Online teaching course recommendation method and system based on Internet

Abstract

The invention discloses an internet-based online teaching course recommendation method and system, and relates to the field of data analysis, and the main scheme is that a learning ability index X is calculated according to a learning progress index XA, a learning initiative index XB, a family-habit cooperation index XC and a learning efficiency index XD, and the learning ability of a user can be comprehensively evaluated; the method solves the problems that course recommendation is carried out only according to initial grade, age and other information during course recommendation, but the difference between learning ability indexes such as user learning progress conditions and learning initiative is neglected, so that the course recommendation difficulty is possibly not matched with the actual ability of the user, and the course recommendation efficiency is improved. According to the method, the problem of gradual disjunction between the recommended course and the actual demand of the user due to the learning ability index X and the learning ability grade threshold set is solved, the learning ability grade of the user is judged according to the learning ability index X and the learning ability grade threshold set, and the course recommendation method is selected, so that the course recommendation accuracy can be improved, and educational resources are utilized more effectively.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to an online teaching course recommendation method and system based on the Internet. Background Art

[0002] In the field of secondary education, recommending online teaching courses to users through the Internet can meet the differentiated learning needs of users at different levels, allowing users to arrange their studies freely according to their own time and learning pace, thereby improving their learning autonomy and enthusiasm. At the same time, it can integrate high-quality teaching resources from different regions, schools, and teachers, breaking geographical and school restrictions, and allowing users to access broader and higher-quality educational content.

[0003] The traditional method of course recommendation is mainly to collect information such as the user's grade, age, gender, and school location, build a user profile, and recommend courses that are suitable for the user's educational stage based on the profile.

[0004] The existing technology still has the following shortcomings: when making course recommendations, it only recommends courses based on initial grade, age and other information, while ignoring the differences between learning ability indicators such as user learning progress and learning initiative, which may make the difficulty of the recommended courses mismatch with the user's actual ability, and thus cause the recommended courses to gradually become out of touch with the user's actual needs. Summary of the invention

[0005] 1. Technical issues to be resolved In view of the deficiencies in the prior art, the present invention provides an Internet-based online teaching course recommendation method and system, which calculates a learning ability index X according to a learning progress index XA, a learning initiative index XB, a family-habit synergy index XC and a learning effectiveness index XD, judges a user's learning ability level according to the learning ability index X and a learning ability level threshold set, and selects a course recommendation method according to the user's learning ability level, thereby solving the problem that when making course recommendations, courses are recommended only based on initial grade, age and other information, while ignoring the differences between learning ability indicators such as user learning progress and learning initiative, which may make the difficulty of the recommended courses mismatch with the actual ability of the user, thereby causing the recommended courses to gradually become out of touch with the actual needs of the user.

[0006] (II) Technical solution To achieve the above objectives, the present invention is implemented through the following technical solutions: an online teaching course recommendation method based on the Internet, comprising the following steps: Collect users' learning performance data, study time data, exercise practice data, interaction data, family data, study habit data, video viewing data and resource access data; Calculate the achievement stability WA based on the learning performance data, and calculate the learning progress index XA based on the achievement stability WA, learning performance data, learning time data and exercise data; calculate the learning initiative index XB based on the learning time data and interaction data; calculate the family resource support index WD based on the family data, calculate the learning habit index WE based on the learning habit data, and calculate the family-habit synergy index XC based on the family resource support index WD and the learning habit index WE; calculate the learning effectiveness index XD based on the learning time data, video viewing data and resource access data; The learning ability index X is calculated according to the learning progress index XA, the learning initiative index XB, the family-habit synergy index XC and the learning effectiveness index XD; a learning ability level threshold set is preset; the user's learning ability level is judged according to the learning ability index X and the learning ability level threshold set, and a course recommendation method is selected according to the user's learning ability level.

