Model training method, learning behavior recognition method, device, equipment and medium

Abstract

The invention provides a model training method, a learning behavior recognition method and device, equipment and a medium. The method comprises the steps that a sample student and a sample course are acquired, a sample label of the sample student for learning the sample course is determined according to the course type of the sample course, the course type comprises a necessary course and a non-necessary course, and the sample label is used for indicating whether the learning behavior of the sample student for learning the sample course is active learning; according to the trainee features of the sample trainee, the course features of the sample course, the historical learning record features of the sample trainee learning the sample course, and the sample label, performing model training on the pre-acquired to-be-trained model to obtain the learning behavior recognition model, and through the learning behavior recognition model, improving the accuracy of learning behavior recognition.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a model training method, a learning behavior recognition method, a device, equipment, and a medium. Background Technology

[0002] With the widespread adoption and deep integration of the internet, corporate online learning platforms have become an important channel for internal education and knowledge sharing. To better recommend courses and enhance learners' interest, it is necessary to identify learners' learning behaviors.

[0003] Existing technologies for learning behavior recognition suffer from poor recognition performance. Summary of the Invention

[0004] This application provides a model training method, a learning behavior recognition method, an apparatus, a device, and a medium to solve the problem of poor learning behavior recognition performance in the prior art.

[0005] Firstly, this application provides a model training method, including:

[0006] Obtain sample students and sample courses. Sample courses are those that sample students have taken.

[0007] Based on the course type of the sample courses, determine the sample labels for the sample students' learning of the sample courses. The course types include compulsory courses and non-compulsory courses. The sample labels are used to indicate whether the learning behavior of the sample students in learning the sample courses is active learning.

[0008] Based on the student characteristics of the sample students, the course characteristics of the sample courses, the historical learning record characteristics of the sample students in learning the sample courses, and the sample labels, the pre-acquired training model is trained to obtain the learning behavior recognition model.

[0009] Furthermore, based on the course type of the sample courses, sample tags for the sample students learning the sample courses are determined, including:

[0010] Based on the course type of the sample courses, determine the initial sample labels for the sample students learning the sample courses;

[0011] If the initial sample label is active learning, then the initial sample label will be determined as the sample label of the sample students learning the sample course;

[0012] If the initial sample label is passive learning, then the initial sample label is corrected according to the learning entry point of the sample students learning the sample course to obtain the sample label of the sample students learning the sample course. The learning entry point includes any one of the following: learning zone, training course, open course, live course, and required course.

[0013] Furthermore, based on the course type of the sample courses, the initial sample labels for the sample students learning the sample courses are determined, including:

[0014] Determine the course type of the sample courses;

[0015] If the course type of the sample course is a required course, then the initial sample label for the sample students learning the sample course is passive learning.

[0016] If the sample course is a non-required course, then the initial sample label for the sample students learning the sample course is active learning.

[0017] Furthermore, if the initial sample label is passive learning, then the initial sample label is modified according to the learning entry point of the sample students learning the sample course, to obtain the sample label of the sample students learning the sample course, including:

[0018] Based on the learning entry point of the sample students to learn the sample courses, obtain the first set of all courses within the learning entry point;

[0019] Determine the first average completion rate of the first course set based on the completion rate of each student in each course in the first course set;

[0020] Based on all students who have achieved the learning objectives in each course in the first course set, the number of students who have achieved the learning objectives, and the total number of times all students who have achieved the learning objectives have completed their first course, the average completion rate of the first course set is determined. The average completion rate is the average of the average completion rates of all courses in the first course set.

[0021] The first indicator value is obtained by comparing the completion rate and number of times the sample students studied the sample courses within the preset learning period, the average completion rate per course and the first average completion rate, and the average completion rate per course with the preset first threshold.

[0022] If the value of the first indicator is less than the preset threshold of the first indicator, then active learning will be identified as the sample label for sample students learning sample courses.

[0023] Furthermore, based on all students who have achieved the learning objectives in each course within the first course set, the total number of students who have achieved the learning objectives, and the total number of times all students who have achieved the learning objectives have completed their first course, the average completion rate per course for the first course set is determined, including:

[0024] Get all students who have achieved the learning goals for each course in the first course set, the number of students who have achieved the learning goals, and the first total number of learning sessions for all students who have achieved the learning goals;

[0025] The average completion rate for each course is determined by the ratio of the number of students who meet the standard to the total number of times the course is completed.

[0026] The average completion rate of the first course set is determined based on the average completion rate of each course.

[0027] Furthermore, based on the comparison results of the completion rate and number of times the sample students studied the sample courses within the preset learning period, the average completion rate per course and the first average completion rate, and the comparison results of the average completion rate per course and the preset first threshold, the first indicator value is obtained, which satisfies:

[0028]

[0029] Among them, f1 k S is the first indicator value. i For the i-th sample student, C j For S i The j-th sample course is studied, where T is the preset time period. For S i Learning C within T j Total study time C j Course duration, For S i Learning C within T j The completion rate, where α is the adjustment coefficient. For S i Learning C within T j Number of times of learning C j The first total number of times you learn within the learning portal. C j The number of people within the learning portal C j The average completion rate is denoted as A, where A is the first average completion rate, and J is the second average completion rate. p B represents the number of courses within the learning portal, where B is the preset first threshold, and I... p This refers to the number of learners who have engaged in learning activities within the learning portal and whose learning volume has reached the threshold.

[0030] Furthermore, if the initial sample label is passive learning, then the initial sample label is modified according to the learning entry point of the sample students learning the sample course, to obtain the sample label of the sample students learning the sample course, including:

[0031] Determine whether the sample students are new students;

[0032] If the sample student is a new student, the initial sample label will be determined as the sample label of the sample student learning the sample course;

[0033] If the sample learners are returning learners, the initial sample labels are modified based on the learning entry points of the sample learners for the sample courses, resulting in sample labels for the sample learners' learning of the sample courses.

[0034] Further, determining whether a sample student is a new student includes:

[0035] The average number of learning sessions per student in the first course set is obtained by comparing the ratio of the second total number of learning sessions to the total number of students in each course within the learning portal.

[0036] The second indicator value is obtained based on the total number of times the sample students studied all courses in the first course set and the average number of times they studied per student.

