CN116541711A

CN116541711A - Model training method, course recommendation method, device, equipment and medium

Info

Publication number: CN116541711A
Application number: CN202310623353.6A
Authority: CN
Inventors: 李素粉; 赵健东
Original assignee: China United Network Communications Group Co Ltd
Current assignee: China United Network Communications Group Co Ltd
Priority date: 2023-05-30
Filing date: 2023-05-30
Publication date: 2023-08-04

Abstract

The application provides a model training method, a course recommendation method, a device, equipment and a medium. The method comprises the following steps: acquiring an initial training sample, wherein the initial training sample is a sample acquired from historical data information of a learning platform, and comprises courses learned by students and students; determining a target course in the courses, learned information of the target course and learning attribute information of a target learner and a target learner in the learner according to the initial training sample; according to the target specialty category and the course attribute category, determining the association degree of the target specialty category and the history learning course; optimizing the initial training sample according to the course quality and the association degree to obtain an optimized training sample; and training the model to be trained according to the optimized training sample to obtain a target training model. According to the method, the training effect of the model to be trained is improved.

Description

Model training method, course recommendation method, device, equipment and medium

Technical Field

The application relates to the technical field of course recommendation, in particular to a model training method, a course recommendation method, a device, equipment and a medium.

Background

The courses which are interested by the students are recommended through the learning platform, so that the students can find the courses which are required to learn more efficiently, information overload is reduced, and the method is a common method in the current society of information burst.

At present, when a learning platform in the prior art carries out course recommendation on a learner, the course recommendation is generally carried out on the learner by collecting information of the learner through a course recommendation model which is internally arranged.

However, the existing recommendation model has the problem that the accuracy of recommended courses is low due to poor model training effect.

Disclosure of Invention

The application provides a model training method, a course recommending method, a device, equipment and a medium, which are used for solving the problem of low accuracy of recommended courses due to poor model training effect.

In a first aspect, the present application provides a model training method, applied to a learning platform, including:

acquiring an initial training sample, wherein the initial training sample is a sample acquired from historical data information of a learning platform, and comprises courses learned by students and students;

determining a target course in the courses, learned information of the target course, target students in the students and learning attribute information of the target students according to the initial training sample, wherein the learned information comprises learned time information of the target course acquired from the historical data information, and the learning attribute information comprises a target specialty class of the target students and a class attribute class of the learned historical learning courses of the target students acquired from the historical data information;

Determining course quality of a target course according to the learned time information;

according to the target specialty category and the course attribute category, determining the association degree of the target specialty category and the history learning course;

optimizing the initial training sample according to the course quality and the association degree to obtain an optimized training sample;

and training the model to be trained according to the optimized training sample to obtain a target training model.

In this application, obtaining an initial training sample includes:

acquiring historical data information;

determining a sample to be trained according to the historical data information, wherein the sample to be trained comprises student information, course information and a student specialty class;

determining a first target to-be-trained sample according to the expert class of the learner, wherein the expert class of the learner in the first target to-be-trained sample is the expert class;

and obtaining an initial training sample according to the first target sample to be trained.

In the application, according to a first target to-be-trained sample, an initial training sample is obtained, including:

determining a second target to-be-trained sample from the to-be-trained sample according to the first target to-be-trained sample, wherein the learner information in the second target to-be-trained sample and the learner information in the first target to-be-trained sample meet the requirement of the preset learner information similarity, and the course information in the second target to-be-trained sample and the course information in the first target to-be-trained sample meet the requirement of the preset course information similarity;

And obtaining an initial training sample according to the first target sample to be trained and the second target sample to be trained.

In the present application, determining a course quality of a target course according to learned time information includes:

according to the learned time information, determining the learned times, average course completion rate and average learned duration of the target courses;

acquiring historical learned time information of each course from the historical data information;

and determining the course quality of the target course according to the learned times, the average course completion rate, the average learned time length and the historical learned time information.

In the present application, determining a course quality of a target course according to a learned number of times, an average course completion rate, an average learned duration, and historical learned time information, includes:

according to the history learned time information, determining the history learned times, the history average course completion rate, the history average learned time, the target history learned times, the target history average course completion rate and the target history average learned time, wherein the target history learned times are the average value of the history learned times which are more than the preset times in the history learned times, the target history average course completion rate is the average value of the history average course completion rate which is more than the preset completion rate in the history average course completion rate, and the target history average learned time is the average value of the history average learned time which is more than the preset time in the history average learned time;

And determining the course quality of the target course according to the learned times, the average course completion rate, the average learned duration, the historical learned times, the historical average course completion rate, the historical average learned duration, the target historical learned times, the target historical average course completion rate and the target historical average learned duration.

In the present application, determining a course quality of a target course according to a learned number, an average course completion rate, an average learned duration, a historical learned number, a historical average course completion rate, a historical average learned duration, a target historical learned number, a target historical average course completion rate, and a target historical average learned duration, includes:

obtaining a first index result according to the learned times and the historical average learned times;

obtaining a second index result according to the average course completion rate and the historical average course completion rate;

obtaining a third index result according to the average learned time length and the historical average learned time length;

obtaining a fourth index result according to the learned times and the target history learned times;

obtaining a fifth index result according to the average course completion rate and the target historical average course completion rate;

Obtaining a sixth index result according to the average learned time length and the target historical average learned time length;

and determining the course quality of the target course according to the first index result, the second index result, the third index result, the fourth index result, the fifth index result and the sixth index result.

