An online education resource adaptive recommendation method and system based on artificial intelligence

Abstract

The application discloses an online education resource adaptive recommendation method and system based on artificial intelligence, relates to the technical field of content recommendation, and comprises the following steps: acquiring candidate online education resources; obtaining an importance degree according to the number of occurrences of a kth knowledge point in the candidate online education resources; determining resource type characteristics according to the resource types of the candidate online education resources; determining resource behavior characteristics according to the education behaviors of the candidate online education resources; acquiring a mastering probability of the kth knowledge point after tth learning of a user, learning type characteristics and learning behavior characteristics of the user; determining a knowledge matching degree by using the importance degree and the mastering probability; determining a type matching degree by using the resource and learning type characteristics; determining a behavior matching degree by using the resource and learning behavior characteristics; determining a target online education resource set according to the knowledge, type and behavior matching degrees; and recommending resources in the target online education resource set to the user. The method can accurately recommend online education resources.

Description

Technical Field

[0001] This application relates to the field of content recommendation technology, and in particular to an adaptive recommendation method and system for online educational resources based on artificial intelligence. Background Technology

[0002] With the rapid development of internet technology and artificial intelligence, online education has become an important part of the education system. A vast amount of educational resources provides learners with abundant learning options, but it also brings the problem of information overload. Faced with a deluge of resources, learners often find it difficult to quickly identify the learning content most suitable for them, leading to low learning efficiency and decreased interest in learning.

[0003] Traditional content-based recommendation methods extract features from educational resources, such as text, video, and audio, as well as learners' historical learning records, and calculate the similarity between the two to make recommendations. While traditional methods can recommend some relevant educational resources to users, these resources often have poor matching accuracy with the user.

[0004] Therefore, traditional solutions suffer from inaccurate recommendations. Summary of the Invention

[0005] This application provides an adaptive recommendation method and system for online educational resources based on artificial intelligence, which can accurately recommend online educational resources.

[0006] To achieve the above objectives, this application adopts the following technical solution: Firstly, this application provides an adaptive recommendation method for online educational resources based on artificial intelligence, including: Obtain candidate online educational resources; The importance of the k-th knowledge point in the candidate online educational resources is determined by the frequency of its occurrence; the resource type characteristics of the candidate online educational resources are determined by their resource type; and the resource behavior characteristics of the candidate online educational resources are determined by their educational behavior. Obtain the probability that a user has mastered the k-th knowledge point after the t-th learning session, as well as the user's learning type characteristics and learning behavior characteristics; The knowledge matching degree between the candidate online education resources and the user is determined by using the importance of the k-th knowledge point in the candidate online education resources and the probability of the user mastering the k-th knowledge point after the t-th learning session; the type matching degree between the candidate online education resources and the user is determined by using resource type features and learning type features; and the behavior matching degree between the candidate online education resources and the user is determined by using resource behavior features and learning behavior features. Based on the knowledge matching degree, the type matching degree, and the behavior matching degree, the overall matching degree between the candidate online education resources and the user is determined; Based on the comprehensive matching degree, the target set of online educational resources is determined; Recommend online educational resources from the target set of online educational resources to the user.

[0007] Optionally, determining the knowledge matching degree between the candidate online educational resources and the user by utilizing the importance of the k-th knowledge point in the candidate online educational resources and the probability of the user mastering the k-th knowledge point after the t-th learning session includes: Based on the probability that a user has mastered the kth knowledge point after the tth learning session, determine the probability that the user has not mastered the kth knowledge point after the tth learning session. The knowledge matching degree between the candidate online education resources and the user is determined by using the importance of the k-th knowledge point in the candidate online education resources and the probability that the user has not mastered the k-th knowledge point after the t-th learning session.

[0008] Optionally, the resource type features include a first visual type score, a first auditory type score, and a first kinesthetic type score, and the learning type features include a second visual type score, a second auditory type score, and a second kinesthetic type score; determining the type matching degree between the candidate online educational resources and the user using the resource type features and learning type features includes: Multiply the first visual type score by the second visual type score to obtain the sub-visual type score; multiply the first auditory type score by the second auditory type score to obtain the sub-auditory type score; multiply the first kinesthetic type score by the second kinesthetic type score to obtain the sub-kinesthetic type score. The sum of the sub-visual type score, sub-auditory type score, and sub-kinesthetic type score is used as the type matching degree between the candidate online educational resource and the user.

[0009] Optionally, the resource behavior features include a first fragmentation score, a first systematic score, and a first interactive score; the learning behavior features include a second fragmentation score, a second systematic score, and a second interactive score; and determining the behavioral matching degree between the candidate online educational resources and the user using the resource behavior features and learning behavior features includes: Multiply the first fragmented score by the second fragmented score to obtain the sub-fragmented score; multiply the first systematic score by the second systematic score to obtain the sub-systematic score; multiply the first interactive score by the second interactive score to obtain the sub-interactive score. The sum of the sub-fragmented score, the sub-systematic score, and the sub-interactive score is used as the behavioral matching degree between the candidate online education resource and the user.

[0010] Optionally, determining the comprehensive matching degree between the candidate online education resources and the user based on the knowledge matching degree, the type matching degree, and the behavior matching degree includes: The first weight for knowledge matching, the second weight for type matching, and the third weight for behavior matching are determined. Using the first weight, the second weight, and the third weight, the knowledge matching degree, the type matching degree, and the behavior matching degree are weighted and summed to obtain the comprehensive matching degree between the candidate online education resources and the user.

