Online education method and system based on artificial intelligence
The AI-based online education method addresses inaccuracies in course recommendations by analyzing learners' behavior data to quantify connections and feedback, resulting in personalized and efficient course suggestions.
Patent Information
- Application Number
- CN202411101893.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-12
- Publication Date
- 2024-12-03
AI Technical Summary
Existing online education systems suffer from inaccurate course recommendations due to the lack of consideration for individual learners' varying learning behavior feedback on different courses, leading to discrepancies between recommended and actual learning outcomes.
An online education method utilizing artificial intelligence to analyze learners' behavior data, quantify connection factors and feedback parameters between courses, and perform clustering to recommend courses based on personalized learning behavior similarities and feedback.
Improves the accuracy of course recommendations by considering individual learning behavior feedback, enhancing the overall learning efficiency of online education learners.
Smart Images

Figure CN119067811A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology, and in particular to an online education method and system based on artificial intelligence. Background Art
[0002] With the development of social economy and the rise of knowledge economy, people's demand for education is becoming stronger and stronger, and the requirements for education quality and educational resources are also getting higher and higher. However, there are many problems and shortcomings in traditional education models and teaching methods, which need to be reformed and innovated. With the rapid development of information technologies such as the Internet, mobile Internet, big data, and artificial intelligence, online education has provided learners with a rich variety of vocational training and skills improvement opportunities, and has become an emerging educational method. Online education breaks the time and space limitations of traditional education. Learners can choose learning content and learning methods according to their own schedule and learning rhythm. At the same time, by analyzing the user's learning behavior and making corresponding course recommendations, learners can obtain a more comprehensive and personalized learning experience. This personalized recommendation can be adjusted according to the learner's learning progress, preferences, and abilities, making the learning process more efficient and targeted.
[0003] In the existing technology, collaborative filtering algorithm is a commonly used online course recommendation method. Its essence is to select related courses for recommendation based on the course similarity and project relevance of different online learning users, but it does not take into account the feedback effect of different learners on the learning status of different courses. As a result, there are large actual differences between the recommended courses and the actual course learning of the learners, which makes the accuracy of course recommendations insufficient. Summary of the invention
[0004] The present invention provides an online education method and system based on artificial intelligence to solve the problem that the learning behavior feedback of different courses in existing online education is different, resulting in poor course recommendation effect. The technical solution adopted is as follows:
[0005] The present invention proposes an online education method based on artificial intelligence, which comprises the following steps:
[0006] Collect the learning behavior data of several learners in online education for the courses corresponding to each learning process, and obtain the learning course sequence of each learner;
[0007] According to the distribution of different courses in each learner's learning course sequence, the connection factors between different courses are obtained;
[0008] According to the learning behavior data of different learners in the process of learning each course, the learning behavior feedback parameters of each course are obtained;
[0009] According to the differences in learning behavior feedback parameters of different courses, the learning behavior similarities between different courses are obtained; combined with the connection factors between different courses, the final connection parameters between different courses are obtained;
[0010] According to the final connection parameters between different courses, a relevant recommendation set for each course is obtained; and courses are recommended to each learner based on the relevant recommendation set for each course.
[0011] Optionally, the collecting of learning behavior data of a number of learners for a course corresponding to each learning process in online education includes the following specific methods:
[0012] For any course corresponding to any learning process of any learner, all viewing behaviors of the course during the learning process are collected, including the rated duration of the course, the total skipping time of the learner in the process of learning the course, the total number of pauses of the learner in the process of learning the course, and the total repeated learning time of the learner in the process of learning the course, which are comprehensively used as the learner's learning behavior data for the course.
[0013] Optionally, the obtaining of the learning course sequence of each learner includes the following specific methods:
[0014] For any learner, the courses corresponding to each learning process of the learner are taken as a number of learning course data of the learner, and all the learning course data are arranged in the order of the learning process to obtain the learning course sequence of the learner.