In the present application, determining, according to the target specialty category and the course attribute category, a degree of association between the target specialty category and the history learning course includes:

comparing the target specialty category with the course attribute category to obtain a comparison result;

and determining the association degree of the target specialty category and the historical learning course according to the comparison result.

In the application, determining the association degree between the target specialty category and the history learning course according to the comparison result comprises the following steps:

if the comparison result represents that the target specialty category is matched with the course attribute category, determining the association degree of the target specialty category and the history learning course according to the comparison result.

if the comparison result represents that the target professional class and the course attribute class are not matched, determining the corresponding group numbers of the students and the courses corresponding to each other in the initial training sample and the sample numbers of the target training samples of the target students and the history learning courses in the initial training sample according to the comparison result;

And determining the association degree of the target students and the historical learning courses according to the corresponding group number and the sample number.

In the application, determining the association degree between the target learner and the historical learning course according to the corresponding group number and the sample number comprises the following steps:

obtaining a group number average value according to the corresponding group number and the sample number;

and determining the association degree of the target students and the historical learning courses according to the group number average value and the sample number.

In a second aspect, the present application provides a course recommendation method, the method comprising:

obtaining login student information of a platform login student, login student course information and a student specialty category;

inputting login student information of a platform login student, login student course information and a student specialty category into a target training model to obtain a target recommended course;

and pushing the target recommended course to the terminal equipment of the platform login student.

In a third aspect, the present application provides a model training apparatus comprising:

the first acquisition module is used for acquiring an initial training sample, wherein the initial training sample is a sample acquired from historical data information of a learning platform and comprises a learner and courses learned by the learner;

a first determining module, configured to determine, according to an initial training sample, a target course in the course, learned information of the target course, learning attribute information of a target learner in the learner, and the target learner, where the learned information includes learned time information of the target course acquired from the history data information, and the learning attribute information includes a target specialty class of the target learner acquired from the history data information and a class attribute class of the learned history learning course of the target learner;

The second determining module is used for determining the course quality of the target course according to the learned time information;

the third determining module is used for determining the association degree of the target specialty category and the history learning course according to the target specialty category and the course attribute category;

the optimizing module is used for optimizing the initial training sample according to the course quality and the association degree to obtain an optimized training sample;

and the training module is used for training the model to be trained according to the optimized training sample to obtain a target training model.

In a fourth aspect, the present application provides a course recommendation device, including:

the second acquisition module is used for acquiring login student information, login student course information and student specialty categories of the platform login students;

the obtaining module is used for inputting login student information of a platform login student, login student course information and a student professional category into the target training model to obtain a target recommended course;

and the pushing module is used for pushing the target recommended course to the terminal equipment of the platform login student.

In a fifth aspect, the present application provides an electronic device, comprising: a processor, a memory communicatively coupled to the processor;

the memory stores computer-executable instructions;

The processor executes the computer-executable instructions stored in the memory to implement the methods described herein.

In a fifth aspect, the present application provides a computer-readable storage medium having stored therein computer-executable instructions that, when executed by a processor, are configured to implement the methods described herein.

According to the model training method, the course recommending method, the device, the equipment and the medium, initial training samples are obtained from historical data information of a learning platform, and the initial training samples comprise courses learned by students and students; determining a target course in the courses, learned information of the target course, target students in the students and learning attribute information of the target students according to the initial training sample, wherein the learned information comprises learned time information of the target course acquired from the historical data information, and the learning attribute information comprises a target specialty class of the target students and a class attribute class of the learned historical learning courses of the target students acquired from the historical data information; determining course quality of a target course according to the learned time information; according to the target specialty category and the course attribute category, determining the association degree of the target specialty category and the history learning course; optimizing the initial training sample according to the course quality and the association degree to obtain an optimized training sample; according to the optimized training sample, training the model to be trained to obtain a target training model, determining the clicked times and the learned time of the target course according to the learned time information of the target course, thereby determining the course quality of the target course, and determining the association degree of the target professional category and the historical learning course according to the target professional category of the target student and the course attribute category of the historical learning course learned by the target student.

Drawings

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

FIG. 1 is a schematic flow chart of a model training method according to an embodiment of the present application;

FIG. 2 is a schematic flow chart of a course recommendation method according to an embodiment of the present disclosure;

FIG. 3 is a flow chart of another model training method according to an embodiment of the present disclosure;

fig. 4 is a schematic structural diagram of a model training device according to an embodiment of the present application;

FIG. 5 is a schematic structural diagram of a course recommendation device according to an embodiment of the present disclosure;

fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Specific embodiments thereof have been shown by way of example in the drawings and will herein be described in more detail. These drawings and the written description are not intended to limit the scope of the inventive concepts in any way, but to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments.

Detailed Description

Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples are not representative of all implementations consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present application as detailed in the accompanying claims.

In the prior art, when a course recommendation model is used for course recommendation, the learning platform generally inputs the historical data of the learning platform as a training sample to a model to be trained to obtain the course recommendation model, however, when the historical data is input into the model to be trained as the training sample, the historical data is generally subjected to simple de-duplication processing, and the historical data is not specifically cleaned according to the actual condition of course learning by a learner, so that the accuracy of the recommended course of the obtained course recommendation model is lower, and the actual requirement of the learner cannot be met.

