Course recommendation method, device and equipment of online learning platform and storage medium

Abstract

The application provides a course recommendation method and device of an online learning platform, equipment and a storage medium. The method comprises the following steps: obtaining attribute information set and role category set of a user from a database of the online learning platform; extracting attributes related to user role category information from the attribute information set of the user to generate role-related attribute information set of the user; obtaining role category information when the role category information of a target user is not missing; obtaining attribute information of the target user when the role category is missing; determining the role category information of the target user by using a Bayesian classification method according to the attribute information of the target user, the role category set of the user and the role-related attribute information set of the user; and filtering candidate course data according to the role category information of the target user, and recommending the candidate course data to the target user. The method improves the accuracy and efficiency of course recommendation.

Description

Methods, devices, equipment, and storage media for recommending courses on online learning platforms Technical Field

[0001] This application relates to the field of network technology, and in particular to a course recommendation method, apparatus, device, and storage medium for an online learning platform. Background Technology

[0002] In the process of providing training and learning services to users, precise learning recommendations are a key aspect of online learning platforms. To accurately deliver course content that matches users' learning needs, users' personal attributes and historical learning information are crucial reference data.

[0003] In order to recommend suitable courses to target users, existing technologies use a method of predicting target courses based on users' natural attribute information and historical learning information.

[0004] However, incomplete natural attribute information entered by users when logging in, or the lack of historical learning information for newly logged-in users, can reduce the accuracy of course recommendations. Summary of the Invention

[0005] This application provides a course recommendation method, apparatus, device, and storage medium for an online learning platform to address the problems of low efficiency and low accuracy in user job determination.

[0006] Firstly, this application provides a course recommendation method for an online learning platform, including:

[0007] The system retrieves user attribute information sets and role category sets from the online learning platform database. The user attribute information set includes multiple attribute information categories, and the user role category set includes multiple role category information. The attribute information includes: natural attribute information excluding the role category information and historical learning information.

[0008] Extract attributes related to user role category information from the user's attribute information set to generate a user role-related attribute information set;

[0009] When the target user's role category information is not missing, the target user's role category information is obtained, and candidate course data is filtered from the online learning platform database based on the target user's role category information, and the candidate course data is recommended to the target user.

[0010] When the target user role category is missing, the attribute information of the target user is obtained. Based on the target user's attribute information, the user's role category set, and the user's role-related attribute information set, the target user's role category information is determined using Bayesian classification. Based on the target user's role category information, candidate course data is filtered from the online learning platform database, and the candidate course data is recommended to the target user.

[0011] Secondly, this application provides a device for determining the user's job position on an online learning platform, comprising:

[0012] The acquisition module is used to acquire a user's attribute information set and role category set from the online learning platform database. The attribute information category includes multiple sub-attribute information, and the user's role category set includes multiple role category information. The attribute information includes: natural attribute information excluding the role category information and historical learning information;

[0013] The generation module is used to extract attributes related to user role category information from the user's attribute information set and generate a user role-related attribute information set;

[0014] The recommendation module is used to obtain the target user's role category information when the target user's role category information is not missing, filter candidate course data from the online learning platform database based on the target user's role category information, and recommend the candidate course data to the target user. When the target user's role category is missing, the module obtains the target user's attribute information, determines the target user's role category information using Bayesian classification based on the target user's attribute information, the user's role category set, and the user's role-related attribute information set, and filters candidate course data from the online learning platform database based on the target user's role category information, and recommends the candidate course data to the target user.

[0015] Thirdly, this application provides a device for determining the job position of an online learning platform user, comprising:

[0016] Processor, memory, communication interface;

[0017] The memory is used to store the executable instructions of the processor;

[0018] The processor is configured to execute the course recommendation method of the online learning platform described in the first aspect above by executing the executable instructions.

[0019] Fourthly, this application provides a readable storage medium, comprising: a computer program stored thereon, wherein the computer program, when executed by a processor, implements the course recommendation method of the online learning platform described in the first aspect above.