[0007] In the preferred embodiment of the above-mentioned online teaching course recommendation method based on the Internet: the method for calculating the score stability WA is: Academic performance data includes the subject grade AA this semester ac ; The performance stability WA is calculated based on the learning performance data according to the following formula:

[0008] Among them, AA ac It is the score of the user in the cth test of the ath subject. a is the serial number corresponding to different subjects, and its value is [1, b]; b is the total number of subjects, and its value is a positive integer; c is the serial number corresponding to different tests, and its value is [1, d]; d is the total number of tests, and its value is a positive integer.

[0009] In the preferred embodiment of the above-mentioned method for recommending online teaching courses based on the Internet, the method for calculating the learning progress index XA is: Academic performance data also includes the comprehensive grades AB for this semester and AC for the previous semester; Study time data include study time CA for this semester; The exercise practice data include the number of correct exercises BA, the total number of exercises practice BB, and the number of repeated incorrect questions BC; The learning progress index XA is calculated based on the score stability WA, the comprehensive score AB of this semester, the comprehensive score AC of the previous semester, the study time data and the exercise practice data. The formula is: .

[0010] In the preferred embodiment of the above-mentioned method for recommending online teaching courses based on the Internet, the method for calculating the learning initiative index XB is: The study time data also includes the continuous study time DA and the total study time DB of this semester; The interaction data include the number of active speeches, EA, and the number of correct answers to questions, EB; The learning initiative index XB is calculated based on the continuous learning time DA, the total learning time DB and the interaction data of this semester. The formula is:

[0011] Among them, ɑ1 is the weight coefficient of the learning persistence index, which ranges from 0.3 to 0.7; ɑ2 is the weight coefficient of the interactive participation index, which ranges from 0.3 to 0.7; and ɑ1+ɑ2=1.

[0012] In the preferred solution of the above-mentioned Internet-based online teaching course recommendation method: the method for calculating the family resource support index WD is: Family data include the quietness index of learning space FA, the number of books owned FB, the average number of books owned FC, the number of learning devices FD and the parent education index FE; The family resource support index WD is calculated based on family data according to the following formula: .

[0013] In the preferred embodiment of the above-mentioned method for recommending online teaching courses based on the Internet: the method for calculating the family-habit synergy index XC is: Study habit data include preview frequency score GA, review frequency score GB and wrong question sorting frequency score GC; The formula for calculating the learning habit index WE based on the learning habit data is:

[0014] The family-habit synergy index XC is calculated based on the family resource support index WD and the learning habit index WE, and the formula is as follows: .

[0015] In the preferred embodiment of the above-mentioned method for recommending online teaching courses based on the Internet, the method for calculating the learning effectiveness index XD is: The learning time data also includes the time spent using non-learning applications, HB; Video viewing data includes the number of videos played at a speed higher than 1x (HC) and the number of video playbacks (HE). m and the average number of video playbacks HF m ; Resource access data include the number of visits to academic websites HG, the number of disciplines included in the collection resources HI, the number of academic journal subscriptions HJ, and the proportion of academic journal reading HK; The learning effectiveness index XD is calculated based on the non-learning application usage time HB, the learning time CA this semester, the video viewing data and the resource access data. The formula is:

[0016] Among them, HE m is the number of times the mth video is played back, m is the serial number corresponding to different videos, and its value is [1, n]; n is the total number of videos watched by the user, and its value is a positive integer; HF m is the average number of playback times of the mth video.

[0017] In the preferred embodiment of the above-mentioned method for recommending online teaching courses based on the Internet, the method for calculating the learning ability index X is: The learning ability index X is calculated based on the learning progress index XA, the learning initiative index XB, the family-habit synergy index XC and the learning effectiveness index XD, and the formula is:

[0018] Among them, β1 is the weight coefficient of the learning progress index XA, and its value is 0.2~0.4; β2 is the weight coefficient of the learning initiative index XB, and its value is 0.2~0.3; β3 is the weight coefficient of the family-habit synergy index XC, and its value is 0.2~0.3; β4 is the weight coefficient of the learning effectiveness index XD, and its value is 0.3~0.4; and β1+β2+β3+β4=1.