In order to solve the problems, the method for training the model is provided, the clicked times and the learned time of the target course can be determined according to the learned time information of the target course, so that the course quality of the target course is determined, and the association degree of the target professional category and the history learning course is determined according to the target professional category of the target student and the course attribute category of the history learning course learned by the target student.

The following describes the technical solutions of the present application and how the technical solutions of the present application solve the above technical problems in detail with specific embodiments. The following embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. Embodiments of the present application will be described below with reference to the accompanying drawings.

The execution subject of the recommendation method provided in the embodiment of the present application may be a server. The server can be a mobile phone, a tablet, a computer and other devices. The implementation manner of the execution body is not particularly limited in this embodiment, as long as the execution body can obtain an initial training sample, where the initial training sample is a sample obtained from the historical data information of the learning platform, and the initial training sample includes courses learned by a learner and a learner; determining a target course in the courses, learned information of the target course, target students in the students and learning attribute information of the target students according to the initial training sample, wherein the learned information comprises learned time information of the target course acquired from the historical data information, and the learning attribute information comprises a target specialty class of the target students and a class attribute class of the learned historical learning courses of the target students acquired from the historical data information; determining course quality of a target course according to the learned time information; according to the target specialty category and the course attribute category, determining the association degree of the target specialty category and the history learning course; optimizing the initial training sample according to the course quality and the association degree to obtain an optimized training sample; training the model to be trained according to the optimized training sample to obtain the target training model.

Among them, machine Learning (ML) is a multi-domain interdisciplinary, and involves multiple disciplines such as probability theory, statistics, approximation theory, convex analysis, and computational complexity theory. Is a multi-field intersection subject, and relates to a plurality of subjects such as probability theory, statistics, approximation theory, convex analysis, calculation complexity theory and the like. Machine learning is the core of artificial intelligence, a fundamental approach to letting computers have intelligence, which is applied throughout various areas of artificial intelligence. Machine learning and deep learning typically include techniques such as artificial neural networks, belief networks, reinforcement learning, transfer learning, induction learning, teaching learning, and the like.

Fig. 1 is a flow chart of a model training method according to an embodiment of the present application. The implementation body of the method may be a server or other servers, and the embodiment is not particularly limited herein, as shown in fig. 1, and the method may include:

s101, acquiring an initial training sample, wherein the initial training sample is a sample acquired from historical data information of a learning platform, and comprises a learner and courses learned by the learner.

The initial training sample may refer to a sample obtained from historical data information of the learning platform. There may be multiple initial training samples. The initial training sample may be a sample for model training, the initial training sample may be obtained from historical data information, and the initial training sample may include a learner and a course, wherein the learner may include an ID of the learner, and after obtaining the ID of the learner, a learner attribute of the learner and a talent class of a professional to which the learner belongs may be determined. The course may be a course learned by a learner, and in the embodiment of the present application, at least one course learned by the learner is included.

The historical data information may be historical data of learning by a learner logging in the learning platform, the historical data information may include learning behavior class attributes, learner class attributes, curriculum class attributes and professional talent class, in this embodiment of the present application, the learning behavior class attributes may include a learner identification ID (Identity document, identification), learning time, learning curriculum identification, the learner class attributes may include a learner identification ID, a learner attribute 1, a learner attribute 2, a learner attribute 3, …, and a professional talent class number, the curriculum class attributes may include a curriculum identification ID, a curriculum attribute 1, a curriculum attribute 2, a curriculum attribute 3, …, and the professional talent class may include a professional talent class number, a professional talent class name, a professional talent class keyword 1, a professional talent class keyword 2, and a professional talent class keyword 3, ….

In an embodiment of the present application, the method for obtaining an initial training sample may include:

acquiring historical data information;

The sample to be trained can comprise a student ID, a student attribute 1, a student attribute 2, a student attribute 3, …, a learning time, a learning course ID and a professional talent category number. In some embodiments, after obtaining the historical data information, "learning behavior class attribute, learner class attribute, curriculum class attribute, and professional talent class" may be filled into "learner ID, learner attribute 1, learner attribute 2, learner attribute 3, …, learning time, learning curriculum ID, professional talent class number" to obtain an initial training sample. For example, the initial training sample may be "Zhang San, new employee, communication engineering specialty, 24 years old, master, wireless communication post, 2021, 3 months 30 No. 9:00-10:00, cloud computing technology course ID,100223".

The learner information may refer to "learner ID, learner attribute 1, learner attribute 2, learner attribute 3, …" of the learner.

Course information may refer to a "learning time, learning course ID" of a learner.

The trainee specialty category may refer to a "professional talent category" of a trainee, where the professional talent category may be a trainee specialty category determined according to a trainee's specialty or post. In the embodiment of the application, the talent category of the specialty may include a specialty category, a non-specialty category, and a potential specialty category, wherein a learner of the specialty category may refer to a learner with a specific specialty or post; a non-professional class of trainees may refer to trainees whose profession or post is not determinable; a potential professional class learner may refer to a learner whose profession or post is not determinable, but whose determined learner information and course information have a high similarity by comparing the learner information and course information of the professional class learner.

The first target sample to be trained may be a sample to be trained in which the learner's professional category is a professional category.