[0020] The online learning platform course recommendation method, apparatus, device, and storage medium provided in this application obtain user attribute information sets and role category sets from the online learning platform database. The user attribute information set includes multiple attribute information categories, and the user role category set includes multiple role category information. The attribute information includes: natural attribute information excluding role category information and historical learning information. Attributes related to the user's role category information are extracted from the user attribute information set to generate a user role-related attribute information set. When the target user's role category information is not missing, the target user's role category information is obtained. Based on the target user's role category information, candidate course data is filtered from the online learning platform database, and the candidate course data is recommended to the target user. When the target user's role category is missing, the target user's attribute information is obtained. Based on the target user's attribute information, the user's role category set, and the user role-related attribute information set, the target user's role category information is determined using Bayesian classification. Based on the target user's role category information, candidate course data is filtered from the online learning platform database, and the candidate course data is recommended to the target user. Among them, user role category information refers to the basic attribute information of users that plays a representative role in the selection of platform courses. This application recommends courses to target users by using the role category information of users whose role types are not missing, thereby improving the accuracy of course recommendation. This application uses Bayesian classification to determine the role category information of users whose role types are missing, and further recommends courses to target users, thereby improving the accuracy and efficiency of course recommendation. Attached Figure Description

[0021] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0022] Figure 1 is a flowchart illustrating the course recommendation method of the online learning platform provided in this embodiment of the application;

[0023] Figure 2 is a schematic diagram of the process of extracting attributes related to user role category information from the user's attribute information set and generating a user role-related attribute information set according to an embodiment of this application;

[0024] Figure 3 is a flowchart illustrating the process of determining the target user's role category information using Bayesian classification based on the target user's attribute information, the user's role category set, and the user's role-related attribute information set, according to an embodiment of this application.

[0025] Figure 4 is a schematic diagram of the structure of a course recommendation device for an online learning platform provided in an embodiment of this application;

[0026] Figure 5 is a schematic diagram of the structure of a course recommendation device for another online learning platform provided in an embodiment of this application;

[0027] Figure 6 is a schematic diagram of the structure of a course recommendation device for an online learning platform provided in an embodiment of this application.

[0028] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concepts of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0029] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0030] Existing technologies recommend courses to target users by estimating target courses based on users' natural attribute information and historical learning information. However, problems such as incomplete natural attribute information entered by users when logging in, or the lack of historical learning information for newly logged-in users, can reduce the accuracy of course recommendations.

[0031] This application obtains user attribute information sets and role category sets from an online learning platform database. The user attribute information set includes multiple attribute information categories, and the user role category set includes multiple role category information. Attribute information includes: natural attribute information excluding role category information and historical learning information. Attributes related to the user's role category information are extracted from the user attribute information set to generate a user role-related attribute information set. When the target user's role category information is not missing, the target user's role category information is obtained. Based on the target user's role category information, candidate course data is filtered from the online learning platform database and recommended to the target user. When the target user's role category is missing, the target user's attribute information is obtained. Based on the target user's attribute information, the user's role category set, and the user role-related attribute information set, the target user's role category information is determined using Bayesian classification. Based on the target user's role category information, candidate course data is filtered from the online learning platform database and recommended to the target user. Among them, user role category information refers to the basic attribute information of users that plays a representative role in the selection of platform courses. This application recommends courses to target users by using the role category information of users whose role types are not missing, thereby improving the accuracy of course recommendation. This application uses Bayesian classification to determine the role category information of users whose role types are missing, and further recommends courses to target users, thereby improving the accuracy and efficiency of course recommendation.

[0032] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0033] Figure 1 is a flowchart illustrating the course recommendation method of the online learning platform provided in the first embodiment of this application.

[0034] As shown in Figure 1, the course recommendation method of the online learning platform in this embodiment may include the following steps:

[0035] Step S101: Obtain the user's attribute information set and role category set from the online learning platform database. The user's attribute information set includes multiple attribute information categories, and the user's role category set includes multiple role category information. The attribute information includes: natural attribute information excluding role category information and historical learning information.

[0036] Specifically, the online learning platform database stores the user's natural attribute information entered when logging into the platform, as well as historical learning information saved during the user's past use of the platform. Natural attribute information refers to the user's basic attribute information. For example, for a user of an enterprise learning platform, the user's natural attribute information, i.e., basic attribute information, may include: user ID, user age, user gender, user department, user position, and user's major. Historical learning information refers to the user's historical course learning information during their past use of the platform. For example, for a user of an enterprise learning platform, the user's historical learning information may include: the number of times the user has studied, the duration of their study, and the courses they have studied. For users who log into the online learning platform but have not started learning, the database only stores their natural attribute information.

[0037] Among the various natural attribute information of users, there is user role category information. User role category information refers to the basic attribute information of users that plays a representative role in the user's choice of courses on the platform. For example, for users of enterprise learning platforms, their role category information can be user job information.

