the user's interest characteristics and ability characteristics based on the user status tag and constructing a user status vector;

[0084] S302: Based on the user state vector, matching is performed in a preset course knowledge base to generate an activity generation path.

[0085] In this embodiment, the user's interest characteristics and ability characteristics are extracted according to the user status label, and the corresponding characteristics are mapped into the vector space to construct a user state vector; according to the interest label in the user status label, the interest keywords and the corresponding interest levels are extracted as interest characteristics. For example, if the user has a high interest in the field corresponding to the interest keyword, a higher value is assigned, such as 10; if the interest is average, a value of 5 is assigned; if there is no interest, a value of 0 is assigned; according to the learning ability label, the learning ability is quantified to obtain a quantitative indicator reflecting the strength of the user's learning ability; for example, when the learning ability label is "strong mathematics learning ability", a value is set, and in the range of 0-10, "strong" is assigned a value of 8; if it is "weak programming learning ability", "weak" is assigned a value of 3; the extracted interest characteristics and ability characteristics are combined to construct a user state vector. By constructing the user state vector, courses and learning resources that match the user status can be quickly matched.

[0086] Specifically, based on the user state vector, matching is performed in the preset course knowledge base to search for course knowledge that is highly matched with the user's interests and abilities. According to the logical relationship between course knowledge, including the logical relationship and progressive relationship of knowledge points, the matched courses are arranged in a corresponding logical order, and an activity generation path is generated to plan a learning route that meets the user's needs. By planning the activity generation path, direction is provided for the formulation of learning activities, and courses that are consistent with the user's interests and abilities are accurately found to avoid users learning content that they are not interested in or that exceeds their abilities. By reasonably planning the course sequence, the learning process can be ensured to be coherent and systematic, helping users to achieve their learning goals more efficiently and improving the user experience.

[0087] Furthermore, the matching is performed in a preset course knowledge base based on the user state vector to generate an activity generation path, including:

[0088] S401. Based on the preset course knowledge base, knowledge activities are used as activity nodes, and associations are established between activity nodes according to knowledge relevance and activity difficulty to construct a knowledge graph;

[0089] S402: Match corresponding learning activities in the knowledge graph according to the user state vector to obtain a set of candidate paths;

[0090] S403 : Calculate the matching value between each path in the candidate path set and the user state vector, and select the path with the highest matching value in the candidate path set as the activity generation path.

[0091] In this embodiment, first, based on the preset course knowledge base, knowledge activities are taken as activity nodes, and the internal associations between knowledge and the progressive relationship of activity difficulty and other associations are taken as edges between activity nodes; by analyzing the logical relationships between the knowledge points involved in each knowledge activity in the course knowledge base, including sequence, inclusion relationship, parallel relationship, etc.; for example, in the mathematics course, "the concept of function" is the basis of "the image and properties of function", and there is a knowledge association between the two in sequence. According to the corresponding knowledge association, edges are established between the corresponding activity nodes, and the determined activity nodes and the established association edges are integrated to form a complete knowledge graph; by constructing a knowledge graph, the knowledge activities in the course knowledge base can be associated, and the connection between different learning activities and the direction of the learning path can be displayed, so as to quickly match the corresponding activity generation path in the knowledge graph.

[0092] like Figure 2According to the user state vector, matching is performed in the knowledge graph to screen out activity nodes that meet the user's interests and abilities. Based on the association relationship between nodes in the knowledge graph, starting from these matching nodes, corresponding learning paths are searched to form a set of candidate paths; according to the interest characteristics and ability characteristics in the user state vector, feature matching is performed separately. When the value corresponding to an interest keyword in the user state vector is high and the knowledge field of an activity node contains the keyword, the activity node is considered to match the user's interest; according to the user's ability quantification value, activity nodes with matching difficulty levels are screened out; combining the results of interest matching and ability matching, activity nodes that meet both the user's interests and abilities are matched; starting from the matching nodes, combined with the association edges between activity nodes in the knowledge graph, corresponding learning paths are searched to form a set of candidate paths; by screening out multiple learning paths that meet the user's personalized needs, the user's interests and abilities are fully considered, a learning route reference is provided, and the problem that a single fixed learning path cannot meet the needs of different users is avoided, thereby improving the flexibility and adaptability of learning path planning.