[0037] If the value of the second indicator is less than the preset threshold of the second indicator, the sample student is determined to be a new student.

[0038] If the value of the second indicator is greater than or equal to the preset threshold of the second indicator, the sample student is determined to be a returning student.

[0039] Furthermore, based on the third total number of times the sample students studied all courses in the first course set, and the second average number of times they studied each course, the second indicator value is obtained, which satisfies:

[0040]

[0041] in, The second indicator value, This is the third total number of learning sessions. C j The second total number of learning sessions, C j Total number of students γ represents the average number of learning sessions per person for the second person, and γ is the adjustment coefficient.

[0042] Furthermore, if the sample learners are returning students, the initial sample tags are modified based on the learning entry point of the sample learners for the sample courses, resulting in sample tags for the sample learners' learning of the sample courses, including:

[0043] Obtain the second set of courses already studied by the sample learners within the learning portal;

[0044] Based on the completion rate of each course studied in the second course set by the sample students, determine the second average completion rate of the second course set by the sample students.

[0045] The third indicator value is obtained based on the completion rate and number of times the sample students studied the sample courses within the preset learning period, the average completion rate of the sample courses per session, the first average completion rate of the first set of courses within the learning portal, and the second average completion rate.

[0046] If the value of the third indicator is less than the preset threshold for the third indicator, then active learning will be identified as the sample label for the sample students learning the sample course.

[0047] Furthermore, based on the completion rate and number of times sample students studied the sample courses within the preset learning period, the average completion rate per course, the first average completion rate of the first set of courses within the learning portal, and the second average completion rate, a third indicator value is obtained, satisfying the following:

[0048]

[0049] Among them, f2 k J1 is the third indicator value. p The number of courses in the second course set. This is the second-highest average completion rate.

[0050] Secondly, this application provides a learning behavior recognition method, including:

[0051] Obtain the student to be identified and the course to be identified. The course to be identified is the course that the student to be identified has taken.

[0052] The student characteristics of the student to be identified, the course characteristics of the course to be identified, and the historical learning record characteristics of the student to be identified in learning the course to be identified are input into the learning behavior identification model for identification, and the identification result is obtained. The learning behavior identification model is the learning behavior identification model provided in this application. The identification result is used to characterize whether the learning behavior of the student to be identified in learning the course to be identified is active learning.

[0053] Furthermore, the method also includes:

[0054] When the recognition result indicates active learning, relevant courses for the course to be recognized are obtained.

[0055] Based on the relevant courses, course recommendations are made to the students to be identified.

[0056] Thirdly, this application provides a model training apparatus, comprising:

[0057] The first acquisition module is used to acquire sample students and sample courses. The sample courses are the courses that the sample students have studied.

[0058] The tag module is used to determine the sample tags of sample students learning sample courses based on the course type of the sample courses. The course types include required courses and non-required courses. The sample tags are used to indicate whether the learning behavior of sample students in learning sample courses is active learning.

[0059] The training module is used to train the pre-acquired model to obtain a learning behavior recognition model based on the student characteristics of the sample students, the course characteristics of the sample courses, the historical learning record characteristics of the sample students learning the sample courses, and the sample labels.

[0060] Fourthly, this application provides a learning behavior recognition device, comprising:

[0061] The second acquisition module is used to acquire the student to be identified and the course to be identified. The course to be identified is the course that the student to be identified has studied.

[0062] The identification module is used to input the student characteristics of the student to be identified, the course characteristics of the course to be identified, and the historical learning record characteristics of the student to be identified in learning the course to be identified into the learning behavior identification model for identification, and obtain the identification result. The learning behavior identification model is the learning behavior identification model provided in this application. The identification result is used to characterize whether the learning behavior of the student to be identified in learning the course to be identified is active learning.

[0063] Fifthly, this application provides an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0064] The memory stores instructions that the computer executes;

[0065] The processor executes computer execution instructions stored in memory to implement the method provided in this application.

[0066] Sixthly, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the method provided in this application.

[0067] The model training method, learning behavior recognition method, device, equipment, and medium provided in this application acquire sample students and sample courses. Based on the course type of the sample courses, sample labels are determined for the sample students' learning of the sample courses. The course types include compulsory and non-compulsory courses. The sample labels are used to indicate whether the sample students' learning behavior of learning the sample courses is active learning. Based on the student characteristics of the sample students, the course characteristics of the sample courses, the historical learning record characteristics of the sample students' learning of the sample courses, and the sample labels, the pre-acquired model to be trained is trained to obtain a learning behavior recognition model. Through the learning behavior recognition model, the accuracy of learning behavior recognition is improved. Attached Figure Description

[0068] 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.

[0069] Figure 1 A schematic diagram of a model training scenario provided in an embodiment of this application;

[0070] Figure 2 A schematic flowchart illustrating a model training method provided in an embodiment of this application;

[0071] Figure 3 A flowchart illustrating a learning behavior recognition method provided in an embodiment of this application;

[0072] Figure 4 This is a schematic diagram of the structure of a model training device provided in an embodiment of this application;

[0073] Figure 5 This is a schematic diagram of the structure of a learning behavior recognition device provided in an embodiment of this application;

[0074] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0075] 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 concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0076] 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.

[0077] To clearly understand the technical solution of this application, the solutions of the prior art will be described in detail first.

[0078] The enterprise online learning platform is based on Internet technology and adopts an open online learning platform model. With learning resources as the core, it meets the needs of various training scenarios for enterprises, builds an internal training ecosystem for enterprises, and helps enterprises achieve talent leadership.

[0079] With the widespread adoption and deep integration of the internet, online learning platforms have become an important channel for internal education and knowledge sharing. Learner behavior data is one of the main bases for guiding platform operations, and how to conduct effective data analysis is a major challenge for platform operation.

[0080] The modeling process for online learner behavior characteristics aims to analyze learner behavior, acquire and maintain learner preferences, and ultimately form a model that reflects learners' personalized needs, knowledge background, or preferences. This involves acquiring data on learners' interests, needs, and all interactive behaviors, analyzing and summarizing this data to obtain a computational and formatted learner behavior characteristic model, and continuously recording changes in learner behavior, which in turn change the learner behavior characteristic model in response to changes in learner preferences.