[0038] Specifically, user's natural attribute information and historical learning information can be obtained from the aforementioned learning platform to generate online learning platform users, namely, user attribute information set B and role category set C. The attribute information includes: natural attribute information excluding role category information and historical learning information. The user attribute information set includes multiple attribute information categories. For example, for the aforementioned enterprise learning platform, the user attribute information set includes multiple attribute information categories, such as user age category, user gender category, etc. The user role category set includes multiple role category information. For example, for the aforementioned enterprise learning platform, the user role category set includes multiple role category information, such as operations and maintenance position, human resources position, etc.

[0039] Step S102: Extract attributes related to user role category information from the user's attribute information set to generate a user role-related attribute information set.

[0040] As described in step S101, user role category information refers to basic user attribute information that plays a representative role in the user's selection of platform courses. Specifically, attributes related to user role category information can be extracted from the user attribute information set B obtained in step S101 to generate user role-related attribute information set A.

[0041] In this context, the elements in user role-related attribute information set A are attribute information set elements of the user associated with role category information. Specifically, in the process of filtering elements in user attribute information set B to generate user role-related attribute information set A, various methods can be used. For example, in user attribute information set B, elements that can be directly judged to have a high correlation with the user's user role category information, such as the learning courses in the aforementioned enterprise learning platform, which are highly correlated with the user's role category information, can be directly processed and set as elements in user role-related attribute information set A. Conversely, in user attribute information set B, elements whose correlation with role category information cannot be directly judged can be determined using parameters such as itemset confidence to determine whether the element can be set as an element in user role-related attribute information set A.

[0042] Specifically, the elements of the role-related attribute information set A as described above can be combined to form the user's role-related attribute information set A.

[0043] Step S103: When the target user's role category information is not missing, obtain the target user's role category information, filter candidate course data from the online learning platform database based on the target user's role category information, and recommend the candidate course data to the target user.

[0044] Specifically, as described in step S101, user role category information refers to basic user attribute information that plays a representative role in the user's selection of platform courses. Therefore, the course recommendation method of the online learning platform provided in this embodiment can filter candidate course data from the above-mentioned online learning platform database based on the target user's role category information to improve the accuracy of platform course recommendation.

[0045] The target users refer to those currently logged into the learning platform (including historical students and first-time users) who need the learning platform to recommend courses based on their user characteristics. Target users can include users whose role category information is not missing and users whose role category information is missing. Users whose role category information is missing refer to users who did not fill in role category information when filling in their natural attribute information.

[0046] Specifically, when the target user's role category information is not missing, the target user's role category information can be obtained, and candidate course data can be filtered from the online learning platform database based on the target user's role category information, and the candidate course data can be recommended to the target user.

[0047] Step S104: When the target user's role category is missing, obtain the target user's attribute information. Based on the target user's attribute information, the user's role category set, and the user's role-related attribute information set, determine the target user's role category information using Bayesian classification. Based on the target user's role category information, filter candidate course data from the online learning platform database and recommend the candidate course data to the target user.

[0048] Specifically, when the target user's role category is missing, the target user's attribute information can be obtained. Based on the target user's attribute information, the user's role category set obtained in step S101, and the user role-related attribute information set generated in step S103, the target user's role category information is determined by Bayesian classification. Then, based on the target user's role category information, candidate course data is filtered from the online learning platform database and recommended to the target user.

[0049] As described in step S101, the attribute information includes: natural attribute information excluding role category information and historical learning information. Specifically, target users with missing role categories can include: users with course learning history and users without course learning history. Users without course learning history include: students logging in for the first time and users who have logged in previously but have not taken any courses. Specifically, for users with course learning history, the target user's attribute information can include: natural attribute information excluding role category information and historical learning information; for users without course learning history, the target user's attribute information can include: natural attribute information excluding role category information.

[0050] The course recommendation method for an online learning platform provided in this embodiment obtains a user's attribute information set and role category set from the online learning platform database. The user's attribute information set includes multiple attribute information categories, and the user's role category set includes multiple role category information. The attribute information includes: natural attribute information excluding role category information and historical learning information. Attributes related to the user's role category information are extracted from the user's attribute information set to generate a user's role-related attribute information set. When the target user's role category information is not missing, the target user's role category information is obtained. Based on the target user's role category information, candidate course data is filtered from the online learning platform database, and the candidate course data is recommended to the target user. When the target user's role category is missing, the target user's attribute information is obtained. Based on the target user's attribute information, the user's role category set, and the user's role-related attribute information set, the target user's role category information is determined using Bayesian classification. Based on the target user's role category information, candidate course data is filtered from the online learning platform database, and the candidate course data is recommended to the target user. Among them, user role category information refers to the basic attribute information of users that plays a representative role in the selection of platform courses. This application recommends courses to target users by using the role category information of users whose role types are not missing, thereby improving the accuracy of course recommendation. This application uses Bayesian classification to determine the role category information of users whose role types are missing, and further recommends courses to target users, thereby improving the accuracy and efficiency of course recommendation.