[0011] Optionally, obtaining the probability that a user has mastered the k-th knowledge point after the t-th learning session includes: When you answer correctly:

[0012] in, This indicates that after a user completes the t-th learning session and answers correctly, the knowledge points obtained are updated. The probability of mastering Indicates the user at time t The probability of mastering knowledge point k after one learning session. This represents the probability of a user having mastered knowledge point k but answering incorrectly due to a mistake. This represents the difficulty coefficient of the corresponding question in the t-th learning session. This represents the probability that a user guesses the answer correctly even though they haven't mastered knowledge point k. This represents the user's mastery status event of knowledge point k after the t-th learning session. This represents the result event of the user's interaction with the question in the t-th learning session. A slippage event indicates that the user has mastered the knowledge point but answered incorrectly. This refers to a guessing event in which a user answered a question correctly even though they did not understand the knowledge point. When answering incorrectly:

[0013] in, This represents the probability of a user mastering knowledge point k after completing the t-th learning session and answering incorrectly.

[0014] Optionally, the step of determining the importance of the k-th knowledge point in the candidate online educational resources based on the frequency of its occurrence includes:

[0015] in, This indicates the importance of the k-th knowledge point in the j-th candidate online educational resource. This represents the number of times the k-th knowledge point appears in the j-th candidate online educational resource. This represents the weight coefficient of the position of the k-th knowledge point in the j-th candidate online educational resource. This represents the total number of knowledge points covered by the j-th candidate online educational resource. This represents the number of times the m-th knowledge point appears in the j-th candidate online educational resource. This represents the weight coefficient of the position of the m-th knowledge point in the j-th candidate online educational resource.

[0016] Secondly, this application provides an adaptive recommendation system for online educational resources based on artificial intelligence, including: The acquisition module is used to acquire candidate online educational resources; The processing module is used to determine the importance of the kth knowledge point in the candidate online education resources based on the frequency of its occurrence; to determine the resource type characteristics of the candidate online education resources based on their resource type; and to determine the resource behavior characteristics of the candidate online education resources based on their educational behavior. The acquisition module is used to acquire the probability of a user mastering the k-th knowledge point after the t-th learning session, as well as the user's learning type characteristics and learning behavior characteristics. The processing module is used to determine the knowledge matching degree between the candidate online educational resources and the user by utilizing the importance of the k-th knowledge point in the candidate online educational resources and the probability of the user mastering the k-th knowledge point after the t-th learning session; to determine the type matching degree between the candidate online educational resources and the user by utilizing resource type features and learning type features; to determine the behavior matching degree between the candidate online educational resources and the user by utilizing resource behavior features and learning behavior features; to determine the comprehensive matching degree between the candidate online educational resources and the user based on the knowledge matching degree, the type matching degree, and the behavior matching degree; and to determine the target online educational resource set based on the comprehensive matching degree. The recommendation module is used to recommend online educational resources from the target set of online educational resources to the user.

[0017] Thirdly, this application provides a computing device, including a memory and a processor; The memory stores one or more computer programs, the one or more computer programs including instructions; when the instructions are executed by the processor, the computing device performs the method as described in any one of the first aspects.

[0018] Fourthly, this application provides a computer-readable storage medium for storing a computer program for performing the method as described in any one of the first aspects.

[0019] As can be seen from the above technical solution, this application has at least the following beneficial effects: This application acquires candidate online educational resources, determines the importance of knowledge points based on the frequency of their appearance in the resources, and identifies resource type characteristics and resource behavior characteristics based on resource type and educational behavior, respectively. After acquiring the probability of user mastery of knowledge points, learning type characteristics, and learning behavior characteristics after learning, it calculates the knowledge matching degree, type matching degree, and behavior matching degree between resources and users, respectively. Finally, it obtains a comprehensive matching degree based on these three types of matching degrees, thereby determining the target resource set and completing the recommendation. This application can correlate the importance of resource knowledge points with the probability of user mastery of knowledge points, the resource type characteristics with user learning type characteristics, and the resource behavior characteristics with user learning behavior characteristics. This reduces the bias caused by matching based on only a single dimension, mitigates the impact of information overload on user learning choices, and provides users with educational resources that are more closely aligned with their learning status and habits.

[0020] Furthermore, after acquiring candidate online educational resources, this application determines the importance of knowledge points based on their frequency of occurrence within the resources. Simultaneously, it identifies resource type features and resource behavior features based on resource type and educational behavior, respectively. After acquiring the user's mastery probability of knowledge points, learning type features, and learning behavior features, the application first determines the non-mastery probability based on the user's mastery probability. Then, it combines this with the importance of knowledge points in the resources to obtain the knowledge matching degree. The type matching degree is calculated by mapping resource type features to learning type features, and the behavior matching degree is calculated by mapping resource behavior features to learning behavior features. Finally, the weights of each matching degree are used to weighted sum the three types of matching degrees to obtain the comprehensive matching degree, thereby determining the target resource set and completing the recommendation. This method can correlate the importance of resource knowledge points with the user's mastery probability, resource type features with the user's learning type features, and resource behavior features with the user's learning behavior features. This reduces the bias caused by matching based on only a single dimension and mitigates the impact of information overload on user learning choices. Thus, educational resources are more closely matched with users, facilitating user learning of educational resources.

[0021] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0022] Figure 1 A flowchart illustrating an adaptive recommendation method for online educational resources based on artificial intelligence, provided for embodiments of this application; Figure 2 A schematic diagram of an adaptive recommendation system for online educational resources based on artificial intelligence, provided for an embodiment of this application; Figure 3 This is a schematic diagram of a computing device provided in an embodiment of this application. Detailed Implementation

[0023] The terms "first," "second," and "third," etc., used in this application specification and accompanying drawings are used to distinguish different objects, not to limit a specific order.