[0015] Optionally, the method of obtaining the connection factors between different courses includes:
[0016]
[0017] In the formula, CT n,n′ represents the connection factor between the nth course and the n′th course; M n,n′ It represents the total number of times the nth course and the n′th course appear in the learning course sequence of the same learner among all M learning course sequences; M represents the total number of learning course sequences corresponding to all learners.
[0018] Optionally, the specific method of obtaining the learning behavior feedback parameter of each course includes:
[0019]
[0020] In the formula, CS n represents the learning behavior feedback parameter of the nth course; N represents the total number of different courses in the learning course sequence of all learners; represents the mean of the total number of pauses of all learners in the process of learning the nth course; represents the mean of the total number of pauses of all learners in the process of learning the n′th course; RT m′,n M′ represents the total duration of repeated learning for the m′th learner who has repeated learning behavior in the process of learning the nth course; n represents the total number of learners who have repeated learning behaviors in the process of learning the nth course; T n represents the rated duration of the nth course; JT m″,n M′ represents the total skipping time of the m″th learner who has skipping learning behavior in the process of learning the nth course; n ′ represents the total number of learners who skip learning behavior during the process of learning the nth course.
[0021] Optionally, the learning behavior similarity between the different courses is obtained by:
[0022]
[0023] In the formula, CY n,n′ represents the similarity parameter of learning behavior between the nth course and the n′th course; CS n represents the learning behavior feedback parameter of the nth course; CS n′ represents the learning behavior feedback parameter of the n′th course; CS n″ represents the learning behavior feedback parameter of the n″th course; N represents the total number of different courses in the learning course sequence of all learners;
[0024] Obtain the learning behavior similarity parameters between the nth course and each other course except the nth course, perform linear normalization on all the obtained learning behavior similarity parameters, and obtain the result as the learning behavior similarity between the nth course and each other course except the nth course.
[0025] Optionally, the final connection parameters between different courses are obtained by:
[0026]
[0027] In the formula, FR n,n′ Indicates the final connection parameter between the nth course and the n′th course; CT n,n′ Represents the connection factor between the nth course and the n′th course; CY n,n′ represents the similarity of learning behaviors between the nth course and the n′th course; represents the mean of the similarity of learning behaviors between the nth course and all other courses except the nth course; || represents the absolute value function; exp() represents the exponential function with a natural constant as the base.
[0028] Optionally, the relevant recommendation set for each course is obtained by:
[0029] DBSCAN clustering is performed on all courses except the nth course. The distance measurement uses the absolute value of the difference between the final connection parameters between different courses and the nth course to obtain several clusters. The set of all courses in any cluster is taken as a related recommendation set for the nth course, and several related recommendation sets for the nth course are obtained.
[0030] Optionally, the method of recommending courses to each learner based on the relevant recommendation set of each course includes:
[0031] Obtain the mean of the final connection parameters between the nth course and all courses in any related recommendation set, record it as the final connection of the related recommendation set of the nth course, and obtain the final connection of each related recommendation set of the nth course;
[0032] A judgment threshold is preset, and the product of the final maximum value of the connection and the judgment threshold is used as the connection threshold;
[0033] A course whose final connection parameter with the n-th course is greater than or equal to the connection threshold is used as a recommended course for the learner whose learning course sequence includes the n-th course.
[0034] The present invention also proposes an online education system based on artificial intelligence, which includes a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of the above method when executing the computer program.