[0081] To better analyze employee learning behavior, the inventors discovered that the learners on corporate online learning platforms are primarily company employees. Their learning behaviors on these platforms vary: some learners passively engage in course learning to complete specific tasks, while others spontaneously pursue learning to enhance their domain knowledge. Therefore, based on whether a sample course is mandatory, the learning behavior of sample learners can be categorized. This categorization considers the differences in learning behavior across different course types, determining whether the learners' behavior is active or passive. Furthermore, based on learner characteristics, course characteristics, historical learning records, and these labels, a model is trained to obtain a learning behavior recognition model, improving the accuracy of learning behavior identification.

[0082] The following describes the application scenarios of the model training method provided in the embodiments of this application.

[0083] Figure 1 This is a schematic diagram of a model training scenario provided in an embodiment of this application, such as... Figure 1 As shown, the scenario includes a course platform and a server. The course platform includes multiple courses and students for each course. The server obtains courses from the course platform as sample courses, and then obtains student information from the students of the sample courses as sample students. Based on the course type of the sample courses, the server determines the sample tags of the sample students learning the sample courses. Finally, based on the student characteristics of the sample students, the course characteristics of the sample courses, the historical learning record characteristics of the sample students learning the sample courses, and the sample tags, the server trains the pre-obtained model to obtain a learning behavior recognition model.

[0084] Figure 2 This is a flowchart illustrating a model training method provided in an embodiment of this application, as shown below. Figure 2 As shown, the method includes:

[0085] S201. Obtain sample students and sample courses. Sample courses are the courses that the sample students have studied.

[0086] In this embodiment, in order to obtain a learning behavior recognition model for analyzing students' learning behavior in courses, it is necessary to obtain samples of different students learning different courses, construct a sample set, and then train the model. In this embodiment, the obtained students are used as sample students, and the courses that the sample students have learned are used as sample courses. The courses that have been learned can refer to the courses that the sample students have opened and played when they logged into the course platform in their own name. The sample courses can be obtained from the course history records of the sample students in the course platform's backend.

[0087] One possible approach is to first select a subset of mandatory courses based on the platform's historical operational experience. Using course IDs as the basis, the learning behaviors of corresponding students can be collected as passive learning behaviors. Examples include courses in the professional talent training section. Simultaneously, select a subset of non-mandatory courses. Based on the number of learners for each course, identify courses whose learner ranking falls within a certain range (e.g., 10%-60%) over a specific time period. Using course IDs as the basis, the learning behaviors of corresponding students can be collected as active learning behaviors. Any one of these courses can be used as a sample course, and any student corresponding to a sample course can be used as a sample student.

[0088] S202. Based on the course type of the sample courses, determine the sample labels for the sample students' learning of the sample courses. The course types include required courses and non-required courses. The sample labels are used to indicate whether the learning behavior of the sample students in learning the sample courses is active learning.

[0089] In this embodiment, passive learning courses generally appear as compulsory courses. Compulsory courses can refer to courses that require a certain range of students to learn, such as learning zones, training courses, and mandatory courses. Therefore, when selecting data, it is necessary to determine the course type of the sample course based on its category. For example, if the sample course is a training course, then the course type of the sample course can be determined as a compulsory course. Then, the learning behavior of compulsory courses is determined as passive learning, and other courses are determined as non-compulsory courses. The learning behavior of non-compulsory courses is determined as active learning. Thus, the sample label of the sample course can be determined based on the course type of the sample course.

[0090] Furthermore, considering that some students may actively learn required courses, further adjustments are needed when the sample label indicates passive learning in order to obtain more accurate sample labels. For ease of differentiation, the sample labels obtained above based on the course type of the sample courses are used as the initial sample labels, which are then adjusted.

[0091] Specifically, the initial sample labels for sample students learning the sample courses can be determined based on the course type of the sample courses.

[0092] If the initial sample label is active learning, then the initial sample label will be determined as the sample label of the sample students learning the sample course;

[0093] If the initial sample label is passive learning, then the initial sample label is corrected according to the learning entry point of the sample students learning the sample course to obtain the sample label of the sample students learning the sample course. The learning entry point includes any one of the following: learning zone, training course, open course, live course, and required course.

[0094] In this embodiment, the method for determining the initial sample label can be to first determine the course type of the sample course; if the course type of the sample course is a required course, then the initial sample label of the sample student learning the sample course is passive learning; if the course type of the sample course is a non-required course, then the initial sample label of the sample student learning the sample course is active learning.

[0095] In this embodiment, given that student and course learning data can reflect learning motivation tendencies to some extent, pseudo-passive learning data tuples can be calculated and identified using student and course learning data. Considering that the learning situation of courses within a learning portal has certain patterns, learning data is aggregated according to the learning portal. The learning portal can refer to the channel through which students access courses. Courses within the same learning portal have similar learning situations and exhibit the following two characteristics: first, the completion rate of the corresponding course is relatively high, and the completion rate of courses within the same learning portal is generally high; second, passive learners also have a high completion rate for other courses within the same learning portal. Based on the above analysis, the relevant information of the learning portal is crucial for correcting passive learning. Therefore, by obtaining the learning portals of sample courses, the initial sample labels are corrected based on the relevant data of the learning portals.

[0096] Furthermore, the initial sample labels can be corrected using corresponding methods based on the two features.

[0097] In the first approach, if the initial sample label is passive learning, then the initial sample label is modified based on the learning entry point of the sample learners in the sample course to obtain the sample label of the sample learners in the sample course, which may include:

[0098] Based on the learning entry point of the sample students to learn the sample courses, obtain the first set of all courses within the learning entry point;

[0099] Determine the first average completion rate of the first course set based on the completion rate of each student in each course in the first course set;

[0100] Based on all students who have achieved the learning objectives in each course in the first course set, the number of students who have achieved the learning objectives, and the total number of times all students who have achieved the learning objectives have completed their first course, the average completion rate of the first course set is determined. The average completion rate is the average of the average completion rates of all courses in the first course set.