[0051] Figure 2 is a flowchart illustrating the process of extracting attributes related to user role category information from the user's attribute information set and generating a user role-related attribute information set according to the second embodiment of this application. Based on the embodiment shown in Figure 1, this embodiment elaborates on the process of extracting attributes related to user role category information from the user's attribute information set and generating a user role-related attribute information set.

[0052] As shown in Figure 2, this embodiment can generate a user role-related attribute information set by extracting attributes related to user role category information from the user attribute information set, which may include the following steps:

[0053] Step S201: Directly extract the first attribute element group of the user role-related attribute information set from the user's attribute information set.

[0054] Specifically, as described in step S102, the elements in the user's role-related attribute information set A are attribute information set elements of the user that are associated with the role category information. Among them, in the user's attribute information set B, elements that can be directly determined to be highly related to the user's role category information can be directly extracted, processed, and set as elements in the user's role-related attribute information set A.

[0055] Specifically, elements in the user's attribute information set B can be directly evaluated. If an element in the user's attribute information set B is highly correlated with the role category—for example, in the aforementioned enterprise learning platform, the user's learning courses in the user's attribute information set B are highly correlated with the role category—then the element of that user's attribute information set, i.e., the user's learning courses, can be directly set as an element of the user's role-related attribute information set A. This type of element is the first attribute element of the user's role-related attribute information set A. Specifically, the first attribute element of the user's role-related attribute information set can be determined based on the user's attribute information set, and all the first attribute elements of the user's role-related attribute information set A can be merged together to form the first attribute element group of the user's role-related attribute information set A, which is then used to construct the user's role-related attribute information set A.

[0056] Step S202: Obtain the natural attribute information and historical learning information of multiple users. Based on the natural attribute information of multiple users, the historical learning information of multiple users, the attribute set of users, and the role category set of users, extract the second attribute element group of the user role-related attribute information set from the user attribute information set.

[0057] Specifically, as described in step S103, the elements in the user's role-related attribute information set A are the user's natural attributes associated with role category information. In the user's attribute information set B, for elements whose association with role category information cannot be directly determined, parameters such as itemset confidence can be used to determine whether the element can be set as an element in the user's role-related attribute information set A.

[0058] Specifically, the natural attribute information and historical learning information of multiple users can be obtained from the online learning platform database. Based on the natural attribute information of the multiple users, the historical learning information of the multiple users, the attribute information set of users obtained in step S101, and the user role category set, the second attribute element group of the user role-related attribute information set is extracted from the user attribute information set.

[0059] Specifically, a transaction set can be constructed based on the natural attribute information and historical learning information of multiple users. The transaction set is a collection of all information of multiple users.

[0060] Specifically, we can filter the natural attribute information and role category information of multiple users whose role category information is not missing from the online learning platform database. Each piece of information from these multiple users' natural attribute information and role category information can be called a transaction, and the set of all transactions is called the transaction set T.

[0061] Multiple items can be constructed based on the user's attribute information set, with different items used to mark different sub-attribute information under the same attribute information category.

[0062] Specifically, the attribute information category includes multiple sub-attribute information. For example, the user gender category includes two sub-attribute information: male and female.

[0063] Specifically, based on the user's attribute information set obtained in step S101, all items of the user's attribute information set can be constructed. Different items are used to mark different sub-attribute information under the same attribute information category. For example, an element can be selected from the user's attribute information set, namely attribute information category B. n - User's learning period; determine attribute information category B n All sub-attribute information, i.e., attribute values, such as time periods like morning (8:00-12:00), afternoon (12:00-18:00), and evening (18:00-22:00); the above attribute information can be categorized into B. n All sub-attribute information, i.e., attribute values, are used as attribute information category B. n The term is denoted as Y. j Where Y1 = 8:00-12:00, Y2 = 12:00-18:00, ..., j = 1, 2 ... J, and J is the attribute information category B. n The number of items, and attribute information category B n The number of sub-attribute information corresponds to this. Specifically, based on the above method, all elements in the user's attribute information set, i.e., the items of the attribute category, can be constructed.

[0064] Multiple itemsets can be constructed based on the user's role category set and multiple items.