In this embodiment of the present application, according to a first target sample to be trained, a method for obtaining an initial training sample may include:

The second target sample to be trained is a sample to be trained with a learner professional category as a potential professional category. In the embodiment of the present application, after the first target to-be-trained sample is determined, a second target to-be-trained sample in the to-be-trained samples may be determined according to the learner information and the course information in the first target to-be-trained sample. The requirement of the preset similarity may refer to similarity of learner information and similarity of course information, for example, if the similarity of the learner information in the sample to be trained and the text of the learner information in the first target sample to be trained meets the requirement of the preset similarity, and the similarity of the course information in the sample to be trained and the text of the course information in the first target sample to be trained meets the requirement of the preset similarity, the sample to be trained may be determined to be the second target sample to be trained, where the method for determining the similarity of the text may include determining through a natural language model.

After the first target sample to be trained and the second target sample to be trained are obtained, an initial training sample can be obtained.

It should be noted that, the user information (including, but not limited to, user equipment information, user personal information, etc.) and the data (including, but not limited to, data for analysis, stored data, presented data, etc.) according to the embodiments of the present application are information and data authorized by the user or fully authorized by each party, and the collection, use and processing of the related data need to comply with the related laws and regulations and standards of the related country and region, and provide corresponding operation entries for the user to select authorization or rejection.

S102, determining a target course in courses, learned information of the target course, and learning attribute information of a target learner and a target learner in a learner according to an initial training sample, wherein the learned information comprises learned time information of the target course acquired from historical data information, and the learning attribute information comprises a target specialty class of the target learner and a class attribute class of the learned historical learning course of the target learner acquired from the historical data information.

The target course may be any course selected from the initial training samples. The learned information of the target course may be obtained from the history data information based on the ID of the target course.

The target trainee may be any trainee selected from the initial training sample. The learning attribute information of the target learner may be acquired from the history data information according to the ID of the target learner.

The target specialty category may refer to a category corresponding to a professional category of a learner, and in this embodiment of the present application, the target specialty category may be obtained by extracting keywords in an information base corresponding to the professional category of the learner, for example, the learner is "information and communication engineering specialty", and the target specialty category may include "communication", "multimedia", and so on. The information base may be a talent category database, which may be stored in historical data information of the learning platform.

The history learning course may be all courses learned by the target learner, and the course attribute category may refer to a classification of the history learning course, for example, the classification of the history learning course is a communication classification, and the learning attribute information may be "communication". In the embodiment of the application, the course attribute category can be obtained by extracting the keyword corresponding to the course attribute category from a course information database, and the database can be stored in the historical data information of the learning platform.

S103, determining the course quality of the target course according to the learned time information.

Wherein the learned time information may refer to the learned time of the target course.

Course quality may characterize the duration and number of times a user learns a target course. The longer the duration and the higher the number of times of learning the target course, the better the course quality. In the embodiment of the application, the course quality of the target course can be determined through the learned times and the average learned time duration.

In an embodiment of the present application, a method for determining a course quality of a target course according to learned time information may include:

The average course completion rate may refer to a progress of learning the target course, for example, when the learning duration of the target course is 60 minutes, if the average learned duration of the target course is 30 minutes, the average course completion rate is 50%.

The historical learned time information may refer to the learned time of each course.

In this embodiment of the present application, the method for determining the course quality of the target course according to the learned times, the average course completion rate, the average learned duration and the historical learned time information may include:

The historical average course completion rate may refer to a progress of learning courses.

In this embodiment of the present application, the method for determining the course quality of the target course according to the learned times, the average course completion rate, the average learned time period, the historical learned times, the historical average course completion rate, the historical average learned time period, the target historical learned times, the target historical average course completion rate, and the target historical average learned time period may include:

The method for determining the course quality of the target course can be represented by the formula:

and (5) determining.

Wherein ST is _j The number of learned times of the target course may be,the average learned times for the history of all courses, < +.>The first index result can be 1 when the learned times of the target courses are larger than the historical average learned times of all courses;

the target historical learned times greater than a preset time threshold in the historical learned times in all courses can be referred, wherein the preset time threshold can be the first 30% of the set historical learned times, and when the learned times of the target courses are greater than the historical target average learned times, the fourth index result is 1;

CST _j can refer to the average learned times of target courses, CT _j May refer to the course duration of the target course,average course completion rate for target courses, < >>Can refer to the historical average course completion rate of all courses,/-for all courses>The second index result is 1 when the average course completion rate of the target course is greater than the historical average course completion rate of all courses;

the target history learned times greater than a preset completion rate threshold in the historical course completion rates may be referred to, where the preset completion rate threshold may be the times of the first 20% in all preset historical course completion rates, and when the average course completion rate of the target courses is greater than the target historical average course completion rate, the fourth index result is 1;

CTST _j May refer to an average learned duration of a target course,the method can refer to the historical average learned time length of all courses, and when the average learned time length of the target course is longer than the historical average learned time length of all courses, the third index result is 1;

the target historical average learned duration greater than a preset duration threshold in the historical average learned duration may be referred to, where the preset duration threshold may be the first 40% of the preset historical average learned durations, and when the average learned time of the target course is greater than the target historical average learned duration, the fifth index result is 1;

alpha, beta and gamma are constant coefficients, and may take a default value, for example, alpha, beta and gamma may each be 1.

In the embodiment of the present application, when f1 is set in advance _j When not less than 3, the course quality of the target course meets the quality requirement, and when f1 _j <And 3, characterizing that the course quality of the target course does not meet the quality requirement.

S104, determining the association degree of the target specialty category and the historical learning course according to the target specialty category and the course attribute category.