[0101] The first indicator value is obtained by comparing the completion rate and number of times the sample students studied the sample courses within the preset learning period, the average completion rate per course and the first average completion rate, and the average completion rate per course with the preset first threshold.

[0102] If the value of the first indicator is less than the preset threshold of the first indicator, then active learning will be identified as the sample label for sample students learning sample courses.

[0103] In this embodiment, each student in each course can refer to a student who has engaged in learning activities at the learning entry point and whose learning volume has reached a threshold. Students can be filtered using a formula:

[0104]

[0105] in, For S i Learning C within T j Total study time, J p To determine the number of courses available within the learning portal, C j The course duration is set, and β is an adjustment coefficient, where 0 ≤ β ≤ 1. It can be set as needed. For example, taking β = 0.2 means that the current student has completed 20% of the course time within the learning portal.

[0106] The completion rate refers to the ratio of the total learning time to the total course duration for students who have studied the same course through the same learning portal within a preset time period. For example, if student A studied course C twice through learning portal B, with the first session lasting 20 minutes and the second session lasting 30 minutes, assuming there is a progress record after the first session, and the second session builds upon that, the total learning time is 50 minutes. Since course C has a total course duration of 60 minutes, student A's completion rate for course C through learning portal B is 5 / 6. After obtaining the completion rate for each student in each course within the first course set, the average completion rate for each course can be obtained. Then, the average completion rate for the first course set is calculated based on the average completion rate for each course. For clarity, this first average completion rate is used as the average completion rate for the first course set. For example, the first course set is the course set of learning entry point A. The first course set includes course 1 and course 2. Course 1 includes student 1 and student 2, and course 2 includes student 3 and student 4. The completion rate of student 1 learning course 1 through learning entry point A is 0.7, the completion rate of student 2 learning course 1 through learning entry point A is 0.8, the completion rate of student 3 learning course 2 through learning entry point A is 0.8, and the completion rate of student 4 learning course 2 through learning entry point A is 0.9. The average completion rate of course 1 is 0.75, the average completion rate of course 2 is 0.85, and the average completion rate of the first course set is 0.8.

[0107] Students who have achieved the learning target can be defined as those whose completion rate reaches a preset completion rate threshold, such as 90%. The number of students who have achieved the learning target is the number of students who have reached the target. The total number of times each student who has reached the target completes the course is the first total number of times they have studied. For example, if there is a student who has reached the learning target for course 1, and student 2 has achieved the target, and student 1 completes the target completion rate for course 1 in 3 sessions, and student 2 completes the target completion rate for course 1 in 2 sessions, then the first total number of times they have studied for course 1 is 5. Considering that different learning portals may include the same course, to avoid data errors, the number of students who have reached the target and the first total number of times they have studied only include learning behaviors generated through this learning portal, and are different from learning behaviors generated through other learning portals.

[0108] Furthermore, after obtaining the total number of students who have achieved the learning objectives for each course in the first course set, the total number of students who have achieved the learning objectives, and the total number of times all students who have achieved the learning objectives have completed their first course, the average completion rate per course can be obtained. Then, based on the average of the average completion rates per course, the average completion rate per course for the first course set can be obtained. Specific methods may include:

[0109] Get all students who have achieved the learning goals for each course in the first course set, the number of students who have achieved the learning goals, and the first total number of learning sessions for all students who have achieved the learning goals;

[0110] The average completion rate for each course is determined by the ratio of the number of students who meet the standard to the total number of times the course is completed.

[0111] The average completion rate of the first course set is determined based on the average completion rate of each course.

[0112] In this embodiment, after obtaining the number of students who met the requirements and the first total number of learning sessions for each course in the first course set, the average completion rate per learning session for each course can be obtained by using the ratio of the number of students who met the requirements to the first total number of learning sessions. This is expressed as the average completion rate per session. For example, if the number of students who met the requirements is 3 and the first total number of learning sessions is 6, the ratio of the number of students who met the requirements to the first total number of learning sessions is 0.5, which means that the average completion rate per learning session is 50% of the preset completion rate threshold. Based on the average of the average completion rates per session for all courses in the first course set, the average average completion rate per session for the first course set can be obtained.

[0113] After obtaining the first average completion rate and average completion rate per session for the first set of courses, it can be compared with the learning data of sample students. For ease of comparison, it is necessary to obtain the completion rate and number of times sample students studied sample courses within a preset learning period. Based on the comparison results obtained by comparing the sample students' completion rate and number of times they studied sample courses within the preset learning period, the average completion rate per session of sample courses, and the first average completion rate of the first set of courses, it is determined whether the learning situation of the sample students matches the overall situation of the first set of courses. Then, based on the comparison results obtained by comparing the average completion rate per session of the first set of courses with a preset first threshold, it is determined whether the learning situation of the first set of courses meets the preset requirements. Finally, the two comparison results are combined to obtain the first indicator value. The preset first indicator threshold is, for example, 2. Based on the relationship between the first indicator value and the preset first indicator threshold, it is determined whether to adjust the initial sample label. If the first indicator value is less than the preset first indicator threshold, then active learning is determined as the sample label for sample students studying sample courses. If the first indicator value is greater than or equal to the preset first indicator threshold, then the initial sample label is determined as the sample label for sample students studying sample courses.

[0114] Specifically, the first indicator value can also satisfy:

[0115]

[0116] Among them, f1 k S is the first indicator value. i For the i-th sample student, C j For S iThe j-th sample course is studied, where T is the preset time period. For S i Learning C within T j Total study time C j Course duration, For S i Learning C within T j The completion rate, where α is the adjustment coefficient. For S i Learning C within T j Number of times of learning Sample Course C j The first total number of times you learn within the learning portal. Sample Course C j The number of people within the learning portal C j The average completion rate is denoted as A, where A is the first average completion rate, and J is the second average completion rate. p B represents the number of courses within the learning portal, where B is the preset first threshold, and I... p This refers to the number of learners who have engaged in learning activities within the learning portal and whose learning volume has reached the threshold.