[0024] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0025] Current conventional recommendation methods only extract surface-level content features of online educational resources and learners' historical learning records, and complete resource recommendations through simple similarity comparisons, without fully considering learners' real-time learning status and personalized learning needs. Specifically, existing methods do not combine learners' probability of mastering each knowledge point when selecting resources, and cannot match content on knowledge points that learners have not mastered; at the same time, they do not take into account learners' learning type characteristics and learning behavior characteristics, and cannot recommend resources whose presentation format and learning pace are adapted to learners' habits. Ultimately, this results in a mismatch between recommended content and learners' learning progress, receiving habits, and learning pace, making it difficult to meet learners' true learning needs.

[0026] In view of this, embodiments of this application provide an adaptive recommendation method for online educational resources based on artificial intelligence, which can be executed by a processing device. The processing device can be a terminal or a server. Terminals include, but are not limited to, smartphones, tablets, laptops, personal digital assistants, or smart wearable devices. The server can be a cloud server, such as a central server in a central cloud computing cluster or an edge server in an edge cloud computing cluster. Alternatively, the server can be a server in a local data center. A local data center refers to a data center directly controlled by the user.

[0027] To address the issue of insufficient alignment between recommended content and learners' actual learning needs in existing online education resource recommendation methods, this application focuses on the matching degree between online education resources and learners. It extracts multi-dimensional features of both resources and learners, calculates the matching degree across dimensions, and weights and integrates these features to obtain a comprehensive matching degree, thus achieving resource-learner adaptation. Specifically, it first extracts features from candidate online education resources, obtaining the importance of knowledge points, resource type features, and resource behavior features. Then, it acquires the learner's real-time learning status and personalized features, including the probability of mastering each knowledge point, learning type features, and learning behavior features. Subsequently, it calculates the matching degree across the knowledge, type, and behavior dimensions, combines the weights of each dimension to obtain a comprehensive matching degree, and selects and pushes target resources based on the comprehensive matching degree. This solves the problem that existing recommendation methods do not take into account learners' real-time status and personalized needs, achieving adaptive resource recommendation.

[0028] To make the technical solution of this application clearer and easier to understand, the following description, in conjunction with the accompanying drawings, introduces an adaptive recommendation method for online educational resources based on artificial intelligence, as provided in the embodiments of this application. Figure 1 As shown, this figure is a flowchart of an adaptive recommendation method for online educational resources based on artificial intelligence, provided in an embodiment of this application. In this embodiment, the method includes: S201, The processing device acquires candidate online educational resources.

[0029] Candidate online educational resources are various online learning resources that can be screened and compared, and are used to push to users.

[0030] The processing device retrieves various online learning content stored in the storage area (including online and local), completes unified collection and organization, and obtains candidate online education resources.

[0031] S202. The processing device determines the importance of the kth knowledge point in the candidate online educational resources based on the frequency of its occurrence; it determines the resource type characteristics of the candidate online educational resources based on their resource type; and it determines the resource behavior characteristics of the candidate online educational resources based on their educational behavior.

[0032] Knowledge points are independent learning units formed by dividing subject learning content. Frequency of occurrence refers to the number of times a single knowledge point appears within the same online educational resource. Importance refers to the proportion of content a knowledge point occupies within the corresponding online educational resource. Resource type refers to the category of content presentation format corresponding to the online educational resource. Resource type characteristics are the corresponding data scores reflecting the presentation format of the online educational resource. Educational behavior is the learning activity attribute inherent in the online educational resource. Resource behavior characteristics are the corresponding data scores reflecting the learning activity format of the online educational resource.

[0033] The processing device counts the frequency of each knowledge point within each candidate online educational resource, performs calculations based on the location of each knowledge point, and determines the importance of the corresponding knowledge point within the resource. The processing device also categorizes and assigns values ​​to candidate online educational resources according to their content presentation format, determining the corresponding resource type characteristics. Furthermore, the processing device numerically classifies candidate online educational resources based on their inherent learning activity attributes, determining the corresponding resource behavior characteristics. The resulting knowledge point importance reflects the proportion of knowledge points within the resource content; the resulting resource type characteristics reflect the content presentation format characteristics of the resource; and the resulting resource behavior characteristics reflect the learning activity characteristics adapted to the resource.

[0034] The expression for the importance of the k-th knowledge point in candidate online educational resources is:

[0035] in, This indicates the importance of the k-th knowledge point in the j-th candidate online educational resource. This represents the number of times the k-th knowledge point appears in the j-th candidate online educational resource. This represents the weight coefficient of the position of the k-th knowledge point in the j-th candidate online educational resource. This represents the total number of knowledge points covered by the j-th candidate online educational resource. This represents the number of times the m-th knowledge point appears in the j-th candidate online educational resource. This represents the weight coefficient of the position of the m-th knowledge point in the j-th candidate online educational resource.

[0036] Resource type characteristics include primary visual type score, primary auditory type score, and primary kinesthetic type score. The primary visual type score represents the percentage of visual content in the candidate online educational resources. The primary auditory type score represents the percentage of auditory content in the candidate online educational resources. The primary kinesthetic type score represents the percentage of kinesthetic / hands-on content in the candidate online educational resources.

[0037] The expression for resource type characteristics is:

[0038] in, This represents the resource type feature vector of the j-th candidate online education resource. Indicates the first visual type score. This indicates the score for the first auditory type. This indicates the score for the first kinesthetic type.