The association degree can represent whether the target specialty category of the learner is associated with the category of the history learning course. In this embodiment of the present application, according to the target specialty category and the course attribute category, the method for determining the association degree between the target specialty category and the history learning course may include:

The comparison result may refer to a result of whether the target specialty category and the course attribute category match. In the embodiment of the application, when the target professional category and the course attribute category have the same keyword, the comparison result can be determined to be matching. For example, when the target specialty category and the course attribute category are both communication, then the target specialty category and the course attribute category match.

In this embodiment of the present application, according to the comparison result, the method for determining the association degree between the target specialty category and the history learning course may include:

The corresponding group number may refer to the group number of a learner and a course corresponding to the learner, which occur in the initial training sample.

In an embodiment of the present application, the method for determining the association degree between the target learner and the historical learning course according to the corresponding group number and the sample number may include:

When the comparison result is that the target professional category and the course attribute category are not matched, the method for determining the association degree of the target learner and the history learning course can be realized through the formula:

and (5) determining.

Wherein Y is _k,j For the presence of a history learning course in the initial training sample k, which characterizes the course attribute category, Y is _k,j 1, when the history learning course does not appear in the initial training sample k, then Y _k,j Is 0;

Z _k,r whether or not a target specialty class, which may be a target learner, is present in the initial training sample k, when the target specialty class is present in the initial training sample k, then Z _k,r 1, when the target specialty class does not appear in the initial training sample k, then Z _k,r Is 0.

May refer to the number of samples of the target training sample.

T1 may be a predetermined number threshold, i.e. when the number of samples of the target training sample is greater than T1,

group that can refer to the presence of a target specialty class and historical learning courses in a learner's specialty class and courses that characterize a course attribute classA number average value, where δ is a constant coefficient, is used to adjust the scale, and in the embodiment of the present application, δ may be set to 1.

In the embodiment of the present application, when f2 is set in advance _j,r <1, when the association degree of the target learner and the history learning course does not meet the requirement, f2 _j,r And when the correlation degree is more than or equal to 1, the correlation degree between the target student and the history learning course meets the requirement.

In this embodiment of the present application, a target learner in one initial training sample may correspond to a plurality of historical learning courses, and when at least one group of target learner and historical learning courses with a satisfactory association degree exist in one initial training sample, the initial training sample is characterized as a training sample with a satisfactory association degree.

And S105, optimizing the initial training sample according to the course quality and the association degree to obtain an optimized training sample.

The optimization may refer to a process of reserving or deleting the target course and the target learner according to the course quality and the association degree. For example, when the course quality of the target course does not meet the quality requirement, the initial training sample with the target course can be deleted, or the course quality of all courses in the initial training sample set can be determined, and the target course with the quality which does not meet the quality requirement can be deleted; when the association degree between the target specialty category and the history learning course does not meet the association degree requirement, the initial training sample set with the target specialty category and the history learning course can be deleted, and the association degree between all specialty categories and the history learning course in the initial training sample set can be determined, so that all the initial training samples with the target specialty category and the history learning course which do not meet the association degree can be deleted, and therefore the optimized training sample can be obtained.

S106, training the model to be trained according to the optimized training sample to obtain a target training model.

The model to be trained can be a neural network model, training is carried out on the model to be trained by taking a learner, a learner attribute, learning time and a professional category in an optimized training sample as input items and taking courses as labels, and the weight of each input item is continuously adjusted until the output result of the model to be trained meets the expected requirement, so that the target training model can be obtained.

In the embodiment of the application, the data mining software IBM SPSS Statistics is used to input the learner attribute and the learning course to the data mining software IBM SPSS Statistics, so that the target training model can be obtained through training.

According to the model training method, the clicked times and the learned time of the target course can be determined according to the learned time information of the target course, so that the course quality of the target course is determined, the association degree of the target professional category and the history learning course is determined according to the target professional category of the target student and the course attribute category of the history learning course learned by the target student, therefore, an initial training sample is optimized through the course quality and the association degree, and data which are not in accordance with requirements in the course quality and the association degree are deleted, so that the data quality of the optimized training sample is higher, and the training effect of the model to be trained is improved.

Fig. 2 is a flow chart of a course recommendation method according to an embodiment of the present application. The implementation body of the method may be a server or other servers, and the embodiment is not particularly limited herein, as shown in fig. 2, and the method may include:

S201, obtaining login student information of a platform login student, course information of the login student and professional categories of the student;

s202, inputting login student information, login student course information and student specialty categories of a platform login student into a target training model to obtain a target recommended course;

s203, pushing the target recommended course to terminal equipment of the platform login student.

Wherein, the platform login learner may refer to a learner logging in the learning platform. After a student logs in the platform, the learning platform can obtain student information, log-in student course information and student professional category of the platform log-in student after the authorization of the student, input the student information, the log-in student course information and the student professional category into a target training model, obtain a target recommended course through the target training model, and push the target recommended course to terminal equipment of the platform log-in student.

Fig. 3 is a schematic flow chart of another model training method provided in the embodiment of the present application, where an execution subject of the method may be a computer, and the embodiment is not particularly limited herein, and as shown in fig. 3, the method may include:

s301, acquiring historical learning data of a known professional talent list in a learning platform.