[0117] In this embodiment, there may be cases where sample learners have not yet completed the sample course within the preset time period. However, the completion rate of each learning session for the sample learners is high. For example, if a sample learner studies the sample course once and achieves a completion rate of 60% for that single session, which is higher than the average completion rate of 50% for the sample course, it indicates that the learning performance of the sample learners is consistent with the overall performance of the sample course. Therefore, the first average completion rate A is adjusted by multiplying the average completion rate of the sample course by the number of times the sample learner studies the sample course, making the obtained first indicator value more accurate. The average completion rate is the first threshold, which is a preset parameter used to compare the average completion rate. Generally, the completion rate of passive learning courses is higher, for example, let B = 80%.

[0118] In one possible implementation, if f1 k <2 indicates that the sample students' learning behavior in the sample course did not meet the expectations of passive learning, but rather was an active choice. Although they only watched a part of the sample course and then found it uninteresting and did not finish watching it, resulting in a low completion rate, it can be seen that the sample students' learning of the sample course was not passive, nor was it a course-brushing activity to complete a task. In this case, the initial sample label needs to be changed from passive learning to active learning.

[0119] In the second approach, if the initial sample label is passive learning, then the initial sample label is modified based on the learning entry point of the sample learners in the sample course, resulting in sample labels for the sample learners in the sample course, including:

[0120] Determine whether the sample students are new students;

[0121] If the sample student is a new student, the initial sample label will be determined as the sample label of the sample student learning the sample course;

[0122] If the sample students are returning students, the learning entry point for the sample courses is used to correct the initial sample labels, resulting in sample labels for the sample courses learned by the sample students.

[0123] In this embodiment, since passive learners have a high completion rate for their respective courses, or a high overall completion rate (even reaching 100%) for all courses within the learning portal, if a learner's overall completion rate for all courses within the learning portal is low, and the average completion rate is significantly lower than the average completion rate for other learners, then the learner's learning behavior is considered to deviate from the typical characteristics of passive learning. In this case, considering the low learning volume and completion rate of new learners, the comparison results may differ significantly from the actual situation; therefore, no initial sample label correction is performed.

[0124] Specifically, methods for determining whether a sample student is a new student may include:

[0125] The average number of learning sessions per student in the first course set is obtained by comparing the ratio of the second total number of learning sessions to the total number of students in each course within the learning portal.

[0126] The second indicator value is obtained based on the total number of times the sample students studied all courses in the first course set and the average number of times they studied per student.

[0127] If the value of the second indicator is less than the preset threshold of the second indicator, the sample student is determined to be a new student.

[0128] If the value of the second indicator is greater than or equal to the preset threshold of the second indicator, the sample student is determined to be a returning student.

[0129] In this embodiment, by obtaining the average number of times each student learns in the first course set, and then comparing the total number of times each student learns all courses in the first course set with the average number of times each student learns in the first course set, a second indicator value is obtained. If the second indicator value is less than a preset second indicator threshold, it means that the number of times each student learns in the sample is much lower than the average number of times each student learns in other students, and the student can be identified as a new student of the learning entry point.

[0130] Specifically, the second indicator value can satisfy:

[0131]

[0132] in, The second indicator value, This is the third total number of learning sessions. C j The second total number of learning sessions, C j Total number of students The average number of learning sessions per person is γ, which is an adjustment coefficient that can be set as needed, for example, γ = 0.8.

[0133] Furthermore, if the sample learners are returning students, the initial sample tags are modified based on the learning entry points they used to access the sample courses, resulting in sample tags for the sample learners' access to the sample courses. These tags may include:

[0134] Obtain the second set of courses already studied by the sample learners within the learning portal;

[0135] Based on the completion rate of each course studied in the second course set by the sample students, determine the second average completion rate of the second course set by the sample students.

[0136] The third indicator value is obtained based on the completion rate and number of times the sample students studied the sample courses within the preset learning period, the average completion rate of the sample courses per session, the first average completion rate of the first set of courses within the learning portal, and the second average completion rate.

[0137] If the value of the third indicator is less than the preset threshold for the third indicator, then active learning will be identified as the sample label for the sample students learning the sample course.

[0138] In this embodiment, a second set of courses already studied by the sample students is obtained through their learning records on the course platform. These studied courses refer to those courses that the sample students have accessed and played when they logged into the platform under their own names. A third indicator value is obtained by comparing the sample students' completion rate and number of times they studied the sample courses within a preset learning period, the average completion rate per course, and the first average completion rate with the second average completion rate. This comprehensive consideration of the comparison between the sample students' completion rate and the first average completion rate of the first course set, as well as the sample students' learning of courses within the learning platform, makes the obtained third indicator value more accurate and better reflects the learning behavior characteristics of the sample students.

[0139] Specifically, the third indicator value can satisfy:

[0140]

[0141] Among them, f2 k J1 is the third indicator value. p The number of courses in the second course set. This is the second-highest average completion rate.

[0142] In this embodiment, the threshold value of the third indicator can be 1, when f2 k If the score is <1, it can be considered that the learning behavior of the sample students in learning the sample courses does not meet the expectations of passive learning. The sample students did not show a high completion rate for courses within the same learning portal. It can be seen that the sample students' learning of the sample courses is not passive, nor is it to complete the course by scrambling for the purpose of completing the task. In this case, the initial sample label needs to be changed from passive learning to active learning.

[0143] S203. Based on the student characteristics of the sample students, the course characteristics of the sample courses, the historical learning record characteristics of the sample students learning the sample courses, and the sample labels, train the pre-acquired model to obtain the learning behavior recognition model.

[0144] In this embodiment, student characteristics may include attributes such as student ID, age, gender, job title, major, and education level, as well as attributes such as the student's total learning time, total number of learning sessions, completion rate, average learning interval, latest learning time, and percentage of learning time for special category courses within a time period T. Course characteristics may include attributes such as course ID, course duration, course category, course launch time, and course keywords, as well as attributes such as the course's total learning time, total number of learning sessions, completion rate, average learning time per student, and average number of learning sessions per student within a time period T. Historical learning record characteristics may include attributes such as the student's total learning time, total number of learning sessions, total learning progress, last learning time, current learning time, current learning duration, and learning entry point within a preset time period T.

[0145] The model to be trained can be used to learn the features of the sample courses, in order to learn the changing relationships between the course features, student features, historical learning record features, and sample labels, thereby predicting whether the learning behavior of the students is active learning. The model to be trained can refer to any possible model, such as neural network models, data mining software, etc., without any restrictions here.