[0093] Specifically, after screening out the candidate path set, the degree of matching between each path in the candidate path set and the user state vector is analyzed, the corresponding matching value is calculated, and the path with the highest matching value is selected as the activity generation path; by calculating the matching value and selecting the corresponding activity generation path, the learning path that best suits the user's interests and abilities can be screened out, ensuring that the user can access learning content that is of interest and matches their own learning abilities during the learning process, thereby improving the enthusiasm and effectiveness of learning. At the same time, it also optimizes the efficiency of using learning resources and provides users with a better learning experience.

[0094] Exemplarily, for interest matching calculation, set the matching score calculation rules between the knowledge domain keywords of each active node and the interest keywords in the user state vector; for example, if the knowledge domain of the active node completely contains the user's interest keywords, the score is 1; if it partially contains, a score between 0-1 is assigned based on the degree of inclusion; if it does not contain, the score is 0. Add up the interest matching scores of all active nodes in the path to obtain the total interest matching score of the path. For ability matching calculation, set the matching score calculation rules between the difficulty level of the active node and the ability quantification value in the user state vector; for example, when the difficulty level of the active node completely matches the user's ability quantification value, the score is 1; if the difficulty is slightly higher or lower, a score between 0-1 is assigned based on the size of the gap; if the gap is too large, the score is 0. Add up the ability matching scores of all active nodes in the path to obtain the total ability matching score of the path; comprehensively consider the total interest matching score and the total ability matching score, and calculate the final matching value of each path with the user state vector by weighted summation, such as 60% of the total interest matching score and 40% of the total ability matching score.

[0095] Furthermore, generating a path based on the activity, matching corresponding learning activities, and obtaining initial learning activities includes:

[0096] S501: Generate a path based on the activity, and match a corresponding first learning activity to each activity node in a preset activity template library;

[0097] S502: Optimize the first learning activity based on the user state vector to obtain a second learning activity;

[0098] S503: Combine the second learning activities according to the order of activity nodes in the activity generation path to obtain an initial learning activity.

[0099] In this embodiment, according to the activity generation path, a corresponding learning activity is matched for each activity node in the preset activity template library to obtain a matching first learning activity; the preset activity template library stores multiple types of learning activity templates, each learning activity template corresponds to specific knowledge content, teaching objectives and activity forms, including learning activities such as video courses, online tests, and group discussions. The attributes of each learning activity template in the activity template library are analyzed, including information such as the knowledge field, teaching objectives, activity form, and applicable difficulty level targeted by the template. The attributes of the activity node are compared with the attributes of the activity template, and the knowledge field is matched first to ensure that the knowledge involved in the template is consistent with the requirements of the activity node; then the learning objectives are matched to analyze whether the activity template can meet the learning objectives set for the activity node; finally, the difficulty level is considered to ensure that the difficulty of the template is adapted to the difficulty of the activity node; when the attributes of the activity template match the attributes of the activity node, the template is used as the first learning activity corresponding to the activity node, and the corresponding first learning activity is matched in the activity template library for each activity node in the activity generation path; by matching the first learning activity, specific learning activities can be quickly generated according to the activity generation path, thereby improving the efficiency of learning activity generation.

[0100] Specifically, after matching the first learning activity, the user state vector is compared and analyzed with the relevant attributes of the first learning activity, the activity content and form are adjusted according to the user's interest preferences, the activity difficulty and progress are adjusted according to the user's learning ability, and the first learning activity is optimized to obtain a second learning activity that better meets the user's needs; optimizing the first learning activity based on the user state vector can fully consider the different needs of different users, so that the learning activity can better meet the interests and ability needs of each user, and by adjusting the activity content, form, difficulty and progress, it can improve the user's participation and enthusiasm in the learning activity, reduce learning barriers caused by unsuitable learning activities, and help improve the user's learning effect and learning experience.