[0065] Specifically, based on all items in the user attribute information set constructed above and the user role category set obtained in step S101, a complete itemset of the role category set can be constructed. For example, for an enterprise learning platform, an element can be selected from the role category set, namely role category C. m - Operations and Maintenance Position; The items in the attribute categories constructed above can be matched with the role category C. m Correspondingly, construct role category C. m All itemsets, such as: Role Category C m With attribute category B n The itemset constructed by item Y1 is {Operations and Maintenance Position, Learning Time - Morning}, and the role category is C. m With attribute category B n The itemset constructed by item Y2 is {Operations and Maintenance Position, Learning Time - Afternoon}. Specifically, based on the method described above, all elements in the role category set can be constructed, i.e., the itemset of the role category.

[0066] Among them, the confidence level of multiple itemsets can be calculated based on the transaction set and multiple itemsets.

[0067] Specifically, based on the transaction set and all itemsets of the role category set constructed in the above process, the confidence level *c* of each itemset—that is, the itemset corresponding to each role category and attribute category—can be calculated. For example: Role category C m With attribute category B n Item Y j The expression for the corresponding itemset confidence level c is:

[0068]

[0069] Specifically, c = P(Y) j |C m (For role category C) m With attribute category B n Item Y j The corresponding itemset confidence level indicates that the transaction set T contains sub-attribute information Y. j The transaction also includes role category C m The percentage of transactions, s = P(C m ∪Y j (For role category C) m With attribute category B n Item Y j The corresponding support represents the sub-attribute information Y. j and role category C m The percentages of both in transaction set T, P(C m ) indicates role category C m Percentage in transaction set T.

[0070] Specifically, the confidence scores of all itemsets corresponding to all character categories and all attribute categories can be calculated using the method described above.

[0071] Among them, the second attribute element group of the user role-related attribute information set can be extracted from the user's attribute information set based on the confidence of multiple itemsets.

[0072] Specifically, based on the confidence scores of all itemsets corresponding to all role categories and all attribute categories calculated above, the second attribute element group of the user role-related attribute information set can be extracted from the user's attribute information set.

[0073] Specifically, confidence thresholds for multiple itemsets can be set based on the user's attribute set.

[0074] Specifically, based on all itemsets of the role category set constructed in the above process, a confidence threshold c0 can be set for each itemet corresponding to an attribute category. For example: Attribute category B n The expression for the corresponding itemset confidence threshold c0 is:

[0075]

[0076] Specifically, J represents attribute type B. n The number of items, for example: when attribute category B n - When the user's learning period is 4, J = 4, ∝ is the adjustment coefficient, which can be adjusted according to the platform's needs, ∝ ≤ J. For example, when ∝ = 3, it means that the confidence level needs to reach three times the average probability.

[0077] Specifically, when the confidence level of an itemset is greater than the confidence level threshold of the itemset corresponding to the confidence level of the itemset, the attribute information category corresponding to the confidence level of the itemset in the user attribute set is set as the second attribute element of the user role-related attribute information set.

[0078] Specifically, based on the itemset confidence threshold c0 corresponding to the attribute category and the itemset confidence c corresponding to each item in the attribute category, it can be determined whether a strong correlation exists between the role category and the attribute category. That is, whether an element of the user's attribute set can be set as the second attribute element of the user's role-related attribute information set. Specifically, if the itemset confidence c corresponding to any item in the job category and the attribute category is greater than the itemset confidence threshold c0 corresponding to the attribute category, a strong correlation can be considered between the role category and the attribute category. In this case, the element of that attribute category, i.e., the user's attribute set, can be set as the second attribute element value of the user's role-related attribute information set. For example: Determining the role category C... m and attribute category B n Whether there is a strong correlation between them can be determined by comparing role categories C. m With attribute category B n Item Y j The corresponding itemset confidence level c, and the attribute category B n The relationship between the corresponding itemset confidence thresholds c0, for j = 1, 2, ..., J, if there exists any itemset with a confidence greater than the threshold, i.e., P(Y) j |C m If c > c0, then the role category is considered to be C. m and attribute category B n There is a strong correlation between them. Therefore, attribute category B... n This is the second attribute element of the user's role-related attribute information set A.

[0079] Specifically, the second attribute elements of all user role-related attribute information sets can be merged to form a group of second attribute elements of the user role-related attribute information set.

[0080] Step S203: Merge the first attribute element group and the second attribute element group to generate a set of user role related attribute information.

[0081] Specifically, the first attribute element group formed in step S201 and the second attribute element group formed in step S202 can be merged to generate a user role-related attribute information set A.