[0035] The beneficial effects of the present invention are as follows: the present invention analyzes the probability of different courses appearing in the learning course sequence of the same learner, quantifies the connection factors between different courses, reflects the course connection of different courses in the learning process, and serves as the basis for subsequent learning recommendations; analyzes the learning behavior data of learners in the course learning process, based on the relevant data of positive feedback and negative feedback, comprehensively quantifies the learning behavior feedback parameters of the course, reflects the learning difficulty of the course in the learning process, and further expresses the behavioral feedback differences of different courses; analyzes the learning behavior feedback parameters of different courses, quantifies the learning behavior similarity, and further adjusts the connection factors between different courses, so as to The final connection parameters are obtained, and it is ensured that the final connection parameters not only reflect the connection between courses, but also include the quantification of the similarity of behavioral feedback between different courses, so as to ensure that the recommendation based on the final connection parameters can take into account the reference efficiency of learners' different courses to the greatest extent; cluster analysis is performed through the final connection parameters between courses to obtain the relevant recommendation set of each course, and then course recommendations are made to learners based on the relevant recommendation set and the final connection parameters, so as to improve the accuracy of course recommendations in the process of learners' online education learning, avoid low recommendation effects due to different learning behavior feedback, and finally achieve the improvement of learning efficiency of online education learners. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0037] Figure 1 A flowchart of an online education method based on artificial intelligence provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0039] See also Figure 1 , which shows a flow chart of an online education method based on artificial intelligence provided by an embodiment of the present invention, the method comprising the following steps:
[0040] Step S001: Collect learning behavior data of several learners in online education for the courses corresponding to each learning process, and obtain the learning course sequence of each learner.
[0041] The main purpose of this embodiment is to collect the learning behavior data and learning course data of several learners in online education, and to further improve the collaborative filtering algorithm based on accuracy, so as to recommend different courses for online education learners. Therefore, it is necessary to collect the learning behavior data and learning course data of learners first.
[0042] Specifically, this embodiment obtains the learner's learning behavior data and learning course data by utilizing the background learning record library in the online education platform, and collects data from M learners. This embodiment uses M=100 for description, and the implementer can set it to other values in other embodiments; for the course corresponding to any learning process of any learner, all viewing behaviors of the course in the learning process are collected, and the learner's viewing behavior for the course corresponds to a set of data, including the rated duration of the course, the total skipping duration of the learner in the process of learning the course, the total number of pauses of the learner in the process of learning the course, and the total repeated learning duration of the learner in the process of learning the course, which are comprehensively used as the learner's learning behavior data for the course; then for the learner, the learning behavior data of the course corresponding to each learning process of the learner is obtained, and the course corresponding to each learning process of the learner is used as a number of learning course data of the learner, all learning course data are arranged in the order of the learning process, and the learning course sequence of the learner is obtained.
[0043] Furthermore, for each learner, the course corresponding to each learning process is obtained, and a number of learning course data and a learning course sequence of each learner are obtained. At the same time, the learning behavior data of each learner for the course corresponding to each learning process is obtained.
[0044] At this point, we have obtained some learning behavior data and learning course sequences for each learner.
[0045] Step S002: According to the distribution of different courses in the learning course sequence of each learner, the connection factors between different courses are obtained.
[0046] It should be noted that for all elements in the course sequence of any learner, the specific information they represent is the set of all courses that the learner has learned during the online course learning process; and different courses have a reference for mutual learning, that is, when the same learner is learning a specific course, some of the different courses studied by the learner have a certain connection. This connection can serve as the basic information support for subsequent learning recommendations. The connection between different courses reflected in the course sequence of a single learner is not sufficient to explain the significance. Therefore, it is necessary to use the M learning course sequences corresponding to all M learners for analysis in order to obtain the connection factors between different courses.
[0047] Specifically, taking the nth course and the n'th course as an example, the calculation method of the connection factor between the nth course and the n'th course is:
[0048]
[0049] In the formula, CT n,n′ represents the connection factor between the nth course and the n′th course; M n,n′ It represents the total number of times the nth course and the n′th course appear in the learning course sequence of the same learner among all M learning course sequences; M represents the total number of learning course sequences corresponding to all learners.
[0050] It should be noted that the connection factor between the nth course and the nth course is quantified by analyzing the probability that the nth course and the nth′ course appear in the same learning course sequence. The larger the value, the more likely a large proportion of the M learners are studying the nth course and the nth′ course at the same time, and the higher the possibility that the nth course and the nth′ course are connected. Otherwise, the lower the possibility that the nth course and the nth′ course are connected.