[0039] In some embodiments, the processing device directly assigns the corresponding resource type feature based on the content presentation format of the candidate online educational resources: If the candidate online education resource is a text and image courseware, and the resource is mainly composed of text and images, the processing device can be set to a score of 0.9 for the first visual type, 0.1 for the first auditory type, and 0 for the first kinesthetic type. If the candidate online education resource is an audio course, and the resource is mainly based on voice explanation, the processing device can be set to a score of 0.1 for the first visual type, 0.9 for the first auditory type, and 0 for the first kinesthetic type. If the candidate online educational resources are interactive operation-type experimental tutorials, the resources should focus on practical demonstrations and interactive exercises. The processing device can be set to score 0.2 for the first visual type, 0.1 for the first auditory type, and 0.7 for the first kinesthetic type.

[0040] The processing device combines these three scores into a vector to obtain the resource type characteristics of the candidate online education resource. Different types of resources will receive different score combinations, thereby reflecting the presentation characteristics of the resource.

[0041] Resource behavior characteristics include a first fragmentation score, a first systematic score, and a first interactive score. The first fragmentation score represents the numerical value of a candidate online educational resource's suitability for short-term, fragmented learning scenarios. The first systematic score represents the numerical value of a candidate online educational resource's suitability for coherent, systematic learning scenarios. The first interactive score represents the numerical value of a candidate online educational resource's suitability for interactive, communicative learning scenarios.

[0042] The expression for resource behavior characteristics is:

[0043] in, This represents the resource behavior feature vector of the j-th candidate online education resource. This represents the first fragment score. This indicates the first systematic score. This indicates the first interactive score.

[0044]

[0045] in, This indicates the first fragmented score in the middle. This represents the duration of the j-th candidate online educational resource.

[0046]

[0047] in, This indicates the first systematic score in the middle. This represents the number of knowledge points contained in the j-th candidate online educational resource. This indicates the total number of knowledge points in the chapter to which this resource belongs; The variable is an indicator variable, which takes the value 1 if the j-th candidate online education resource contains a knowledge framework graph, and 0 otherwise.

[0048]

[0049] in, This indicates the first interactive score in the middle. This represents the number of interactive elements in the j-th candidate online educational resource; This represents the duration of the j-th candidate online educational resource; This indicates an indicator variable; it takes the value 1 if the j-th candidate online educational resource supports real-time Q&A, and 0 otherwise. Indicates the first The number of interactive elements in each candidate online educational resource; Indicates the first The duration of each candidate online educational resource.

[0050] Regarding the above , and Perform overall normalization to obtain resource behavior feature vector components (first fragmentation score). First systematic score And the first interactive score ).

[0051] S203. The processing device obtains the probability of a user mastering the k-th knowledge point after the t-th learning session, as well as the user's learning type characteristics and learning behavior characteristics.

[0052] Mastery probability is a quantifiable value representing the degree to which a user understands and applies knowledge points after learning. Learning type characteristics are corresponding data scores reflecting a user's habits of receiving knowledge. Learning behavior characteristics are corresponding data scores reflecting a user's preferred methods of conducting daily learning.

[0053] The processing device retrieves the user's past learning data and study records for each knowledge point, substitutes them into a preset formula to calculate the probability of mastering the k-th knowledge point after the t-th study session. The device also retrieves data on the user's preferences for different presentation formats of resources during daily learning, assigning values ​​to the user's learning type characteristics. Finally, the device retrieves data on the user's daily learning activities, assigning values ​​to the user's learning behavior characteristics. The obtained mastery probability reflects the user's current understanding and application of the corresponding knowledge point; the obtained learning type characteristics reflect the user's knowledge acquisition habits; and the obtained learning behavior characteristics reflect the user's preferred learning activity format.

[0054] The probability that a user has mastered the k-th knowledge point after the t-th learning session is as follows: When you answer correctly:

[0055] in, This indicates that after a user completes the t-th learning session and answers correctly, the knowledge points obtained are updated. The probability of mastering Indicates the user at time t The probability of mastering knowledge point k after one learning session. This represents the probability of a user having mastered knowledge point k but answering incorrectly due to a mistake. This represents the difficulty coefficient of the corresponding question in the t-th learning session. This represents the probability that a user guesses the answer correctly even though they haven't mastered knowledge point k. This represents the user's mastery status event of knowledge point k after the t-th learning session. This represents the result event of the user's interaction with the question in the t-th learning session. A slippage event indicates that the user has mastered the knowledge point but answered incorrectly. This refers to a guessing event in which a user answered a question correctly even though they did not understand the knowledge point. The processing equipment summarizes historical learning data and statistically analyzes the knowledge points that have been determined to be mastered. The frequency of incorrect answers by users on similar test questions is calculated by dividing the total number of incorrect answers by the total number of answers given by the group. .

[0056] The processing device summarizes historical learning data, calculates the frequency of correct answers for similar test questions by users who have not been assessed as having mastered knowledge point k, and divides the total number of correct answers by the total number of answers for this group to obtain the result. .

[0057] When answering incorrectly:

[0058] in, This represents the probability of a user mastering knowledge point k after completing the t-th learning session and answering incorrectly.

[0059] Learning type characteristics include second visual type score, second auditory type score, and second kinesthetic type score. The second visual type score represents the degree to which a user prefers visual learning resources. The second auditory type score represents the degree to which a user prefers auditory learning resources. The second kinesthetic type score represents the degree to which a user prefers kinesthetic / hands-on learning resources.

[0060] In some embodiments, the processing device counts the percentage of time or frequency a user uses visual, auditory, and kinesthetic learning resources within a preset period; and directly uses the percentage of each type of resource as the corresponding score value to obtain the second visual type score, the second auditory type score, and the second kinesthetic type score.