[0101] For example, when the content of the first learning activity involves a field of knowledge that the user is interested in, the depth or breadth of the relevant content can be increased, or the form of the activity can be adjusted to a form that the user prefers. For example, if the user is interested in "artificial intelligence" and the first learning activity is about Python programming, artificial intelligence-related programming cases can be added to the programming exercises; for users with stronger learning abilities, the difficulty of the activity can be increased, such as providing more complex exercises or expanded content; for users with weaker learning abilities, the difficulty can be reduced, complex tasks can be split into multiple simple subtasks, or the learning time can be extended.

[0102] Specifically, the optimized second learning activities are arranged and combined in sequence according to the order of learning activities in the activity generation path to obtain the initial learning activity. For example, if the activity generation path is "English vocabulary accumulation → English grammar learning → English essay writing", the corresponding second learning activities are "English high-frequency vocabulary memorization video course and word spelling practice", "English grammar intensive online course and grammar filling-in-the-blank practice", and "English essay writing skills explanation and theme writing task". These three second learning activities are combined according to the node order of the activity generation path to obtain the initial learning activity. By combining learning activities, users can follow the logical order of knowledge and learning rules during the learning process, avoid confusion and jumps in the learning process, and help improve learning efficiency and effectiveness.

[0103] Furthermore, the activity generation path is optimized by adding path branches, pruning path branches, or replacing path nodes in combination with the user's real-time learning status to obtain an activity optimization path, including:

[0104] S601, identifying a path optimization area by analyzing the density and dependency of activity nodes in an activity-generated path based on the user's real-time learning status;

[0105] S602: Optimize the path optimization area by adding path branches, pruning path branches, or replacing path nodes to obtain a first optimized path;

[0106] S603: Optimize the activity generation path according to the first optimization path to obtain an activity optimization path.

[0107] In this embodiment, the density and dependency of activity nodes in the activity generation path are analyzed based on the user's real-time learning status. A high node density indicates that the learning content is too compact and difficult for the user to digest; a low node density indicates that the learning progress is slow and the content is incoherent; the dependency relationship of activity nodes indicates the sequence and logical association of knowledge learning; by analyzing the density and dependency of activity nodes, the part of the current activity generation path that does not match the user's real-time learning status is identified, and the path optimization area is obtained; the user's real-time learning status is analyzed based on the collected real-time learning data of the user; the activity generation path is divided according to knowledge modules or learning stages , count the number of active nodes in each divided area, calculate the active node density, and by comparing the active node density of different areas, identify areas with abnormal density (too high or too low). By analyzing the sequence and logical dependencies between the individual active nodes in the activity generation path, identify cases where the dependencies are unreasonable. For example, when a user's test score on a certain knowledge module is poor and the node density in that activity area is too high, indicating that the user has difficulty understanding the overly compact learning content, this area can be used as a path optimization area. When a user can complete subsequent activity nodes by skipping a previous activity node, the relevant node area can also be used as a path optimization area. By identifying the path optimization area, it is possible to locate the part that does not match the user's real-time learning status, avoiding blind modifications to the entire path, and improving the targetedness and efficiency of path optimization.

[0108] like Figure 3 The path optimization region is optimized by adding path branches, pruning path branches, or replacing path nodes to obtain the first optimized path. Adding path branches introduces new active nodes or subpaths to the path optimization region to supplement the user's lack of knowledge or skills. Pruning path branches removes active nodes or subpaths that are unnecessary or inappropriate for the user's current learning state, reducing the learning burden and preventing users from wasting time on irrelevant content. Path node replacement replaces existing nodes with active nodes more suitable for the user's current learning state to improve learning outcomes. Based on the specific conditions of the path optimization region and the user's real-time learning status, an appropriate optimization method is selected to adjust it, resulting in the optimized first optimized path. By optimizing the path optimization region, the learning path can be flexibly adjusted according to the user's real-time learning status to meet the user's personalized learning needs, promptly address knowledge gaps, remove redundant content, and improve the relevance and effectiveness of learning activities.