[0051] Furthermore, the above calculation method is only a method for calculating the connection factor between the nth course and the n′th course in one embodiment of the present invention. Similarly, this method is used to calculate the connection factors between all courses to obtain the connection factors between all different courses.
[0052] So far, by analyzing the probability of different courses appearing in the learning course sequence of the same learner, the connection factors between different courses are quantified, reflecting the course connection of different courses in the learning process, and serving as the basis for subsequent learning recommendations.
[0053] Step S003: Obtain learning behavior feedback parameters for each course based on the learning behavior data of different learners in the process of learning each course.
[0054] It should be noted that for any course, different courses have different actual feedback received during the actual learning process, and course feedback is a very important influencing factor for course learning and recommendation. For example, one of the most obvious manifestations is that different courses have different difficulty levels, which leads to different learning feedback from learners. Therefore, the specific feedback behavior of learners in the process of learning a certain course is used to judge the specific learning situation of different courses.
[0055] Specifically, taking the nth course as an example, based on the learning behavior data of different learners in learning the nth course, the calculation method for obtaining the learning behavior feedback parameters of the nth course is:
[0056]
[0057] In the formula, CS n represents the learning behavior feedback parameter of the nth course; N represents the total number of different courses in the learning course sequence of all learners, that is, the total number of courses corresponding to all learning processes of all learners (excluding repeated courses); represents the mean of the total number of pauses of all learners in the process of learning the nth course; represents the mean of the total number of pauses of all learners in the process of learning the n′th course; RT m′,n M′ represents the total duration of repeated learning for the m′th learner who has repeated learning behavior in the process of learning the nth course; n represents the total number of learners who have repeated learning behaviors in the process of learning the nth course; T n represents the rated duration of the nth course; JT m″,n M′ represents the total skipping time of the m″th learner who has skipping learning behavior in the process of learning the nth course; n ′ represents the total number of learners who have skipped learning behavior in the process of learning the nth course. It should be noted that if the total time of repeated learning of the learner in the process of learning the nth course is 0, then he does not have repeated learning behavior in the process of learning the nth course. Similarly, if the total skipping time is 0, there is no skipping learning behavior.
[0058] What needs to be explained is that by using the negative feedback of learners in the process of learning the course as the basis, that is, the mean of the total number of pauses and the total time of repeated learning relative to other courses, combined with the positive feedback of learners in the process of learning the course, that is, the mean of the total skipping time, the feedback parameters of learners in the process of learning the course are comprehensively quantified, and then the learning behavior feedback parameters of the course are obtained; in the process of learning the course, the higher the negative feedback of the learners, that is, the greater the mean of the relative total number of pauses, the greater the mean of the total time of repeated learning, and the smaller the positive feedback, that is, the total skipping time, it means that the learners have more difficulty in learning the course and the greater the learning behavior feedback parameters of the course.
[0059] Furthermore, the above calculation method is only a method using the learning behavior feedback parameters of the nth course as an example in one embodiment of the present invention. Similarly, the learning behavior feedback parameters of all courses are calculated using this method to obtain the learning behavior feedback parameters of each course.
[0060] At this point, by analyzing the learners' learning behavior data during the course learning process, based on the relevant data of positive feedback and negative feedback, the learning behavior feedback parameters of the course are comprehensively quantified to reflect the learning difficulty of the course in the learning process, and further show the differences in behavioral feedback of different courses.
[0061] Step S004: According to the differences in learning behavior feedback parameters of different courses, the learning behavior similarities between different courses are obtained; and the final connection parameters between different courses are obtained by combining the connection factors between different courses.
[0062] Preferably, in one embodiment of the present invention, the learning behavior similarity between different courses is obtained according to the difference in learning behavior feedback parameters of different courses, and the specific method includes:
[0063] It should be noted that for any learner, his or her own learning ability tends to be habitual. When learners are studying courses, the learning process will be easier for them for courses with similar feedback behaviors. Therefore, it is necessary to quantify the similarity of learning behaviors based on the learning feedback behavior parameters of different courses.