[0019] In the preferred embodiment of the above-mentioned method for recommending online teaching courses based on the Internet, the method for judging the user's learning ability level is: The learning ability level threshold set includes a high learning ability threshold YG and a low learning ability threshold YD; The user's learning ability level is determined based on the learning ability indicator X and the learning ability level threshold set. The formula is:

[0020] When the learning ability is high, choose the course recommendation method 1; When learning ability is medium, choose the second course recommendation method; When learning ability is low, choose course recommendation method three.

[0021] The present invention also discloses an Internet-based online teaching course recommendation system, comprising: Data collection module, used to collect users' learning performance data, learning time data, exercise data, interaction data, family data, learning habit data, video viewing data and resource access data; The indicator calculation module can calculate the performance stability WA based on the learning performance data, calculate the learning progress index XA based on the performance stability WA, the learning performance data, the learning time data and the exercise data; calculate the learning initiative index XB based on the learning time data and the interaction data; calculate the family resource support index WD based on the family data, calculate the learning habit index WE based on the learning habit data, and calculate the family-habit synergy index XC based on the family resource support index WD and the learning habit index WE; calculate the learning effectiveness index XD based on the learning time data, the video viewing data and the resource access data; The judgment module can calculate the learning ability index X according to the learning progress index XA, the learning initiative index XB, the family-habit synergy index XC and the learning effectiveness index XD; preset a learning ability level threshold set; judge the user's learning ability level according to the learning ability index X and the learning ability level threshold set, and select a course recommendation method according to the user's learning ability level.

[0022] (III) Beneficial effects The present invention provides an online teaching course recommendation method based on the Internet, which has the following beneficial effects: (1) By collecting users’ learning performance data, study time data, exercise practice data, interaction data, family data, study habit data, video viewing data, and resource access data, we can evaluate users’ learning ability from multiple aspects, thereby improving the accuracy of course recommendations.

[0023] (2) The learning progress index XA is calculated based on the learning performance data, learning time data and exercise data. It can accurately evaluate the user's learning outcomes and discover the bottlenecks and advantages of the user in the learning process. The learning initiative index XB is calculated based on the learning time data and interaction data. It can understand the user's investment in learning and measure the user's interest in the learning content. The family-habit synergy index XC based on the family data and learning habit data can reflect the impact of the family environment on the user's learning habits, which is conducive to discovering family factors that may affect the user's learning in advance. The learning effectiveness index XD is calculated based on the learning time data, video viewing data and resource access data. It can help users objectively understand their own learning situation and provide a basis for users to adjust their learning strategies.

[0024] (3) The learning ability index X is calculated based on the learning progress index XA, the learning initiative index XB, the family-habit synergy index XC and the learning effectiveness index XD. This can comprehensively evaluate the user's learning ability and solve the problem that when making course recommendations, courses are recommended only based on initial information such as grade and age, while ignoring the differences between learning ability indicators such as user learning progress and learning initiative. This may cause the difficulty of the recommended courses to be mismatched with the user's actual ability, and thus lead to the recommended courses gradually becoming out of touch with the user's actual needs. The user's learning ability level is judged based on the learning ability index X and the learning ability level threshold set, and the course recommendation method is selected based on the user's learning ability level. This can improve the accuracy of course recommendations and make more effective use of educational resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 The figure is a schematic diagram of the working steps of the Internet-based online teaching course recommendation method of the present invention. DETAILED DESCRIPTION

[0026] 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.

[0027] See also Figure 1 The present invention provides an online teaching course recommendation method on the Internet, comprising the following steps: Step 1: Collect the user's learning performance data, study time data, exercise data, interaction data, family data, study habit data, video viewing data and resource access data.

[0028] Combined with the content of step 1: By collecting users' academic performance data, study time data, exercise practice data, interaction data, family data, study habit data, video viewing data and resource access data, we can evaluate users' learning ability from multiple aspects, thereby improving the accuracy of course recommendations.

[0029] Step 2: Calculate the performance stability WA based on the learning performance data, and calculate the learning progress index XA based on the performance stability WA, learning performance data, learning time data and exercise practice data; calculate the learning initiative index XB based on the learning time data and interaction data; calculate the family resource support index WD based on the family data, calculate the learning habit index WE based on the learning habit data, and calculate the family-habit synergy index XC based on the family resource support index WD and the learning habit index WE; calculate the learning effectiveness index XD based on the learning time data, video viewing data and resource access data.