[0109] Specifically, according to the first optimization path, the activity generation path is optimized, and the original path optimization area is replaced to obtain the activity optimization path. The replaced activity generation path is checked to see whether the overall logic of the path is reasonable and whether the order and association between the activity nodes are correct. The path is optimized, such as adjusting the transition instructions between activity nodes, supplementing necessary learning prompts, etc., to obtain a complete activity optimization path; the first optimization path is integrated into the activity generation path to obtain the activity optimization path, which ensures the integrity and coherence of the learning path optimization, can better adapt to the user's real-time learning status, improve the quality and practicality of the learning path, and help users complete learning tasks more smoothly and achieve learning goals.

[0110] Furthermore, the optimizing the path optimization area by adding path branches, pruning path branches, or replacing path nodes to obtain a first optimized path includes:

[0111] S701. Analyze the user's real-time interests and learning activity completion rate based on the user's real-time learning status;

[0112] S702: Add a path branch corresponding to the activity in the path optimization region according to the real-time interest to obtain a first optimization region;

[0113] S703: When the completion rate of the learning activity is lower than a preset completion rate threshold, pruning the path branch of the corresponding activity to obtain a second optimized region;

[0114] S704: Analyze the matching degree between the real-time learning state and the knowledge graph, and replace the path nodes whose matching degree is less than a preset matching degree threshold to obtain a third optimized region;

[0115] S705: Combine the first optimization area, the second optimization area, and the third optimization area to obtain a first optimization path.

[0116] In this embodiment, based on the user's real-time learning status, the user's current interests and the completion status of the learning activities are analyzed. By analyzing the real-time interests, the user's current focus and interests can be obtained, so as to supplement the user with learning content that is more in line with his or her interests; the learning activity completion rate reflects the progress and effect of the user in performing the learning task, and determines whether the learning activity is suitable for the user; by analyzing the user's real-time interests and learning activity completion rate, direction is provided for path optimization, thereby meeting the user's personalized learning needs and improving the fit between the learning activity and the user.

[0117] For example, when user F was studying the "Business English" course, the platform collected the following data: user F posted 3 posts about "Business English Negotiation Skills" in the discussion area and searched for relevant learning materials 5 times; in the "Business English Email Writing" learning activity, the user was required to complete 8 email writing exercises and watch 3 teaching videos, and the user completed 3 exercises and watched 1 video; through analysis, it was determined that user F's real-time interest was "Business English Negotiation Skills"; the completion rate of the "Business English Email Writing" learning activity was calculated to be (3 / 8×0.6+1 / 3×0.4)≈0.39, or 39%.

[0118] Specifically, according to the user's real-time interests, corresponding learning activities are matched, and activities suitable for the user's current learning stage and ability level are screened out. In the path optimization area, learning activity path branches corresponding to the real-time interests are inserted. According to the logical relationship of knowledge and the learning order, the connection method between the new branch and the original path node is determined to ensure the continuity and logic of the path. After completing the addition of path branches, the first optimized area containing the new learning activities is obtained; by adding path branches, it is possible to respond to changes in user interests, provide users with more interesting learning content, improve users' learning enthusiasm and participation, make the learning path more personalized and attractive, and enhance users' learning experience.

[0119] Specifically, when the completion rate of a learning activity is lower than a preset threshold, it means that the user encounters difficulties in performing the learning activity or is not interested in the activity. Continuing to retain the activity will affect the user's learning progress and enthusiasm. The corresponding path branches are pruned to remove learning content that is not suitable for the user, reduce the learning burden, and make the learning path more concise and efficient. By pruning path branches, the user's learning burden can be reduced, and the effectiveness and efficiency of the learning path can be improved.