[0064] Specifically, taking the nth course and the n'th course as an example, using the learning behavior feedback parameters of the nth course and the n'th course, the calculation method for obtaining the learning behavior similarity between the nth course and the n'th course is:
[0065]
[0066] In the formula, CY n,n′ represents the similarity parameter of learning behavior between the nth course and the n′th course; CSn represents the learning behavior feedback parameter of the nth course; CS n′ represents the learning behavior feedback parameter of the n′th course; CS n″ represents the learning behavior feedback parameter of the n″th course; N represents the total number of different courses in the learning course sequence of all learners; the learning behavior similarity parameters between the nth course and each course except the nth course are obtained according to the above method, and all the obtained learning behavior similarity parameters are linearly normalized, and the results are obtained as the learning behavior similarity between the nth course and each course except the nth course.
[0067] What needs to be explained is that by obtaining the ratio of the learning behavior feedback parameters of the two courses, and calculating the difference with the mean ratio of the learning behavior feedback parameters between the nth course and all other courses, the larger the difference is, the more it means that the learner has a more similar feedback behavior in the process of learning the nth course and the process of learning the n′th course, that is, the learner's learning behavior in learning the nth course and the process of learning the n′th course has a high similarity, so that the learner, based on the study of the nth course, is more likely to have a higher efficiency in the process of learning the n′th course, and vice versa.
[0068] Furthermore, the above calculation method is only a method in one embodiment of the present invention that uses the similarity of learning behaviors between the nth course and the n′th course as an example. Similarly, this method can be used to calculate the similarity of learning behaviors between all courses, so as to obtain the similarity of learning behaviors between different courses.
[0069] Preferably, in one embodiment of the present invention, the final connection parameters between different courses are obtained by combining the connection factors between different courses, including the specific method of:
[0070] It should be noted that for different courses, the course recommendation based only on the connection factor does not take into account the feedback of the learners' actual learning behavior, resulting in extremely limited accuracy in the course recommendation results. Therefore, it is necessary to combine the feedback of the learners' actual learning behavior, further select courses that have a higher learning behavior similarity with the learners' behavioral feedback in course learning, correct the connection factor, and then obtain the final connection parameters between different courses.
[0071] Specifically, taking the nth course and the n'th course as an example, the calculation method of the final connection parameter between the nth course and the n'th course is:
[0072]
[0073] In the formula, FR n,n′Indicates the final connection parameter between the nth course and the n′th course; CT n,n′ Represents the connection factor between the nth course and the n′th course; CY n,n′ represents the similarity of learning behaviors between the nth course and the n′th course; represents the mean of the similarity of learning behaviors between the nth course and all other courses except the nth course; || represents the absolute value function; exp() represents an exponential function with a natural constant as the base. This embodiment adopts the exp(-x) model to present the inverse proportional relationship. x is the input of the model. The implementer can set the inverse proportional function according to the actual situation.
[0074] It should be noted that based on the connection factors between different courses, the weights are constructed in combination with the learning behavior similarities between different courses to correct the connection factors, and then the final connection parameters between different courses are obtained; the closer the ratio of the learning behavior similarity between two courses to the average of the learning behavior similarity between the nth course and all other courses is to 1, the more similar the behavioral feedback of the two courses in the learning process is, the larger the final connection parameter should be, and therefore the larger the weight is; conversely, the greater the difference between the ratio and 1, the greater the difference in behavioral feedback is, the smaller the final connection parameter should be, and the smaller the weight is; and the similarity in behavioral feedback, that is, the more similar the behavioral feedback between two courses is to the overall behavioral feedback between the nth course and other courses, the more suitable the corresponding two courses are as the learned courses and the courses to be recommended, and the learners' learning efficiency will be correspondingly higher.