[0061] For example: If a user's total learning time in the past month is 100 hours, of which 70 hours were spent on visual resources such as text, images, and videos, 20 hours on auditory resources such as audio courses, and 10 hours on kinesthetic resources such as interactive hands-on activities and simulations, then: Second visual type score = 70 ÷ 100 = 0.7; Second auditory type score = 20 ÷ 100 = 0.2; Second kinesthetic type score = 10 ÷ 100 = 0.1.

[0062] Learning behavior characteristics include a second fragmentation score, a second systematic score, and a second interactive score. The second fragmentation score represents the degree to which a user prefers short, fragmented learning formats. The second systematic score represents the degree to which a user prefers coherent, systematic learning formats. The second interactive score represents the degree to which a user prefers interactive, communicative learning formats.

[0063] In some embodiments, the processing device collects data related to the user's daily learning behavior, defines intervals according to corresponding evaluation criteria, and then determines the scores for each item.

[0064] For example: Single study sessions shorter than 15 minutes are considered fragmented learning; completing an entire chapter consecutively is considered systematic learning; and actively participating in Q&A sessions is considered interactive learning. If we analyze 100 learning activities a user has engaged in over the past month, and 80 were short, fragmented learning sessions, 15 were complete chapter studies, and 5 were interactive Q&A sessions, then: Second fragmentation score = 80 ÷ 100 = 0.8 Second systematic score = 15 ÷ 100 = 0.15 The second interactive score = 5 ÷ 100 = 0.05.

[0065] S204. The processing device uses the importance of the k-th knowledge point in the candidate online educational resources and the probability of the user mastering the k-th knowledge point after the t-th learning session to determine the knowledge matching degree between the candidate online educational resources and the user; it uses resource type features and learning type features to determine the type matching degree between the candidate online educational resources and the user; and it uses resource behavior features and learning behavior features to determine the behavior matching degree between the candidate online educational resources and the user.

[0066] Knowledge matching score represents the numerical value indicating how well the candidate online educational resources match the user's knowledge level. Type matching score represents the numerical value indicating how well the candidate online educational resources match the user's learning type preferences. Behavioral matching score represents the numerical value indicating how well the candidate online educational resources match the user's learning behavior habits.

[0067] In some embodiments, the processing device determines the probability that a user has not mastered the k-th knowledge point after the t-th learning session based on the probability that the user has mastered the k-th knowledge point after the t-th learning session; and determines the knowledge matching degree between the candidate online educational resources and the user by using the importance of the k-th knowledge point in the candidate online educational resources and the probability that the user has not mastered the k-th knowledge point after the t-th learning session.

[0068] The probability of not mastering something is a quantitative value representing a user's failure to understand and apply the corresponding knowledge points.

[0069] The processing device subtracts the user's probability of mastering a corresponding knowledge point from a fixed value to calculate the probability of not mastering that knowledge point. It then integrates the importance of the knowledge point with the corresponding probability of not mastering it to arrive at the knowledge matching degree. The obtained probability of not mastering reflects the user's current deficiency in that knowledge point, while the obtained knowledge matching degree reflects the degree to which the online educational resource content matches the user's knowledge gaps.

[0070] The formula for calculating the knowledge matching degree between candidate online educational resources and users is as follows:

[0071] in, This indicates that user U and the j-th candidate online education resource Knowledge matching degree This indicates the importance of the k-th knowledge point in the j-th candidate online educational resource. This indicates that after a user completes the t-th learning session and answers correctly / incorrectly, the knowledge points obtained are updated. The probability of mastering This represents the total number of knowledge points covered by the j-th candidate online educational resource. This represents the probability that a user has not mastered the k-th knowledge point after the t-th learning session.

[0072] In some embodiments, the processing device multiplies the first visual type score by the second visual type score to obtain a sub-visual type score, multiplies the first auditory type score by the second auditory type score to obtain a sub-auditory type score, and multiplies the first kinesthetic type score by the second kinesthetic type score to obtain a sub-kinesthetic type score; the sum of the sub-visual type score, the sub-auditory type score, and the sub-kinesthetic type score is used as the type matching degree between the candidate online educational resource and the user.

[0073] The expression for the type match between candidate online educational resources and users is:

[0074] in, This indicates that user U and the j-th candidate online education resource Type matching degree, This represents a feature vector representing the user's learning type. This represents the resource type feature vector of the j-th candidate online education resource. This represents the second visual type score in the user's learning type feature vector. This represents the second auditory type score in the user's learning type feature vector. This represents the second kinesthetic type score in the user's learning type feature vector. Indicates the first visual type score. This indicates the score for the first auditory type. This indicates the score for the first kinesthetic type.

[0075] In some embodiments, the processing device multiplies a first fragmented score by a second fragmented score to obtain a sub-fragmented score, multiplies a first systematic score by a second systematic score to obtain a sub-systematic score, and multiplies a first interactive score by a second interactive score to obtain a sub-interactive score; the sum of the sub-fragmented score, the sub-systematic score, and the sub-interactive score is used as the behavioral matching degree between the candidate online education resource and the user.

[0076] The expression for the match between candidate online educational resources and user behavior is:

[0077] in, This indicates that user U and the j-th candidate online education resource Behavioral matching degree This represents a user's learning behavior feature vector. This represents the resource behavior feature vector of the j-th candidate online education resource. This represents the second fragmentation score in the user's learning behavior feature vector. This represents the second systematic score in the user's learning behavior feature vector. This represents the second interactive score in the user's learning behavior feature vector. This represents the first fragment score. This indicates the first systematic score. This indicates the first interactive score.

[0078] S205. The processing device determines the overall matching degree between candidate online educational resources and users based on knowledge matching degree, type matching degree, and behavior matching degree.