[0146] The model training method provided in this application obtains sample students and sample courses. Based on the course type of the sample courses, it determines the sample labels of the sample students learning the sample courses. The course types include compulsory courses and non-compulsory courses. The sample labels are used to indicate whether the learning behavior of the sample students in learning the sample courses is active learning. Based on the student characteristics of the sample students, the course characteristics of the sample courses, the historical learning record characteristics of the sample students in learning the sample courses, and the sample labels, the pre-obtained model to be trained is trained to obtain a learning behavior recognition model. Through the learning behavior recognition model, the accuracy of learning behavior recognition is improved.

[0147] Figure 3 This is a flowchart illustrating a learning behavior recognition method provided in an embodiment of this application, as shown below. Figure 3 As shown, the method includes:

[0148] S301. Obtain the student to be identified and the course to be identified. The course to be identified is the course that the student to be identified has studied.

[0149] In this embodiment, when it is necessary to identify the learning behavior of the student to be identified in the course platform, the relevant information of the student to be identified and the relevant information of the course to be identified are obtained from the course platform.

[0150] S302. Input the student characteristics of the student to be identified, the course characteristics of the course to be identified, and the historical learning record characteristics of the student to be identified in learning the course to be identified into the learning behavior recognition model for recognition, and obtain the recognition result.

[0151] In this embodiment, the learning behavior recognition model is the learning behavior recognition model provided in this application, and the recognition result is used to characterize whether the learning behavior of the student to be identified in learning the course to be identified is active learning.

[0152] In some implementations, when the identification result is active learning, relevant courses for the course to be identified are obtained, and course recommendations are made to the student to be identified based on these relevant courses. Alternatively, the course to be identified can be added to an active learning list, allowing for the construction of a learning profile for the student to be identified based on the courses in the active learning list. This helps determine the courses the student is interested in, leading to better course recommendations and improved operational effectiveness of the course platform.

[0153] The learning behavior recognition method provided in this application embodiment obtains the student to be identified and the course to be identified. The course to be identified is the course that the student to be identified has studied. Then, the student characteristics of the student to be identified, the course characteristics of the course to be identified, and the historical learning record characteristics of the student to be identified in learning the course to be identified are input into the learning behavior recognition model for recognition, and the recognition result is obtained, which improves the accuracy of recognizing the student's learning behavior in learning courses.

[0154] Figure 4 This is a schematic diagram of the structure of a model training device provided in an embodiment of this application, as shown below. Figure 4 As shown, the device includes:

[0155] The first acquisition module 401 is used to acquire sample students and sample courses, where the sample courses are the courses that the sample students have studied.

[0156] The tag module 402 is used to determine the sample tags of the sample students learning the sample courses based on the course type of the sample courses. The course types include compulsory courses and non-compulsory courses. The sample tags are used to indicate whether the learning behavior of the sample students in learning the sample courses is active learning.

[0157] Training module 403 is used to train a pre-acquired model to obtain a learning behavior recognition model based on the student characteristics of the sample students, the course characteristics of the sample courses, the historical learning record characteristics of the sample students learning the sample courses, and the sample labels.

[0158] In some implementations, the tag module 402 is also used for:

[0159] Based on the course type of the sample courses, determine the initial sample labels for the sample students learning the sample courses;

[0160] If the initial sample label is active learning, then the initial sample label will be determined as the sample label of the sample students learning the sample course;

[0161] If the initial sample label is passive learning, then the initial sample label is corrected according to the learning entry point of the sample students learning the sample course to obtain the sample label of the sample students learning the sample course. The learning entry point includes any one of the following: learning zone, training course, open course, live course, and required course.

[0162] In some implementations, the tag module 402 is also used for:

[0163] Determine the course type of the sample courses;

[0164] If the course type of the sample course is a required course, then the initial sample label for the sample students learning the sample course is passive learning.

[0165] If the sample course is a non-required course, then the initial sample label for the sample students learning the sample course is active learning.

[0166] In some implementations, the tag module 402 is also used for:

[0167] Based on the learning entry point of the sample students to learn the sample courses, obtain the first set of all courses within the learning entry point;

[0168] Determine the first average completion rate of the first course set based on the completion rate of each student in each course in the first course set;

[0169] Based on all students who have achieved the learning objectives in each course in the first course set, the number of students who have achieved the learning objectives, and the total number of times all students who have achieved the learning objectives have completed their first course, the average completion rate of the first course set is determined. The average completion rate is the average of the average completion rates of all courses in the first course set.

[0170] The first indicator value is obtained by comparing the completion rate and number of times the sample students studied the sample courses within the preset learning period, the average completion rate per course and the first average completion rate, and the average completion rate per course with the preset first threshold.

[0171] If the value of the first indicator is less than the preset threshold of the first indicator, then active learning will be identified as the sample label for sample students learning sample courses.

[0172] In some implementations, the tag module 402 is also used for:

[0173] Get all students who have achieved the learning goals for each course in the first course set, the number of students who have achieved the learning goals, and the first total number of learning sessions for all students who have achieved the learning goals;

[0174] The average completion rate for each course is determined by the ratio of the number of students who meet the standard to the total number of times the course is completed.

[0175] The average completion rate of the first course set is determined based on the average completion rate of each course.

[0176] In some implementations, the tag module 402 is also used to satisfy:

[0177]

[0178]

[0179] Among them, f1 k S is the first indicator value. i For the i-th sample student, C j For S i The j-th sample course is studied, where T is the preset time period. For S i Learning C within T j Total study time C j Course duration, For S i Learning C within T j The completion rate, where α is the adjustment coefficient. For S i Learning C within T j Number of times of learning C j The first total number of times you learn within the learning portal. C j The number of people within the learning portal C j The average completion rate is denoted as A, where A is the first average completion rate, and J is the second average completion rate. p B represents the number of courses within the learning portal, where B is the preset first threshold, and I... p This refers to the number of learners who have engaged in learning activities within the learning portal and whose learning volume has reached the threshold.