[0120] Specifically, according to the matching degree between the user's real-time learning status and the knowledge graph, the path nodes with a matching degree less than a preset matching degree threshold are replaced to obtain the third optimized area. When the matching degree is less than the preset matching degree threshold, it means that the path node is not suitable for the user's current status. By replacing it with a more suitable node, the learning path is optimized to make it more adaptable to the user's learning needs. The corresponding matching degree is obtained by calculating the cosine similarity between the user's real-time learning status vector and each path node vector in the knowledge graph. For path nodes with a matching degree less than the preset matching degree threshold, the knowledge graph is searched for nodes with a higher matching degree with the user's real-time learning status and more suitable for the user's current learning stage and ability to replace them, and the association relationship between the nodes is updated. By analyzing the matching degree between the real-time learning status and the knowledge graph to replace the path nodes, it can be ensured that the content of the user's learning matches his or her actual situation, thereby improving the pertinence and effectiveness of learning.

[0121] The paths in the first, second, and third optimized areas are merged, duplication removed, and valid learning activity nodes and path branches retained. The integrated paths are then sorted out, the order and logical connections between activity nodes checked for rationality, and the connections between nodes adjusted to obtain the final first optimized path. By optimizing the paths in various ways, multiple factors, including user interests, learning outcomes, and knowledge system compatibility, are fully considered, providing users with learning paths that better meet their needs, helping to improve learning efficiency and enhance learning outcomes, ultimately enhancing their learning experience and outcomes.

[0122] Furthermore, the initial learning activity is optimized through the activity optimization path to obtain an optimized learning activity, so as to dynamically optimize the interactive learning activity, including:

[0123] S801. Perform hierarchical optimization on the initial learning activity according to the activity optimization path to obtain a corresponding first optimized activity;

[0124] S802: Combine the first optimization activities to obtain optimized learning activities, so as to dynamically optimize the interactive learning activities.

[0125] In this embodiment, the initial learning activities are optimized in layers according to the activity optimization path, and the initial learning activities are optimized according to different dimensions. According to the adjustment requirements of each part in the activity optimization path, the learning activities at different levels are optimized separately, so that each learning activity can better fit the user's current learning needs and status, thereby obtaining the first optimized activity; by performing layered optimization on the initial learning activities, the optimization can be made more detailed and comprehensive, meeting the user's learning needs, avoiding the problem of inaccurate optimization caused by general adjustments to learning activities, improving the optimization effect and efficiency, and helping to improve the quality of learning activities.

[0126] Specifically, the first optimization activities are combined according to the corresponding logical order and learning rules to form a complete and coherent sequence of learning activities to obtain optimized learning activities; the combination order of the first optimization activities is determined according to the order of activity nodes in the activity optimization path and the logical relationship between learning activities (including the sequence of knowledge, the progressive relationship of skill training, etc.); for example, in programming learning, learning activities related to programming language basics should be carried out first, and then learning activities related to data structures and algorithms; the learning content in the first optimization activity is integrated in sequence, and it is checked whether the transition between adjacent activities is natural and whether there is duplication or omission of content. The repeated content is streamlined and the omitted key knowledge points or skill training are supplemented to ensure the coherence and completeness of the content of the learning activities; the connection between the learning activities is optimized to obtain optimized learning activities; by combining the optimized activities, the systematicness and coherence of the learning activities are guaranteed, which helps users better understand and master knowledge and skills, can better meet users' real-time changing learning needs, improve learning effects and users' learning experience, and enhance the guiding and promoting role of learning activities on users' learning.

[0127] Furthermore, the initial learning activity is hierarchically optimized according to the activity optimization path to obtain the corresponding first optimized activity, including:

[0128] S901. According to the optimized activity nodes in the activity optimization path, the corresponding learning activities are matched in the preset activity template library to obtain the optimized activities;

[0129] S902: Replace the corresponding learning activity in the initial learning activity with the optimized activity, and adjust the order and duration of the learning activities to obtain a first optimized activity.