[0079] The overall matching degree represents the degree to which candidate online education resources fit the overall learning needs, preferences, and habits of users.

[0080] In some embodiments, the processing device obtains a first weight for knowledge matching, a second weight for type matching, and a third weight for behavior matching; using the first weight, the second weight, and the third weight, it performs a weighted summation of the knowledge matching, type matching, and behavior matching to obtain the comprehensive matching degree between the candidate online education resources and the user.

[0081] The expression for the overall matching degree between candidate online educational resources and users is:

[0082] in, This indicates that user U and the j-th candidate online education resource Overall matching degree The first weight representing the knowledge matching degree The second weight representing the type matching degree, The third weight represents the degree of behavioral matching.

[0083] S206. The processing equipment determines the target set of online educational resources based on the overall matching degree.

[0084] The target online education resource collection is a list of recommended resources filtered based on user matching.

[0085] The processing device sorts all candidate online educational resources by their overall matching degree from highest to lowest, and selects resources with an overall matching degree not lower than a preset threshold, or selects a preset number of resources from the top of the ranking, to form a target online educational resource set. The resulting target online educational resource set is a collection of online educational resources that are well-suited to the user.

[0086] S207. The processing device recommends online educational resources from the target set of online educational resources to the user.

[0087] The processing device displays the resources from the target online educational resource collection to the user through the platform interface, according to a preset order or sorting method. The resulting display presents the online educational resources from the target collection to the user, providing learning resource options tailored to their learning needs.

[0088] Based on the above description, this application has the following beneficial effects: This application determines the importance of knowledge points, resource type characteristics, and resource behavior characteristics of candidate online educational resources. Simultaneously, it collects the user's corresponding knowledge point mastery probability, learning type characteristics, and learning behavior characteristics. Matching scores are calculated at the knowledge, type, and behavioral levels, and then integrated to obtain a comprehensive matching score for resource filtering and recommendation. This method can filter suitable content based on the user's actual learning situation, broaden the reference range for resource matching through multi-dimensional comparison, and reduce the filtering bias caused by a single comparison method. It can adapt to the different learning conditions and daily learning habits of different users, rationally organize massive amounts of online educational resources, reduce the interference of irrelevant learning content, and accurately recommend online educational resources.

[0089] The above text combined Figure 1 The adaptive recommendation method for online educational resources based on artificial intelligence provided in the embodiments of this application has been described in detail. The system and device provided in the embodiments of this application will be described below with reference to the accompanying drawings.

[0090] like Figure 2 As shown in the figure, this is a schematic diagram of an adaptive recommendation system for online educational resources based on artificial intelligence provided in an embodiment of this application. The system includes: Module 301 is used to acquire candidate online educational resources; The processing module 302 is used to determine the importance of the kth knowledge point in the candidate online education resources based on the frequency of its occurrence; to determine the resource type characteristics of the candidate online education resources based on their resource type; and to determine the resource behavior characteristics of the candidate online education resources based on their educational behavior. The acquisition module 301 is used to acquire the probability of a user mastering the k-th knowledge point after the t-th learning session, as well as the user's learning type characteristics and learning behavior characteristics. Processing module 302 is configured to: determine the knowledge matching degree between the candidate online educational resources and the user by utilizing the importance of the k-th knowledge point in the candidate online educational resources and the probability of the user mastering the k-th knowledge point after the t-th learning session; determine the type matching degree between the candidate online educational resources and the user by utilizing resource type features and learning type features; determine the behavior matching degree between the candidate online educational resources and the user by utilizing resource behavior features and learning behavior features; determine the comprehensive matching degree between the candidate online educational resources and the user based on the knowledge matching degree, the type matching degree, and the behavior matching degree; and determine the target online educational resource set based on the comprehensive matching degree. The recommendation module 303 is used to recommend online educational resources from the target online educational resource set to the user.

[0091] Optionally, the processing module 302 is specifically used to determine the probability that the user has not mastered the kth knowledge point after the tth learning session, based on the probability that the user has mastered the kth knowledge point after the tth learning session. The knowledge matching degree between the candidate online education resources and the user is determined by using the importance of the k-th knowledge point in the candidate online education resources and the probability that the user has not mastered the k-th knowledge point after the t-th learning session.

[0092] Optionally, the resource type features include a first visual type score, a first auditory type score, and a first kinesthetic type score, and the learning type features include a second visual type score, a second auditory type score, and a second kinesthetic type score; the processing module 302 is specifically used to multiply the first visual type score and the second visual type score to obtain a sub-visual type score, multiply the first auditory type score and the second auditory type score to obtain a sub-auditory type score, and multiply the first kinesthetic type score and the second kinesthetic type score to obtain a sub-kinesthetic type score; The sum of the sub-visual type score, sub-auditory type score, and sub-kinesthetic type score is used as the type matching degree between the candidate online educational resource and the user.

[0093] Optionally, the resource behavior features include a first fragmented score, a first systematic score, and a first interactive score, and the learning behavior features include a second fragmented score, a second systematic score, and a second interactive score. The processing module 302 is specifically used to multiply the first fragmented score and the second fragmented score to obtain a sub-fragmented score, multiply the first systematic score and the second systematic score to obtain a sub-systematic score, and multiply the first interactive score and the second interactive score to obtain a sub-interactive score. The sum of the sub-fragmented score, the sub-systematic score, and the sub-interactive score is used as the behavioral matching degree between the candidate online education resource and the user.