[0180] In some implementations, the tag module 402 is also used for:

[0181] Determine whether the sample students are new students;

[0182] If the sample student is a new student, the initial sample label will be determined as the sample label of the sample student learning the sample course;

[0183] If the sample learners are returning learners, the initial sample labels are modified based on the learning entry points of the sample learners for the sample courses, resulting in sample labels for the sample learners' learning of the sample courses.

[0184] In some implementations, the tag module 402 is also used for:

[0185] The average number of learning sessions per student in the first course set is obtained by comparing the ratio of the second total number of learning sessions to the total number of students in each course within the learning portal.

[0186] The second indicator value is obtained based on the total number of times the sample students studied all courses in the first course set and the average number of times they studied per student.

[0187] If the value of the second indicator is less than the preset threshold of the second indicator, the sample student is determined to be a new student.

[0188] If the value of the second indicator is greater than or equal to the preset threshold of the second indicator, the sample student is determined to be a returning student.

[0189] In some implementations, the tag module 402 is also used to satisfy:

[0190]

[0191] in, The second indicator value, This is the third total number of learning sessions. C j The second total number of learning sessions, C j Total number of students γ represents the average number of learning sessions per person for the second person, and γ is the adjustment coefficient.

[0192] In some implementations, the tag module 402 is also used for:

[0193] Obtain the second set of courses already studied by the sample learners within the learning portal;

[0194] Based on the completion rate of each course studied in the second course set by the sample students, determine the second average completion rate of the second course set by the sample students.

[0195] The third indicator value is obtained based on the completion rate and number of times the sample students studied the sample courses within the preset learning period, the average completion rate of the sample courses per session, the first average completion rate of the first set of courses within the learning portal, and the second average completion rate.

[0196] If the value of the third indicator is less than the preset threshold for the third indicator, then active learning will be identified as the sample label for the sample students learning the sample course.

[0197] In some implementations, the tag module 402 is also used to satisfy:

[0198]

[0199] Among them, f2 k J1 is the third indicator value. p The number of courses in the second course set. This is the second-highest average completion rate.

[0200] Figure 5 This is a schematic diagram of the structure of a learning behavior recognition device provided in an embodiment of this application, as shown below. Figure 5 As shown, the device includes:

[0201] The second acquisition module 501 is used to acquire the student to be identified and the course to be identified, wherein the course to be identified is the course that the student to be identified has studied.

[0202] The identification module 502 is used to input the student characteristics of the student to be identified, the course characteristics of the course to be identified, and the historical learning record characteristics of the student to be identified in learning the course to be identified into the learning behavior identification model for identification, and obtain the identification result. The learning behavior identification model is the learning behavior identification model provided in this application. The identification result is used to characterize whether the learning behavior of the student to be identified in learning the course to be identified is active learning.

[0203] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 6 As shown, the electronic device 60 includes:

[0204] The electronic device 60 may include a processor 601 with one or more processing cores, a memory 602 with one or more computer-readable storage media, a communication component 603, and other components. The processor 601, memory 602, and communication component 603 are connected via a bus 604.

[0205] In the specific implementation process, at least one processor 601 executes computer execution instructions stored in memory 602, causing at least one processor 601 to execute the above-mentioned model training method or learning behavior recognition method.

[0206] The specific implementation process of processor 601 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0207] In the above Figure 6 In the illustrated embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0208] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0209] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0210] In some embodiments, a computer program product is also provided, comprising a computer program or instructions that, when executed by a processor, implement the steps in any of the model training methods or learning behavior recognition methods described above.

[0211] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0212] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0213] Therefore, embodiments of this application provide a computer-readable storage medium storing a plurality of instructions that can be loaded by a processor to execute steps in any of the model training methods or learning behavior recognition methods provided in embodiments of this application.

[0214] The storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0215] According to one aspect of this application, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium.

[0216] Since the instructions stored in the storage medium can execute the steps in any of the model training methods or learning behavior recognition methods provided in the embodiments of this application, the beneficial effects that any of the model training methods or learning behavior recognition methods provided in the embodiments of this application can achieve can be realized. For details, please refer to the previous embodiments, which will not be repeated here.

[0217] 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.

[0218] 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 model training method, characterized in that, The method comprises the following steps: obtaining a sample student and a sample course, the sample course being a course learned by the sample student; determining a sample label of the sample student learning the sample course according to a course type of the sample course, the course type comprising a compulsory course and a non-compulsory course, the sample label being used to indicate whether a learning behavior of the sample student learning the sample course is active learning; performing model training on a pre-obtained to-be-trained model according to a student feature of the sample student, a course feature of the sample course, a historical learning record feature of the sample student learning the sample course, and the sample label, to obtain a learning behavior recognition model.

2. The method of claim 1, wherein, The step of determining the sample label of the sample student learning the sample course according to the course type of the sample course comprises the following steps: determining an initial sample label of the sample student learning the sample course according to the course type of the sample course; if the initial sample label is active learning, determining the initial sample label as the sample label of the sample student learning the sample course; if the initial sample label is passive learning, correcting the initial sample label according to a learning portal of the sample student learning the sample course to obtain the sample label of the sample student learning the sample course, the learning portal comprising any one of a learning zone, a training class, a public course, a live course, and a compulsory course.

3. The method of claim 2, wherein, The step of determining the initial sample label of the sample student learning the sample course according to the course type of the sample course comprises the following steps: determining the course type of the sample course; if the course type of the sample course is a compulsory course, the initial sample label of the sample student learning the sample course is passive learning; if the course type of the sample course is a non-compulsory course, the initial sample label of the sample student learning the sample course is active learning.

4. The method of claim 2, wherein, The step of correcting the initial sample label according to the learning portal of the sample student learning the sample course if the initial sample label is passive learning to obtain the sample label of the sample student learning the sample course comprises the following steps: obtaining a first course set of all courses in the learning portal according to the learning portal of the sample student learning the sample course; determining a first average completion rate of the first course set according to a completion rate of each student in each course in the first course set; determining an average average completion rate of the first course set according to all learning qualified students of each course in the first course set, a number of the learning qualified students, and a first total number of learning times of the learning qualified students, the average average completion rate being an average value of average completion rates of all courses in the first course set; obtaining a first index value according to a completion rate and a number of learning times of the sample student learning the sample course in a preset learning period, a comparison result of the average average completion rate and the first average completion rate, and a comparison result of the average average completion rate and a preset first threshold value. If the first index value is less than a preset first index threshold, the active learning is determined as a sample label of the sample student learning the sample course.