[0082] This embodiment provides a process for extracting attributes related to user role category information from a user's attribute information set to generate a user role-related attribute information set. The process involves directly extracting a first attribute element group from the user's attribute information set, obtaining natural attribute information and historical learning information from multiple users, and then extracting a second attribute element group from the user's attribute information set based on the natural attribute information, historical learning information, user attribute set, and user role category set. The first and second attribute element groups are then merged to generate the user role-related attribute information set. Different methods are used to confirm each element in the first and second attribute element groups, improving the accuracy of the user role-related attribute information set.

[0083] Figure 3 is a flowchart illustrating the process of determining the target user's role category information using Bayesian classification based on the target user's attribute information, the user's role category set, and the user's role-related attribute information set, according to the third embodiment of this application. Based on the embodiment shown in Figure 1, this embodiment elaborates on the process of determining the target user's role category information using Bayesian classification based on the target user's attribute information, the user's role category set, and the user's role-related attribute information set.

[0084] As shown in Figure 3, the determination of the target user's role category information using Bayesian classification based on the target user's attribute information, the user's role category set, and the user's role-related attribute information set in this embodiment may include the following steps:

[0085] Step S301: Determine the target user's role-related attribute information tuple based on the user role-related attribute information set and the target user's attribute information.

[0086] Specifically, the target user's role-related attribute information tuple X can be confirmed based on the user's role-related attribute information set A confirmed in step S102 and the target user's attribute information obtained in step S104.

[0087] Here, the target user's role-related attribute information tuple X refers to the set of natural attribute information of the target user corresponding to the user's role-related attribute information set A. Specifically, the natural attribute information of the target user can be filtered based on the elements of the user's role-related attribute information set A. For example, when the element of the user's role-related attribute information set A confirmed in step S103 includes the element of the user's department, the user's department information can be filtered from the natural attribute information of the target user obtained in step S104 and used as an element of the target user's role-related attribute information tuple X.

[0088] Optionally, the target user's natural attribute information may be incomplete or incomplete. Therefore, the target user's role-related attribute information tuple determined based on the filtered target user's natural attribute information and the user's role-related attribute information set may have fewer elements than the number of elements in the user's role-related attribute information set. The number of elements does not affect the determination of the target user's role category information.

[0089] Step S302: Based on the user's role category set and the target user's role-related attribute information tuple, analyze the target user's role category using Bayesian classification to determine the target user's role category information.

[0090] Specifically, the role category set C determined in step S101 and the role-related attribute information tuple X of the target user determined in step S301 can be used to analyze the role category of the target user and determine the role category information of the target user.

[0091] Bayesian classification is a statistical classification method that utilizes probability and statistics to perform classification. In many situations, Bayesian classification is comparable to decision trees and neural network classification algorithms. It can be applied to large databases and is simple, accurate, and fast. Bayesian classification can predict the probability of class membership, such as the probability that a given tuple belongs to a specific class.

[0092] Specifically, the probability of the target user's role-related attribute information tuple X belonging to different user role categories can be calculated according to the Bayesian classification algorithm described above. That is, the probability of each user role category in the role category set corresponding to the target user's role-related attribute information tuple X is determined. The target user's job position can be confirmed based on the probability value of each user role category.

[0093] Specifically, the posterior probability of each role category can be calculated using Bayesian classification based on the user's role category set and the target user's role-related attribute information tuple.

[0094] Specifically, based on the set of role categories C determined in step S101 and the target user role-related attribute information tuple X determined in step S301, the posterior probability of each user role can be calculated using the Bayesian classification method described in step S301, that is, the probability that the target user's role-related attribute information tuple X belongs to different user role categories.

[0095] For example: For a specific target user, the role-related attribute information tuple X of the target user can be determined according to step S301. i X i ={x1,x2,…,x N}, i = 1, 2, ..., N. Where, user role category C m The posterior probability, i.e., the tuple X containing the target user's role-related attribute information. i Belongs to user role category C m The expression for the probability is:

[0096]

[0097] Specifically, P(C m |X i The value of ) can be calculated based on the dataset D constructed from the transaction set T described in step S202. Wherein, P(X) i P(C) is a constant for all user role categories; m )=|C m,D | / |D|, where |C m,D | is D in C m The number of training tuples for class D, where |D| is the number of training tuples for class D;

[0098] Among them, since the user's role-related attribute information set A generally contains many attributes, namely the target user's role-related attribute information tuple X i A molecule typically has multiple elements, therefore calculating P(X) requires... i |C m The overhead of calculating P(X) is very high. To reduce the computational cost of P(X)... i |C m The cost of P(X) can be assumed to be that there are no dependencies between attributes and that attribute values ​​are conditionally independent, i.e., P(X) i |C m The expression for ) is:

[0099]

[0100] Where x n This indicates that tuple Xi is in attribute A n The value of P(x). n |Cm ) is attribute A in D n The value is x n C m The number of tuples in class divided by C in D m Number of tuples in a class | C m,D |

[0101] Specifically, based on the algorithm described above, the posterior probability of each user role category information of the target user can be calculated, i.e., the tuple X of the target user's role-related attribute information. i The probability of belonging to different user role categories.