[0094] Optionally, the processing module 302 is specifically used to obtain the first weight of knowledge matching degree, the second weight of type matching degree, and the third weight of behavior matching degree; Using the first weight, the second weight, and the third weight, the knowledge matching degree, the type matching degree, and the behavior matching degree are weighted and summed to obtain the comprehensive matching degree between the candidate online education resources and the user.

[0095] Optionally, processing module 302 is specifically used to obtain the probability that a user has mastered the k-th knowledge point after the t-th learning session, including: When you answer correctly:

[0096] in, This indicates that after a user completes the t-th learning session and answers correctly, the knowledge points obtained are updated. The probability of mastering Indicates the user at time t The probability of mastering knowledge point k after one learning session. This represents the probability of a user having mastered knowledge point k but answering incorrectly due to a mistake. This represents the difficulty coefficient of the corresponding question in the t-th learning session. This represents the probability that a user guesses the answer correctly even though they haven't mastered knowledge point k. This represents the user's mastery status event of knowledge point k after the t-th learning session. This represents the result event of the user's interaction with the question in the t-th learning session. A slippage event indicates that the user has mastered the knowledge point but answered incorrectly. This refers to a guessing event in which a user answered a question correctly even though they did not understand the knowledge point. When answering incorrectly:

[0097] in, This represents the probability of a user mastering knowledge point k after completing the t-th learning session and answering incorrectly.

[0098] Optionally, processing module 302 is specifically used to determine the importance of the k-th knowledge point in the candidate online educational resources based on the frequency of its occurrence, including:

[0099] in, This indicates the importance of the k-th knowledge point in the j-th candidate online educational resource. This represents the number of times the k-th knowledge point appears in the j-th candidate online educational resource. This represents the weight coefficient of the position of the k-th knowledge point in the j-th candidate online educational resource. This represents the total number of knowledge points covered by the j-th candidate online educational resource. This represents the number of times the m-th knowledge point appears in the j-th candidate online educational resource. This represents the weight coefficient of the position of the m-th knowledge point in the j-th candidate online educational resource.

[0100] The AI-based adaptive recommendation system for online educational resources according to the embodiments of this application can correspond to the execution of the methods described in the embodiments of this application, and the other operations and / or functions of each module / unit of the AI-based adaptive recommendation system for online educational resources are respectively for the purpose of implementing... Figure 1 For the sake of brevity, the corresponding processes of each method in the illustrated embodiments will not be described in detail here.

[0101] This application also provides a computing device. For example... Figure 3 As shown in the figure, this is a schematic diagram of a computing device provided in an embodiment of this application. The computing device 700 includes a bus 701, a processor 702, a communication interface 703, and a memory 704. The processor 702, the memory 704, and the communication interface 703 communicate with each other via the bus 701.

[0102] The 701 bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0103] The processor 702 can be any one or more of the following processors: central processing unit (CPU), graphics processing unit (GPU), microprocessor (MP), or digital signal processor (DSP).

[0104] The communication interface 703 is used for communication with external devices.

[0105] Memory 704 may include volatile memory, such as random access memory (RAM). Memory 704 may also include non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid state drive (SSD).

[0106] The memory 704 stores executable code, and the processor 702 executes the executable code to perform the aforementioned AI-based adaptive recommendation method for online educational resources.

[0107] Specifically, in achieving Figure 2 In the case of the illustrated embodiment, and Figure 2 When the modules or units of the AI-based adaptive recommendation system for online educational resources described in the embodiments are implemented in software, the following steps are performed: Figure 2 The software or program code required for the functions of each module / unit can be partially or wholly stored in the memory 704. The processor 702 executes the program code corresponding to each unit stored in the memory 704, and executes the aforementioned adaptive recommendation method for online education resources based on artificial intelligence.

[0108] This application also provides a computer-readable storage medium. The computer-readable storage medium can be any available medium that a computing device can store, or a data storage device such as a data center containing one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive). The computer-readable storage medium includes instructions that instruct the computing device to execute the aforementioned AI-based adaptive recommendation method for online educational resources.

[0109] This application also provides a computer program product comprising one or more computer instructions. When the computer instructions are loaded and executed on a computing device, all or part of the processes or functions described in this application are generated.

[0110] The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, or data center to another website, computer, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means.

[0111] When the computer program product is executed by a computer, the computer executes any of the aforementioned methods of the AI-based adaptive recommendation method for online educational resources. The computer program product can be a software installation package; when any of the aforementioned AI-based adaptive recommendation methods for online educational resources needs to be used, the computer program product can be downloaded and executed on the computer.

[0112] The descriptions of the processes or structures corresponding to the above figures each have their own emphasis. For parts of a process or structure that are not described in detail, please refer to the relevant descriptions of other processes or structures.

[0113] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be covered within the scope of protection of this application.

Claims

1. An adaptive recommendation method for online educational resources based on artificial intelligence, characterized in that, The method includes: Obtain candidate online educational resources; The importance of the k-th knowledge point in the candidate online educational resources is determined by the frequency of its occurrence; the resource type characteristics of the candidate online educational resources are determined by their resource type; and the resource behavior characteristics of the candidate online educational resources are determined by their educational behavior. Obtain the probability that a user has mastered the k-th knowledge point after the t-th learning session, as well as the user's learning type characteristics and learning behavior characteristics; The knowledge matching degree between the candidate online education resources and the user is determined by using the importance of the k-th knowledge point in the candidate online education resources and the probability of the user mastering the k-th knowledge point after the t-th learning session; the type matching degree between the candidate online education resources and the user is determined by using resource type features and learning type features; and the behavior matching degree between the candidate online education resources and the user is determined by using resource behavior features and learning behavior features. Based on the knowledge matching degree, the type matching degree, and the behavior matching degree, the overall matching degree between the candidate online education resources and the user is determined; Based on the comprehensive matching degree, the target set of online educational resources is determined; Recommend online educational resources from the target set of online educational resources to the user.