5. The method of claim 4, wherein, According to all learning qualified students of each course in the first course set, the number of the all learning qualified students and the first total learning times of the all learning qualified students, the average per-course completion rate of the first course set is determined, including: According to all learning qualified students of each course in the first course set, the number of the all learning qualified students and the first total learning times of the all learning qualified students, the average per-course completion rate of the first course set is determined, including: According to the ratio of the number of the all learning qualified students and the first total learning times, the per-course completion rate of each course is determined; According to the per-course completion rate of each course, the average per-course completion rate of the first course set is determined.

6. The method of claim 4, wherein, The comparison result of the sample student learning the sample course in a preset learning period, the sample course completion rate and the first average completion rate, and the comparison result of the average per-course completion rate and a preset first threshold value are obtained, and the first index value is obtained, satisfying: wherein f1 k is the first index value, S i is the i-th sample student, C j is the S i j-th sample course of learning, T is a preset time period, is the total learning time of S i in T for C j , is the course duration of C j , is the completion rate of S i in T for C j , a is an adjustment coefficient, is the learning frequency of S i in T for C j , is the first total learning frequency of C j in the learning portal, is the number of C j in the learning portal, is the secondary average completion rate of C j , A is the first average completion rate, J p is the number of courses in the learning portal, B is the preset first threshold value, I p is the number of learners who have learning behavior and learning amount reaching the threshold value in the learning portal.

7. The method of claim 2, wherein, If the initial sample label is passive learning, the initial sample label is corrected according to the learning entrance of the sample student learning the sample course, and the sample label of the sample student learning the sample course is obtained, including: Determine whether the sample student is a new student; If the sample student is a new student, the initial sample label is determined as the sample label of the sample student learning the sample course; If the sample student is an old student, the initial sample label is corrected according to the learning entrance of the sample student learning the sample course, and the sample label of the sample student learning the sample course is obtained.

8. The method of claim 7, wherein, The determination of whether the sample student is a new student includes: According to the ratio of the second total learning times and the total number of learning persons of each course in the first course set in the learning entrance, the average per-course learning times of the first course set is obtained; According to the third total learning times of the sample student learning all courses in the first course set and the average per-course learning times, a second index value is obtained; If the second index value is less than a preset second index threshold, the sample student is determined to be a new student; If the second index value is greater than or equal to a preset second index threshold, the sample student is determined to be an old student.

9. The method of claim 8, wherein, The second index value is obtained according to the third total learning times of the sample student learning all courses in the first course set and the second average per-course learning times, satisfying: wherein, is the second index value, is the third total number of learning, is C j the second total number of learning, is C j the total number of learners, is the second number of learning per person, and γ is an adjustment coefficient.

10. The method of claim 7, wherein, If the sample student is an old student, the initial sample label is corrected according to the learning entrance of the sample student learning the sample course, and the sample label of the sample student learning the sample course is obtained, including: A second course set of the learned courses of the sample student in the learning entrance is obtained; According to the completion rate of the sample student learning each learned course in the second course set, a second average completion rate of the sample student learning the second course set is determined; According to the completion rate and the learning times of the sample course learned by the sample student within a preset learning period, the average completion rate of the sample course, the first average completion rate of the first course set in the learning portal, and the second average completion rate, a third index value is obtained; If the third index value is less than a preset third index threshold, the active learning is determined as the sample label of the sample student learning the sample course.

11. The method of claim 10, wherein, The third index value is obtained according to the completion rate and the learning times of the sample course learned by the sample student within a preset learning period, the average completion rate of the sample course, the first average completion rate of the first course set in the learning portal, and the second average completion rate, which satisfies: wherein f2 k is the third index value, J1 p is the number of courses in the second course set, is the second average completion rate.

12. A learning behavior recognition method characterized by, It comprises: Obtain a to-be-identified student and a to-be-identified course, the to-be-identified course being a course learned by the to-be-identified student; Input the student features of the to-be-identified student, the course features of the to-be-identified course, and the historical learning record features of the to-be-identified student learning the to-be-identified course into a learning behavior recognition model for recognition to obtain a recognition result, the learning behavior recognition model being the learning behavior recognition model of any one of claims 1-11, and the recognition result being used to represent whether the learning behavior of the to-be-identified student learning the to-be-identified course is active learning.

13. The method of claim 12, wherein, The method further comprises: When the recognition result is active learning, obtain a related course of the to-be-identified course; According to the related course, recommend a course to the to-be-identified student.

14. A model training apparatus, comprising: It comprises: A first obtaining module is configured to obtain a sample student and a sample course, the sample course being a course learned by the sample student; A label module is configured to determine a sample label of the sample student learning the sample course according to a course type of the sample course, the course type including a compulsory course and a non-compulsory course, and the sample label being used to indicate whether the learning behavior of the sample student learning the sample course is active learning; A training module is configured to perform model training on a to-be-trained model pre-obtained according to student features of the sample student, course features of the sample course, historical learning record features of the sample student learning the sample course, and the sample label, to obtain a learning behavior recognition model.

15. A learning behavior recognition apparatus characterized by comprising: It comprises: A second obtaining module is configured to obtain a to-be-identified student and a to-be-identified course, the to-be-identified course being a course learned by the to-be-identified student; An identification module is configured to input student features of the to-be-identified student, course features of the to-be-identified course, and historical learning record features of the to-be-identified student learning the to-be-identified course into a learning behavior recognition model for recognition to obtain a recognition result, the learning behavior recognition model being the learning behavior recognition model of any one of claims 1-11, and the recognition result being used to represent whether the learning behavior of the to-be-identified student learning the to-be-identified course is active learning.

16. An electronic device, comprising: It comprises: A processor, and a memory in communication connection with the processor; The memory stores computer execution instructions; The processor executes computer-executable instructions stored in the memory to implement the method of any of claims 1-13.

17. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions that, when executed by a processor, implement the method of any of claims 1-13.

Images