Wherein the history learning data includes: learning behavior class attribute r1= { learner unique identifier ID, learning time, learning course unique identifier }; student class attribute r2= { student unique identifier ID, student attribute 1, student attribute 2, student attribute 3, …, professional talent class number }; course class attribute r3= { course unique identifier ID, course attribute 1, course attribute 2, course attribute 3, … }; specialty talent category r4= { specialty talent category number, specialty talent category name, specialty talent category keyword 1, specialty talent category keyword 2, specialty talent category keyword 3 … }.

In this embodiment of the present application, the professional talent category R4 may be determined according to registration information of a learner in the learning platform, and obtaining the historical learning data of the known professional talent list in the learning platform may refer to determining the known professional talents by screening the registration information of the learner, and obtaining the corresponding historical learning data by the known professional talents.

S302, obtaining an initial training data set of known professional talents according to the historical learning data.

The initial training data of the known professional talents can be obtained according to the learning behavior class attribute R1, the learner class attribute R2, the course class attribute R3 and the professional talent class R4.

In the embodiment of the present application, the initial training data r0= { learner ID, learner attribute 1, learner attribute 2, learner attribute 3, …, learning time, learning course ID, professional talent category number }. Wherein, the argument in the initial training data R0 may be set as: student attribute 1, student attribute 2, student attribute 3, …, learning time, learning course ID, dependent variables can be set to: professional talent category number. Wherein, the professional talent category number can represent the category of the professional talent, for example, when the professional talent category is the known professional talent, the head number of the professional talent category number is "1"; when the professional talents are potential professional talents, the first number of the professional talents category number is "2"; when the professional talent category is non-professional talents, the professional talent category number is empty.

The learning behavior class attribute R1, the learner class attribute R2, the curriculum class attribute R3 and the professional talent class R4 are filled into R0, so that initial training data X1 of the known professional talents can be obtained, and an initial training data set of the known professional talents can be obtained from a plurality of initial training data, so that the initial training data set x= { X1, X3, …, xi, …, xi }, of the known professional talents.

S303, determining potential professional talents and initial training data sets of the potential professional talents according to the initial training data sets of the known professional talents;

s304, obtaining an initial training sample according to the initial training data set of the known professional talents and the initial training data set of the potential professional talents.

The independent variables of the initial training sample may be: student attribute 1, student attribute 2, student attribute 3, …, learning time, professional talent category number, dependent variables can be: learning course ID1, learning course ID2, …, learning course IDN.

S305, obtaining an initial target course set according to the initial training sample, wherein the initial target course set comprises more than one initial target courses.

The initial target course set may be a set of initial target courses obtained by screening learning courses extracted from an initial training sample.

Screening the learning course may refer to a process of eliminating repeated learning courses.

S306, determining a target course meeting the requirements in the initial target courses.

Wherein, the compliance requirement can refer to a course that meets the quality requirement and the association requirement.

The method for determining whether the initial target course meets the quality requirement can be to determine the learning rate and learning duration index of the course. In the embodiment of the present application, the formula may be:

Wherein f1 of the initial target course is selected _j And (3) courses more than or equal to 3.

The method for determining whether the initial target course meets the association requirement can be the association degree of the course and the professional class. In the embodiment of the present application, the formula may be:

and (5) determining.

Wherein f2 of the initial target course is selected _j,r And (3) courses more than or equal to 1.

In the embodiment of the application, f1 of the initial target course is selected _j Course not less than 3 and f2 of initial target course _j,r After the courses which are more than or equal to 1, the courses which simultaneously meet the quality requirement and the association degree requirement can be determined as target courses, and the courses which respectively meet the quality requirement and the association degree requirement can be selected as target courses.

S307, determining a target training sample according to the target course.

The target training sample comprises a target course, and after the target can be determined, the target training sample comprising the target course can be determined in an initial training data set of known professional talents and an initial training data set of potential professional talents.

S308, inputting the target training sample into a model to be trained for model training, and obtaining a course recommendation model.

In this embodiment of the present application, "learner attribute 1, learner attribute 2, learner attribute 3, …, learning time, and professional talent class number" as "independent variable" and "target course" as "label" may be input into the model for training, so as to obtain a course recommendation model.

In this embodiment of the present application, the target training sample may also be input to the data mining software, so as to train to obtain the course recommendation model, where the data mining software may be IBM SPSS Statistics.

After the course recommendation model is obtained, the learning platform can recommend courses to the students according to the student attributes of the students.

According to the method for training the model, based on the characteristics of the historical learning data, the data can be optimized by combining the data recording time, the time length and the talent type of the learner, invalid data and low-quality data in the historical learning data are removed, so that the quality of the data for training the model is higher, and the effect of training the model is improved.

Fig. 4 is a schematic structural diagram of a model training device according to an embodiment of the present application. As shown in fig. 4, the model training apparatus 40 includes: a first acquisition module 401, a first determination module 402, a second determination module 403, a third determination module 404, an optimization module 405, and a training module 406. Wherein:

the first obtaining module 401 is configured to obtain an initial training sample, where the initial training sample is a sample obtained from historical data information of a learning platform, and the initial training sample includes a learner and a course learned by the learner;

A first determining module 402, configured to determine, according to an initial training sample, a target course in a course, learned information of the target course, learning attribute information of a target learner in a learner, and the target learner, where the learned information includes learned time information of the target course acquired from the history data information, and the learning attribute information includes a target specialty class of the target learner acquired from the history data information and a class attribute class of a learned history learning course of the target learner;

a second determining module 403, configured to determine a course quality of the target course according to the learned time information;

a third determining module 404, configured to determine, according to the target specialty category and the course attribute category, a degree of association between the target specialty category and the history learning course;

the optimizing module 405 is configured to optimize the initial training sample according to the course quality and the association degree, so as to obtain an optimized training sample;

the training module 406 is configured to train the model to be trained according to the optimized training sample, to obtain a target training model.