2. The method according to claim 1, characterized in that, The step of determining the knowledge matching degree between the candidate online educational resources and the user by utilizing the importance of the k-th knowledge point in the candidate online educational resources and the probability of the user mastering the k-th knowledge point after the t-th learning session includes: Based on the probability that a user has mastered the kth knowledge point after the tth learning session, determine the probability that the user has not mastered the kth knowledge point after the tth learning session. The knowledge matching degree between the candidate online education resources and the user is determined by using the importance of the k-th knowledge point in the candidate online education resources and the probability that the user has not mastered the k-th knowledge point after the t-th learning session.

3. The method according to claim 1, characterized in that, The resource type features include a first visual type score, a first auditory type score, and a first kinesthetic type score; the learning type features include a second visual type score, a second auditory type score, and a second kinesthetic type score. The step of determining the type matching degree between the candidate online education resources and the user by utilizing resource type characteristics and learning type characteristics includes: Multiply the first visual type score by the second visual type score to obtain the sub-visual type score; multiply the first auditory type score by the second auditory type score to obtain the sub-auditory type score; multiply the first kinesthetic type score by the second kinesthetic type score to obtain the sub-kinesthetic type score. The sum of the sub-visual type score, sub-auditory type score, and sub-kinesthetic type score is used as the type matching degree between the candidate online educational resource and the user.

4. The method according to claim 1, characterized in that, The resource behavior characteristics include a first fragmentation score, a first systematic score, and a first interactive score; the learning behavior characteristics include a second fragmentation score, a second systematic score, and a second interactive score; and the determination of the behavioral matching degree between the candidate online educational resources and the user using the resource behavior characteristics and learning behavior characteristics includes: Multiply the first fragmented score by the second fragmented score to obtain the sub-fragmented score; multiply the first systematic score by the second systematic score to obtain the sub-systematic score; multiply the first interactive score by the second interactive score to obtain the sub-interactive score. The sum of the sub-fragmented score, the sub-systematic score, and the sub-interactive score is used as the behavioral matching degree between the candidate online education resource and the user.

5. The method according to claim 1, characterized in that, The step of determining the overall matching degree between the candidate online education resources and the user based on the knowledge matching degree, the type matching degree, and the behavior matching degree includes: The first weight for knowledge matching, the second weight for type matching, and the third weight for behavior matching are determined. Using the first weight, the second weight, and the third weight, the knowledge matching degree, the type matching degree, and the behavior matching degree are weighted and summed to obtain the comprehensive matching degree between the candidate online education resources and the user.

6. The method according to claim 2, characterized in that, The process of obtaining the probability that a user has mastered the k-th knowledge point after the t-th learning session includes: When you answer correctly: in, This indicates that after a user completes the t-th learning session and answers correctly, the knowledge points obtained are updated. The probability of mastering Indicates the user at time t The probability of mastering knowledge point k after one learning session. This represents the probability of a user having mastered knowledge point k but answering incorrectly due to a mistake. This represents the difficulty coefficient of the corresponding question in the t-th learning session. This represents the probability that a user guesses the answer correctly even though they haven't mastered knowledge point k. This represents the user's mastery status event of knowledge point k after the t-th learning session. This represents the result event of the user's interaction with the question in the t-th learning session. A slippage event indicates that the user has mastered the knowledge point but answered incorrectly. This refers to a guessing event in which a user answered a question correctly even though they did not understand the knowledge point. When answering incorrectly: in, This represents the probability of a user mastering knowledge point k after completing the t-th learning session and answering incorrectly.

7. The method according to claim 1, characterized in that, The method of determining the importance of the k-th knowledge point in a candidate online educational resource based on the frequency of its occurrence includes: in, This indicates the importance of the k-th knowledge point in the j-th candidate online educational resource. This represents the number of times the k-th knowledge point appears in the j-th candidate online educational resource. This represents the weight coefficient of the position of the k-th knowledge point in the j-th candidate online educational resource. This represents the total number of knowledge points covered by the j-th candidate online educational resource. This represents the number of times the m-th knowledge point appears in the j-th candidate online educational resource. This represents the weight coefficient of the position of the m-th knowledge point in the j-th candidate online educational resource.

8. An adaptive recommendation system for online educational resources based on artificial intelligence, characterized in that, The system includes: The acquisition module is used to acquire candidate online educational resources; The processing module is used to determine the importance of the kth knowledge point in the candidate online education resources based on the frequency of its occurrence; to determine the resource type characteristics of the candidate online education resources based on their resource type; and to determine the resource behavior characteristics of the candidate online education resources based on their educational behavior. The acquisition module is used to acquire the probability of a user mastering the k-th knowledge point after the t-th learning session, as well as the user's learning type characteristics and learning behavior characteristics. The processing module is used to determine the knowledge matching degree between the candidate online educational resources and the user by utilizing the importance of the k-th knowledge point in the candidate online educational resources and the probability of the user mastering the k-th knowledge point after the t-th learning session; to determine the type matching degree between the candidate online educational resources and the user by utilizing resource type features and learning type features; to determine the behavior matching degree between the candidate online educational resources and the user by utilizing resource behavior features and learning behavior features; to determine the comprehensive matching degree between the candidate online educational resources and the user based on the knowledge matching degree, the type matching degree, and the behavior matching degree; and to determine the target online educational resource set based on the comprehensive matching degree. The recommendation module is used to recommend online educational resources from the target set of online educational resources to the user.

Images