[0102] Specifically, the target user's role category information can be determined based on the posterior probability of each role category.

[0103] Specifically, the posterior probabilities of each user role category information of the target user calculated above, that is, the probabilities of the target user's role-related attribute information tuple X belonging to different user role categories, are compared. The user role category information corresponding to the highest posterior probability is the target user's role category information.

[0104] This embodiment provides a method for determining the target user's role-related attribute information tuple based on the target user's attribute information, user role category set, and user role-related attribute information set. It then uses Bayesian classification to analyze the target user's role category information, based on the user role category set and the target user's role-related attribute information tuple. The use of Bayesian classification to calculate the posterior probability of each user's role improves the accuracy and efficiency of role determination.

[0105] Figure 4 is a schematic diagram of the structure of a course recommendation device for an online learning platform provided in the fourth embodiment of this application.

[0106] As shown in Figure 4, the course recommendation device 40 of the online learning platform in this embodiment includes an acquisition module 41, a generation module 42, and a recommendation module 43.

[0107] Module 41 is used to retrieve the user's attribute information set and role category set from the online learning platform database. The user's attribute information set includes multiple attribute information categories, and each attribute information category includes multiple sub-attribute information. The user's role category set includes multiple role category information. The attribute information includes: natural attribute information excluding role category information and historical learning information.

[0108] The generation module 42 is used to extract attributes related to user role category information from the user's attribute information set and generate a user role-related attribute information set.

[0109] Recommendation module 43 is used to obtain the target user's role category information when the target user's role category information is not missing, filter candidate course data from the online learning platform database based on the target user's role category information, and recommend the candidate course data to the target user. When the target user's role category is missing, it obtains the target user's attribute information, determines the target user's role category information using Bayesian classification based on the target user's attribute information, the user's role category set, and the user's role-related attribute information set, and filters candidate course data from the online learning platform database based on the target user's role category information, and recommends the candidate course data to the target user.

[0110] Optionally, the recommendation module 43 includes: a first recommendation unit and a second recommendation unit.

[0111] The first recommendation unit 413 is used to obtain the target user's role category information when the target user's role category information is not missing, and to filter candidate course data from the online learning platform database based on the target user's role category information, and recommend the candidate course data to the target user.

[0112] The second recommendation unit 423 is used to obtain the target user's attribute information when the target user's role category is missing. Based on the target user's attribute information, the user's role category set, and the user's role-related attribute information set, it determines the target user's role category information using Bayesian classification. Based on the target user's role category information, it filters candidate course data from the online learning platform database and recommends the candidate course data to the target user. Figure 5 is a schematic diagram of the structure of another course recommendation device for an online learning platform provided in an embodiment of this application.

[0113] The apparatus provided in this embodiment can be used to execute the technical solutions of Figures 1 to 3 of the above method embodiments. Its implementation principle and technical effect are similar, and will not be described again here.

[0114] Figure 6 is a schematic diagram of the structure of a course recommendation device for an online learning platform provided in the fifth embodiment of this application.

[0115] As shown in Figure 6, the course recommendation device 60 of the online learning platform in this embodiment includes: a processor 61, a memory 62, and a communication interface 63.

[0116] Memory 62 is used to store the processor's executable instructions;

[0117] The processor 61 is configured to execute the course recommendation method of any one of the online learning platforms in Figures 1 to 3 of the above method embodiments by executing executable instructions.

[0118] This application also provides a readable storage medium storing a computer program thereon, wherein when the computer program is executed by a processor, it implements the course recommendation method of the online learning platform according to any one of Figures 1 to 3 of the above-described method embodiments.

[0119] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals are provided for users to choose to authorize or refuse.