[0130] In this embodiment, according to the optimized activity nodes in the activity optimization path, activity matching is performed in the preset activity template library, and the corresponding learning activities are matched to obtain optimized activities; the knowledge field, learning objectives, difficulty level and other attributes of each optimized activity node in the activity optimization path are analyzed to obtain the attributes of the optimized activity node; the attributes of the optimized activity node in the activity optimization path are compared with the attributes of the template in the activity template library, and the knowledge field is matched first to ensure that the knowledge involved in the template is consistent with the node requirements; then the learning objectives are matched to analyze whether the template can meet the learning objectives set by the node; then the difficulty level is matched to ensure that the difficulty of the template is adapted to the difficulty of the node; finally, the matching activity template is selected considering factors such as the activity form and learning time; the corresponding optimized activity is matched for each optimized activity node in the activity optimization path; by matching the optimized activities, learning activities that meet the changes in the user's current learning needs can be found quickly and efficiently to ensure a high degree of match between the optimized activities and user needs.

[0131] Specifically, the matched optimized activity is used to replace the corresponding learning activity in the initial learning activity, and the order and duration of the learning activities are adjusted to obtain the first optimized activity; in the initial learning activity, the learning activity corresponding to the optimized activity node in the activity optimization path is found, and the corresponding optimized activity replaces the corresponding original activity in the initial learning activity, and the order of the replaced learning activities is adjusted according to the order of each activity node in the activity optimization path to ensure that the entire learning activity sequence conforms to the new learning path logic; the duration of the learning activity is adjusted in combination with the learning duration of the optimized activity, the overall learning plan and the user's learning time schedule, and the start time and duration of the subsequent learning activities are adjusted accordingly to ensure that the time schedule of the entire learning activity sequence is reasonable and compact; after activity replacement, sequence adjustment and duration adjustment, the complete first optimized activity is obtained; by replacing the initial learning activity with the optimized activity and adjusting the order and duration, the learning activity can quickly adapt to changes in the user's learning status, and the learning content can be updated in time, ensuring the logic and smoothness of the learning process, meeting the user's current learning needs, and improving the user's learning experience and learning effect.

[0132] For example, the original sequence of learning activities is "Python Basic Syntax Learning (3 hours) → Python Web Development Fundamentals (5 hours) → Python Data Processing (4 hours)." After replacing the "Python Web Development Fundamentals" learning activity with the "Django Advanced Framework Application Practice Course," the sequence of activities optimized is adjusted to "Python Basic Syntax Learning (3 hours) → Django Advanced Framework Application Practice Course (8 hours) → Python Data Processing (4 hours)." Furthermore, considering the overall learning progress and the user's daily learning time, the duration of "Python Basic Syntax Learning" is adjusted to 2 hours, and the duration of "Python Data Processing" is adjusted to 3 hours, resulting in the first optimized activity: "Python Basic Syntax Learning (2 hours) → Django Advanced Framework Application Practice Course (8 hours) → Python Data Processing (3 hours)."

[0133] Example 2:

[0134] In this embodiment, if Figure 4 , providing an interactive learning activity dynamic optimization system for implementing the interactive learning activity dynamic optimization method, comprising:

[0135] The user status analysis module obtains the user's learning data, extracts the user's interest tags and learning ability tags through the preset user behavior analysis model, and generates user status tags;

[0136] An activity generation path formulation module generates a corresponding activity generation path based on the user status tag and in combination with a preset course knowledge base;

[0137] A learning activity formulation module generates a path based on the activity, matches the corresponding learning activity, and obtains the initial learning activity;

[0138] An activity generation path optimization module optimizes the activity generation path by adding path branches, pruning path branches, or replacing path nodes in combination with the user's real-time learning status to obtain an activity optimization path;

[0139] The learning activity optimization module optimizes the initial learning activity through the activity optimization path to obtain an optimized learning activity, so as to dynamically optimize the interactive learning activity.