[0075] Furthermore, the above calculation method is only a method using the final connection parameters between the nth course and the n′th course as an example in one embodiment of the present invention. Similarly, the method is used to calculate the final connection parameters between all courses to obtain the final connection parameters between different courses.
[0076] At this point, by analyzing the learning behavior feedback parameters of different courses, quantifying the similarity of learning behaviors, and further adjusting the connection factors between different courses, the final connection parameters are obtained. It is ensured that the final connection parameters not only reflect the connection between courses, but also include the quantification of the behavioral feedback similarities between different courses, ensuring that the recommendation based on the final connection parameters can take into account the reference of learners' learning efficiency of different courses to the greatest extent.
[0077] Step S005: According to the final connection parameters between different courses, obtain the relevant recommendation set of each course; and recommend courses to each learner based on the relevant recommendation set of each course.
[0078] It should be noted that after obtaining the final connection parameters, for any course, clustering is performed based on the final connection parameters between other courses and the course, and different clustering results can be obtained for each course, and then several related recommendation sets for each course are obtained, and then the final recommended courses are obtained based on the final connection parameters between other courses in the related recommendation sets and the course.
[0079] Specifically, taking the nth course as an example, DBSCAN clustering is performed on all courses except the nth course. The distance measurement uses the absolute value of the difference between the final connection parameters between different courses and the nth course, and several clusters are obtained; the set of all courses in any cluster is taken as a related recommendation set for the nth course, and several related recommendation sets for the nth course are obtained.
[0080] Furthermore, the mean of the final connection parameters between the nth course and all the courses in any related recommendation set is obtained, recorded as the final connection of the related recommendation set of the nth course, and the final connection of each related recommendation set of the nth course is obtained; a judgment threshold is preset, and the judgment threshold in this embodiment is described as 0.95, and the product of the maximum value of the final connection and the judgment threshold is used as the connection threshold; the course whose final connection parameter with the nth course is greater than or equal to the connection threshold is used as the recommended course for the learner whose course sequence includes the nth course.
[0081] Furthermore, the above calculation method is only a method for recommending courses using the nth course as an example in one embodiment of the present invention. Similarly, this method is used to obtain a relevant recommendation set for each course, so as to obtain a learning course sequence including recommended courses for learners of each course, thereby realizing course recommendations for each learner.
[0082] At this point, cluster analysis is performed through the final connection parameters between courses to obtain the relevant recommendation sets of each course, and then course recommendations are made to learners based on the relevant recommendation sets and the final connection parameters, thereby improving the accuracy of course recommendations during learners' online education course learning process, avoiding low recommendation effects due to different learning behavior feedback, and ultimately achieving improved learning efficiency for online education learners.
[0083] Another embodiment of the present invention provides an artificial intelligence-based online education system, which includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, steps S001 to S005 of the above method are implemented.
[0084] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.
Claims
1. An online education method based on artificial intelligence, characterized in that: The method comprises the following steps: Collect the learning behavior data of several learners in online education for the courses corresponding to each learning process, and obtain the learning course sequence of each learner; According to the distribution of different courses in each learner's learning course sequence, the connection factors between different courses are obtained; According to the learning behavior data of different learners in the process of learning each course, the learning behavior feedback parameters of each course are obtained; According to the differences in learning behavior feedback parameters of different courses, the learning behavior similarities between different courses are obtained; combined with the connection factors between different courses, the final connection parameters between different courses are obtained; According to the final connection parameters between different courses, a relevant recommendation set for each course is obtained; and courses are recommended to each learner based on the relevant recommendation set for each course.
2. The online education method based on artificial intelligence according to claim 1, characterized in that: The specific method of collecting the learning behavior data of several learners for each course corresponding to the learning process in online education includes: For any course corresponding to any learning process of any learner, all viewing behaviors of the course during the learning process are collected, including the rated duration of the course, the total skipping time of the learner in the process of learning the course, the total number of pauses of the learner in the process of learning the course, and the total repeated learning time of the learner in the process of learning the course, which are comprehensively used as the learner's learning behavior data for the course.