[0120] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0121] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A course recommendation method for an online learning platform, characterized in that, include: The system retrieves user attribute information sets and role category sets from the online learning platform database. The user attribute information set includes multiple attribute categories, and the user role category set includes multiple role category information. The attribute information includes natural attribute information excluding role category information and historical learning information. Attributes related to the user's role category information are extracted from the user attribute information set to generate a user role-related attribute information set. When the target user's role category information is not missing, the target user's role category information is retrieved. Based on the target user's role category information, candidate course data is filtered from the online learning platform database, and the candidate courses are selected... The system recommends course data to the target user; when the target user's role category is missing, it obtains the target user's attribute information, determines the target user's role-related attribute information tuple based on the user's role-related attribute information set and the target user's attribute information; calculates the posterior probability of each role category using Bayesian classification based on the user's role category set and the target user's role-related attribute information tuple; determines the target user's role category information based on the posterior probability of each role category; and filters candidate course data from the online learning platform database based on the target user's role category information, recommending the candidate course data to the target user.

2. The method according to claim 1, characterized in that, The step of extracting attributes related to user role category information from the user's attribute information set to generate a user role-related attribute information set includes: directly extracting a first attribute element group of the user role-related attribute information set from the user's attribute information set; obtaining natural attribute information and historical learning information of multiple users; extracting a second attribute element group of the user role-related attribute information set from the user's attribute information set based on the natural attribute information of the multiple users, the historical learning information of the multiple users, the user's attribute set, and the user's role category set; and merging the first attribute element group and the second attribute element group to generate the user role-related attribute information set.

3. The method according to claim 2, characterized in that, The attribute information category includes multiple sub-attribute information. The step of extracting the second attribute element group of the user role-related attribute information set from the user's attribute information set based on the user's natural attribute information, the user's historical learning information, the user's attribute set, and the user's role category set includes: constructing a transaction set based on the user's natural attribute information and the user's historical learning information, where the transaction set is a collection of all information for the user; constructing multiple items based on the user's attribute set, where different items are used to label different sub-attribute information under the same attribute information category; constructing multiple itemsets based on the user's role category set and the multiple itemsets; calculating the confidence level of the multiple itemsets based on the transaction set and the multiple itemsets; and extracting the second attribute element group of the user role-related attribute information set from the user's attribute information set based on the confidence level of the multiple itemsets.

4. The method according to claim 3, characterized in that, The step of extracting a second attribute element group of the user role-related attribute information set from the user's attribute information set based on the confidence levels of multiple itemsets includes: setting multiple confidence thresholds for itemsets based on the user's attribute information set; when the confidence level of an itemset is greater than the confidence threshold corresponding to the confidence level of the itemset, setting the attribute information category corresponding to the confidence level of the itemset in the user attribute set as a second attribute element of the user role-related attribute information set; and merging all the second attribute elements of the user role-related attribute information sets to form a second attribute element group of the user role-related attribute information set.

5. A course recommendation device for an online learning platform, characterized in that, include: The acquisition module is used to acquire a user's attribute information set and role category set from the online learning platform database. The user's attribute information set includes multiple attribute information categories, and the user's role category set includes multiple role category information. The attribute information includes: natural attribute information excluding the role category information and historical learning information. The generation module is used to extract attributes related to the user's role category information from the user's attribute information set to generate a user's role-related attribute information set. The recommendation module is used to acquire the target user's role category information when the target user's role category information is not missing, and to filter candidates from the online learning platform database based on the target user's role category information. The system retrieves course data and recommends candidate course data to the target user. When the target user's role category is missing, it obtains the target user's attribute information and determines the target user's role-related attribute information tuple based on the user's role-related attribute information set and the target user's attribute information. Based on the user's role category set and the target user's role-related attribute information tuple, it calculates the posterior probability of each role category using Bayesian classification. Based on the posterior probability of each role category, it determines the target user's role category information. Based on the target user's role category information, it filters candidate course data from the online learning platform database and recommends the candidate course data to the target user.

6. The apparatus according to claim 5, characterized in that, The recommendation module includes: a first recommendation unit, configured to, when the target user's role category information is not missing, obtain the target user's role category information, filter candidate course data from the online learning platform database based on the target user's role category information, and recommend the candidate course data to the target user; and a second recommendation unit, configured to, when the target user's role category is missing, obtain the target user's attribute information, determine the target user's role category information using Bayesian classification based on the target user's attribute information, the user's role category set, and the user's role-related attribute information set, filter candidate course data from the online learning platform database based on the target user's role category information, and recommend the candidate course data to the target user.

7. A course recommendation device for an online learning platform, characterized in that, include: Processor, memory, communication interface; The memory is used to store executable instructions of the processor; wherein the processor is configured to execute the course recommendation method of the online learning platform according to any one of claims 1 to 4 by executing the executable instructions.

8. A readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the course recommendation method of the online learning platform according to any one of claims 1 to 4.

Images