[0140] In this embodiment, the user status analysis module collects multi-dimensional learning data such as user behavior logs, learning time, interaction records, etc., and uses a preset user behavior analysis model to extract user interest tags and learning ability tags, and integrates the two to generate user status tags that comprehensively reflect user interests and abilities, providing accurate user status basis for formulating and optimizing learning activities; the activity generation path formulation module extracts user interest and ability characteristics based on the user status tags generated by the user status analysis module and converts them into vectors. In the preset course knowledge base, by constructing a knowledge graph, it analyzes the correlation and difficulty between knowledge activities, calculates the similarity between the user state vector and the course vector, matches the corresponding courses and resources, arranges them in the logical order of knowledge, and generates personalized activity generation paths to provide direction for the formulation of learning activities.

[0141] Specifically, the learning activity formulation module matches the first learning activity for each node in the preset activity template library according to the activity generation path, and then optimizes the activity content, form, difficulty and progress from the perspective of interest and ability based on the user state vector to obtain the second learning activity. Finally, the initial learning activity is formed by combining the order of path nodes to generate specific executable learning task activities; the activity generation path optimization module monitors the user's learning process data in real time, identifies the areas that need to be optimized by analyzing the node density and dependency relationship of the activity generation path, and dynamically adjusts the path by adding path branches that meet the user's real-time interests, pruning inefficient branches with low completion rates, and replacing path nodes with low matching with the user status, etc., to ensure that the learning path always fits the user's actual learning situation and improve the pertinence and effectiveness of learning; the learning activity optimization module uses the activity optimization path as the basis to perform hierarchical optimization on the initial learning activity to obtain the first optimized activity, and then integrates the content and adjusts the connection according to the path node order and the logical relationship of the activities, and combines the first optimized activity into the optimized learning activity, realizing dynamic optimization of interactive learning activities and bringing users a better learning experience.

[0142] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for dynamic optimization of interactive learning activities, characterized in that: include: Obtain the user's learning data, extract the user's interest tags and learning ability tags through the preset user behavior analysis model, and generate user status tags; Based on the user status tag and in combination with a preset course knowledge base, a corresponding activity generation path is generated; Generate a path based on the activity, match the corresponding learning activity, and obtain the initial learning activity; In combination with the user's real-time learning status, the activity generation path is optimized by adding path branches, pruning path branches, or replacing path nodes to obtain an activity optimization path; Optimizing the initial learning activity through the activity optimization path to obtain an optimized learning activity, so as to dynamically optimize the interactive learning activity; The step of generating a corresponding activity generation path based on the user status tag and in combination with a preset course knowledge base includes: Extract the user's interest characteristics and ability characteristics based on the user status label and construct the user status vector; Based on the preset course knowledge base, knowledge activities are used as activity nodes. The logical relationship between the knowledge points included in each knowledge activity is analyzed according to the knowledge association and activity difficulty. Associations are established between activity nodes to construct a knowledge graph. The logical relationships include sequence, inclusion, and parallel relationships. According to the user state vector, the corresponding learning activities are matched in the knowledge graph, and the activity nodes that meet the user's interests and abilities are screened out. Based on the association relationship between the nodes in the knowledge graph, the corresponding learning paths are searched starting from the matched nodes to obtain a set of candidate paths; The matching value between each path in the candidate path set and the user state vector is calculated by configuring a matching score calculation rule, and the path with the highest matching value is selected from the candidate path set as the activity generation path. The matching score calculation rule includes: for the interest matching score, the interest matching score of each activity node is assigned according to the degree of inclusion of the knowledge field of the activity node and the user's interest keywords, and the interest matching scores of all activity nodes in the path are added to obtain the total interest matching score of the path; for the ability matching score, the ability matching score of each activity node is assigned according to the degree of matching between the difficulty level of the activity node and the ability quantification value in the user state vector, and the ability matching scores of all activity nodes in the path are added to obtain the total ability matching score of the path; the total interest matching score and the total ability matching score of each path are weighted and summed to obtain the matching value between each path and the user state vector.