In this embodiment of the present application, the first obtaining module 401 may be further specifically configured to:

acquiring historical data information;

In this embodiment of the present application, the second determining module 403 may be further specifically configured to:

In the embodiment of the present application, the third determining module 404 may be further specifically configured to:

As can be seen from the above, the model training device of the present embodiment is configured to obtain an initial training sample by using the first obtaining module 401, where the initial training sample is a sample obtained from the historical data information of the learning platform, and the initial training sample includes courses learned by a learner and a learner; determining, by the first determining module 402, a target course in the course, learned information of the target course, learning attribute information of the target learner and the target learner in the learner, according to the initial training sample, wherein the learned information includes learned time information of the target course acquired from the history data information, and the learning attribute information includes a target specialty class of the target learner acquired from the history data information and a class attribute class of the learned history learning course of the target learner; a second determining module 403, configured to determine a course quality of the target course according to the learned time information; a third determining module 404, configured to determine a degree of association between the target specialty category and the history learning course according to the target specialty category and the course attribute category; the optimizing module 405 is configured to optimize the initial training sample according to the course quality and the association degree, so as to obtain an optimized training sample; the training module 406 is configured to train the model to be trained according to the optimized training sample, so as to obtain a target training model. Thereby, the effect of model training is improved.

Fig. 5 is a schematic structural diagram of a course recommendation device according to an embodiment of the present application. As shown in fig. 5, the course recommendation device 50 includes: a second obtaining module 501, an obtaining module 502 and a pushing module 503.

Wherein:

a second obtaining module 501, configured to obtain login student information, login student course information, and a student specialty class of a platform login student;

the obtaining module 502 is configured to input login learner information, login learner course information and a learner professional category of a platform login learner into the target training model to obtain a target recommended course;

and the pushing module 503 is configured to push the target recommended course to the terminal device of the platform login student.

As can be seen from the above, the course recommendation device in this embodiment is configured to obtain, by using the second obtaining module 501, information of a login learner, course information of the login learner, and a professional category of the learner; the obtaining module 502 is configured to obtain module 502, and input login learner information, login learner course information and a learner professional category of a platform login learner to a target training model to obtain a target recommended course; and pushing the target recommended course to the terminal equipment of the platform login student by a pushing module 503. Therefore, the recommended course is more accurate.

Fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present application. As shown in fig. 6, the electronic device 60 includes:

the electronic device 60 may include one or more processing cores 'processors 601, one or more computer-readable storage media's memory 602, communication components 603, and the like. The processor 601, the memory 602, and the communication section 603 are connected via a bus 604.

In a specific implementation, at least one processor 601 executes computer-executable instructions stored in memory 602, causing at least one processor 601 to perform the model training method and the course recommendation method as described above.

The specific implementation process of the processor 601 may refer to the above-mentioned method embodiment, and its implementation principle and technical effects are similar, and this embodiment will not be described herein again.

In the embodiment shown in fig. 6, it should be understood that the processor may be a central processing unit (english: central Processing Unit, abbreviated as CPU), or may be other general purpose processors, digital signal processors (english: digital Signal Processor, abbreviated as DSP), application specific integrated circuits (english: application Specific Integrated Circuit, abbreviated as ASIC), or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the present invention may be embodied directly in a hardware processor for execution, or in a combination of hardware and software modules in a processor for execution.

The Memory may comprise high-speed Memory (Random Access Memory, RAM) or may further comprise Non-volatile Memory (NVM), such as at least one disk Memory.

The bus may be an industry standard architecture (Industry Standard Architecture, ISA) bus, an external device interconnect (Peripheral Component, PCI) bus, or an extended industry standard architecture (Extended Industry Standard Architecture, EISA) bus, among others. The buses may be divided into address buses, data buses, control buses, etc. For ease of illustration, the buses in the drawings of the present application are not limited to only one bus or one type of bus.

In some embodiments, a computer program product is also presented, comprising a computer program or instructions which, when executed by a processor, implement the steps of any of the model training methods or course recommendation methods described above.

The specific implementation of each operation above may be referred to the previous embodiments, and will not be described herein.

Those of ordinary skill in the art will appreciate that all or a portion of the steps of the various methods of the above embodiments may be performed by instructions, or by instructions controlling associated hardware, which may be stored in a computer-readable storage medium and loaded and executed by a processor.

To this end, embodiments of the present application provide a computer readable storage medium having stored therein a plurality of instructions capable of being loaded by a processor to perform steps in any one of the model training methods or the course recommendation methods provided by embodiments of the present application.

Wherein the storage medium may include: read Only Memory (ROM), random access Memory (RAM, random Access Memory), magnetic or optical disk, and the like.

According to one aspect of the present application, there is provided a computer program product or computer program comprising computer instructions stored in a computer readable storage medium.

Because the instructions stored in the storage medium may perform steps in any model training method or course recommendation method provided in the embodiments of the present application, the beneficial effects that may be achieved in any model training method or course recommendation method provided in the embodiments of the present application may be achieved, which are detailed in the previous embodiments and are not described herein.

Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the application following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice within the art to which the application pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the application being indicated by the following claims.

It is to be understood that the present application is not limited to the precise arrangements and instrumentalities shown in the drawings, which have been described above, and that various modifications and changes may be effected without departing from the scope thereof. The scope of the application is limited only by the appended claims.

Claims

1. A model training method, applied to a learning platform, the method comprising:

acquiring an initial training sample, wherein the initial training sample is a sample acquired from historical data information of the learning platform, and comprises a learner and courses learned by the learner;

determining a target course in the courses, learned information of the target courses, target students in the students and learning attribute information of the target students according to the initial training samples, wherein the learned information comprises learned time information of the target course acquired from the historical data information, and the learning attribute information comprises target specialty categories of the target students and learned course attribute categories of the target students acquired from the historical data information; determining course quality of the target course according to the learned time information;

Determining the association degree of the target specialty category and the history learning course according to the target specialty category and the course attribute category;

2. The method of claim 1, wherein the obtaining initial training samples comprises:

acquiring historical data information;

determining a sample to be trained according to the historical data information, wherein the sample to be trained comprises learner information, course information and a learner specialty class;

determining a first target to-be-trained sample according to the learner specialty category, wherein the learner specialty category in the first target to-be-trained sample is the specialty category;

3. The method according to claim 2, wherein obtaining an initial training sample from the first target sample to be trained comprises:

4. The method of claim 1, wherein said determining a course quality of said target course based on said learned time information comprises:

determining the learned times, average course completion rate and average learned duration of the target courses according to the learned time information;

and determining the course quality of the target course according to the learned times, the average course completion rate, the average learned duration and the historical learned time information.

5. The method of claim 4, wherein said determining a course quality for said target course based on said learned times, said average course completion rate, said average learned duration, and said historical learned time information comprises:

according to the historical learned time information, determining historical learned times, historical average course completion rates, historical average learned time periods, target historical learned times, target historical average course completion rates and target historical average learned time periods, wherein the target historical learned times are average values of the historical learned times which are more than preset times in the historical learned times, the target historical average course completion rates are average values of the historical average course completion rates which are more than preset completion rates in the historical average course completion rates, and the target historical average learned time periods are average values of the historical average learned time periods which are more than preset time periods in the historical average learned time periods;

And determining the course quality of the target courses according to the learned times, the average course completion rate, the average learned time length, the historical learned times, the historical average course completion rate, the historical average learned time length, the target historical learned times, the target historical average course completion rate and the target historical average learned time length.

6. The method of claim 5, wherein the determining the course quality of the target course based on the number of times learned, the average course completion rate, the average learned duration, the historical number of times learned, the historical average course completion rate, the historical average learned duration, the target historical number of times learned, the target historical average course completion rate, and the target historical average learned duration comprises:

7. The method of claim 1, wherein the determining the association of the target specialty category with the historic learning lesson based on the target specialty category and the lesson attribute category comprises:

and determining the association degree of the target specialty class and the historical learning course according to the comparison result.

8. The method of claim 7, wherein determining the relevance of the target specialty category to the historical learning course based on the comparison result comprises:

And if the comparison result represents that the target specialty class is matched with the course attribute class, determining the association degree of the target specialty class and the history learning course according to the comparison result.

9. The method of claim 7, wherein determining the relevance of the target specialty category to the historical learning course based on the comparison result comprises:

if the comparison result represents that the target specialty class and the course attribute class are not matched, determining the corresponding group numbers of the students and the courses corresponding to each other in the initial training sample and the sample numbers of the target training samples of the target students and the history learning courses in the initial training sample according to the comparison result;

and determining the association degree of the target learner and the historical learning course according to the corresponding group number and the sample number.

10. The method of claim 9, wherein said determining a degree of association of the target learner with the historic learning course based on the corresponding group number and the sample number comprises:

And determining the association degree of the target student and the historical learning course according to the group number average value and the sample number.

11. A course recommendation method, the method comprising:

inputting login student information, login student course information and student specialty categories of the platform login students into a target training model trained by the method according to claims 1-10 to obtain a target recommended course;

12. A model training device, comprising:

the first acquisition module is used for acquiring an initial training sample, wherein the initial training sample is a sample acquired from historical data information of a learning platform and comprises a learner and a course learned by the learner;

a first determining module, configured to determine, according to the initial training sample, a target course of the courses, learned information of the target course, a target learner of the trainees, and learning attribute information of the target learner, where the learned information includes learned time information of the target course acquired from the history data information, and the learning attribute information includes a target specialty class of the target learner and a class attribute class of a learned history learning course of the target learner acquired from the history data information;

the third determining module is used for determining the association degree of the target specialty category and the historical learning course according to the target specialty category and the course attribute category;

13. A course recommendation device, comprising:

the obtaining module is used for inputting login student information, login student course information and student specialty categories of the platform login students into a target training model obtained by training according to the method of claims 1-10 to obtain a target recommended course;

14. An electronic device, comprising: a processor, and a memory communicatively coupled to the processor;

the memory stores computer-executable instructions;

the processor executes computer-executable instructions stored in the memory to implement the method of any one of claims 1 to 11.

15. A computer readable storage medium having stored therein computer executable instructions which when executed by a processor are adapted to carry out the method of any one of claims 1 to 11.