3. The online education method based on artificial intelligence according to claim 1, characterized in that: The specific method of obtaining the learning course sequence of each learner includes: For any learner, the courses corresponding to each learning process of the learner are taken as a number of learning course data of the learner, and all the learning course data are arranged in the order of the learning process to obtain the learning course sequence of the learner.
4. The online education method based on artificial intelligence according to claim 1, characterized in that: The specific method of obtaining the connection factors between different courses is as follows: In the formula, CT n,n′ represents the connection factor between the nth course and the n′th course; M n,n′ It represents the total number of times the nth course and the n′th course appear in the learning course sequence of the same learner among all M learning course sequences; M represents the total number of learning course sequences corresponding to all learners.
5. The online education method based on artificial intelligence according to claim 2, characterized in that: The specific method of obtaining the learning behavior feedback parameters of each course includes: In the formula, CS n represents the learning behavior feedback parameter of the nth course; N represents the total number of different courses in the learning course sequence of all learners; represents the mean of the total number of pauses of all learners in the process of learning the nth course; represents the mean of the total number of pauses of all learners in the process of learning the n′th course; RT m′,n It represents the total duration of repeated learning of the m′th learner who has repeated learning behavior in the process of learning the nth course; M′ n It represents the total number of learners who have repeated learning behaviors in the process of learning the nth course; T n represents the rated duration of the nth course; JT m″,n M′ represents the total skipping time of the m″th learner who has skipping learning behavior in the process of learning the nth course; n ′ represents the total number of learners who skip learning behavior during the process of learning the nth course.
6. The online education method based on artificial intelligence according to claim 1, characterized in that: The learning behavior similarity between different courses is obtained by: In the formula, CY n,n′ represents the similarity parameter of learning behavior between the nth course and the n′th course; CS n represents the learning behavior feedback parameter of the nth course; CS n′ represents the learning behavior feedback parameter of the n′th course; CS n″ represents the learning behavior feedback parameter of the n″th course; N represents the total number of different courses in the learning course sequence of all learners; Obtain the learning behavior similarity parameters between the nth course and each other course except the nth course, perform linear normalization on all the obtained learning behavior similarity parameters, and obtain the result as the learning behavior similarity between the nth course and each other course except the nth course.
7. The online education method based on artificial intelligence according to claim 1, characterized in that: The final connection parameters between different courses are obtained by: In the formula, FR n,n′ Indicates the final connection parameter between the nth course and the n′th course; CT n,n′ Represents the connection factor between the nth course and the n′th course; CY n,n′ represents the similarity of learning behaviors between the nth course and the n′th course; represents the mean of the similarity of learning behaviors between the nth course and all other courses except the nth course; || represents the absolute value function; exp() represents the exponential function with a natural constant as the base.
8. The online education method based on artificial intelligence according to claim 1, characterized in that: The specific method for obtaining the relevant recommendation set for each course is as follows: DBSCAN clustering is performed on all courses except the nth course. The distance measurement uses the absolute value of the difference between the final connection parameters between different courses and the nth course to obtain several clusters. The set of all courses in any cluster is taken as a related recommendation set for the nth course, and several related recommendation sets for the nth course are obtained.
9. The online education method based on artificial intelligence according to claim 1, characterized in that: The specific method of recommending courses to each learner based on the relevant recommendation set of each course includes: Obtain the mean of the final connection parameters between the nth course and all courses in any related recommendation set, record it as the final connection of the related recommendation set of the nth course, and obtain the final connection of each related recommendation set of the nth course; A judgment threshold is preset, and the product of the final maximum value of the connection and the judgment threshold is used as the connection threshold; A course whose final connection parameter with the n-th course is greater than or equal to the connection threshold is used as a recommended course for the learner whose learning course sequence includes the n-th course.
10. An online education system based on artificial intelligence, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of an online education method based on artificial intelligence as described in any one of claims 1-9 are implemented.