2. The method for dynamic optimization of interactive learning activities according to claim 1, characterized in that: The method of obtaining the user's learning data, extracting the user's interest tags and learning ability tags through a preset user behavior analysis model, and generating the user status tags includes: Obtain user learning data by collecting user behavior logs, learning time, and interaction records; Based on the learning data, extract the user's interest tags and learning ability tags through a preset user behavior analysis model; The interest tag and the learning ability tag are integrated to generate a user status tag.

3. The method for dynamic optimization of interactive learning activities according to claim 1, characterized in that: Generating a path according to the activity, matching corresponding learning activities, and obtaining initial learning activities includes: According to the activity generation path, a corresponding first learning activity is matched for each activity node in a preset activity template library; Optimizing the first learning activity based on the user state vector to obtain a second learning activity; The second learning activities are combined according to the order of activity nodes in the activity generation path to obtain an initial learning activity.

4. The method for dynamic optimization of interactive learning activities according to claim 1, characterized in that: The step of optimizing the activity generation path by adding path branches, pruning path branches, or replacing path nodes in combination with the user's real-time learning status to obtain an activity optimization path includes: Based on the user's real-time learning status, the path optimization area is identified by analyzing the activity node density and dependency relationships in the activity generation path; Optimizing the path optimization area by adding path branches, pruning path branches, or replacing path nodes to obtain a first optimized path; The activity generation path is optimized according to the first optimization path to obtain an activity optimization path.

5. The method for dynamic optimization of interactive learning activities according to claim 4, characterized in that: The step of optimizing the path optimization area by adding path branches, pruning path branches, or replacing path nodes to obtain a first optimized path includes: Analyze users' real-time interests and learning activity completion rates based on their real-time learning status; According to the real-time interest, a path branch corresponding to the activity is added to the path optimization area to obtain a first optimization area; When the completion rate of the learning activity is lower than a preset completion rate threshold, the path branch of the corresponding activity is pruned to obtain a second optimized area; By analyzing the matching degree between the real-time learning state and the knowledge graph, the path nodes whose matching degree is less than a preset matching degree threshold are replaced to obtain a third optimized region; The first optimization region, the second optimization region, and the third optimization region are combined to obtain a first optimization path.

6. The method for dynamic optimization of interactive learning activities according to claim 1, characterized in that: The step of optimizing the initial learning activity through the activity optimization path to obtain an optimized learning activity, so as to dynamically optimize the interactive learning activity, includes: According to the activity optimization path, the initial learning activity is optimized in layers to obtain the corresponding first optimized activity; The first optimization activities are combined to obtain optimized learning activities, so as to dynamically optimize the interactive learning activities.

7. The method for dynamic optimization of interactive learning activities according to claim 6, characterized in that: The initial learning activity is hierarchically optimized according to the activity optimization path to obtain the corresponding first optimization activity, including: According to the optimized activity nodes in the activity optimization path, the corresponding learning activities are matched in the preset activity template library to obtain the optimized activities; The optimized activity replaces the corresponding learning activity in the initial learning activity, and the order and duration of the learning activities are adjusted to obtain a first optimized activity.

8. A dynamic optimization system for interactive learning activities, characterized in that: A method for dynamically optimizing interactive learning activities according to any one of claims 1 to 7, comprising: The user status analysis module obtains the user's learning data, extracts the user's interest tags and learning ability tags through the preset user behavior analysis model, and generates user status tags; An activity generation path formulation module generates a corresponding activity generation path based on the user status tag and in combination with a preset course knowledge base; A learning activity formulation module generates a path based on the activity, matches the corresponding learning activity, and obtains the initial learning activity; An activity generation path optimization module optimizes the activity generation path by adding path branches, pruning path branches, or replacing path nodes in combination with the user's real-time learning status to obtain an activity optimization path; The learning activity optimization module optimizes the initial learning activity through the activity optimization path to obtain an optimized learning activity, so as to dynamically optimize the interactive learning activity.

Images