[0030] Step 2 includes the following steps: Step 201: The method for calculating the performance stability WA is: Academic performance data includes the subject grade AA this semester ac .

[0031] It should be noted that the subject grades for this semester are AA ac It reflects the results achieved by users in studying different subjects during this semester. It is obtained by querying the personal learning center or score query section of the online learning platform. The online learning platform includes NetEase Cloud Classroom, primary and secondary school synchronous classrooms, etc.

[0032] The performance stability WA is calculated based on the learning performance data according to the following formula:

[0033] Among them, AA ac It is the score of the user in the cth test of the ath subject. a is the serial number corresponding to different subjects, and its value is [1, b]; b is the total number of subjects, and its value is a positive integer; c is the serial number corresponding to different tests, and its value is [1, d]; d is the total number of tests, and its value is a positive integer.

[0034] It should be noted that this formula measures the fluctuation of scores in different subjects by calculating the standard deviation of the test scores of each subject, and then adds up the standard deviations of all subjects and calculates their average to obtain the score stability WA.

[0035] Step 202: The method for calculating the learning progress indicator XA is: The academic performance data also includes the comprehensive grades AB for this semester and AC for the previous semester.

[0036] It should be noted that the comprehensive score AB for this semester is a comprehensive evaluation of the user's learning achievements in this semester. The method of obtaining it is: by querying the personal learning center or the score query section of the online learning platform to obtain the scores of various tests that the user has participated in this semester, these scores are summarized and divided by the number of tests, as the comprehensive score AB for this semester. The comprehensive score AC for the previous semester is a comprehensive consideration and overall evaluation of the user's learning situation in the previous semester. The method of obtaining it is: by querying the personal learning center or the score query section of the online learning platform to obtain the scores of various tests that the user has participated in the previous semester, these scores are summarized and divided by the number of tests, as the comprehensive score AC for the previous semester.

[0037] The study time data includes the study time CA for this semester.

[0038] It should be noted that the study time CA this semester refers to the total time the user spends on studying various subjects every day this semester. The method of obtaining it is: log in to the online learning platform to view the course introduction or study plan of the course the user participates in, which will contain specific arrangements for the study time of the course, including the course start time, end time, course duration, weekly learning tasks and corresponding time requirements, etc. The course duration of all courses the user participates in is accumulated as the study time CA this semester.

[0039] The exercise practice data includes the number of correct exercises BA, the total number of exercise practice BB, and the number of repeated incorrect questions BC.

[0040] It should be noted that the number of correct exercises BA refers to the number of questions answered correctly among all the exercises completed by the user this semester, which reflects the user's mastery of knowledge and application ability. It is obtained in the following way: when the user completes the exercise and submits it on the online learning platform, the system will automatically judge the right or wrong according to the preset answers, and count the number of correct questions. The number of correct questions in each exercise of the user is summarized as the number of correct exercises BA. The total number of exercise exercises BB refers to the total number of questions that the user has practiced in this semester, including the number of correct questions and the number of incorrect questions. It is obtained in the following way: the online learning platform will count the user's learning data. The number of exercises completed on the platform can be viewed through the personal learning data page of the platform. The number of exercises completed by the user on different platforms is summarized as the total number of exercises BB. The number of repeated incorrect questions BC refers to the number of questions that users repeatedly make mistakes in multiple exercise practices, reflecting the user's weak links in certain knowledge points and the stubborn mistakes in the learning process. It is obtained as follows: online learning platforms and educational software have built-in analysis tools, which can automatically identify the questions that users repeatedly make mistakes in multiple exercises, and count the number of these questions, which is used as the number of repeated incorrect questions BC.

[0041] The learning progress index XA is calculated based on the score stability WA, the comprehensive score AB of this semester, the comprehensive score AC of the previous semester, the study time data and the exercise practice data. The formula is:

[0042] It should be noted that the operating principle of this formula is: is the user's progress this semester. By calculating the ratio of the user's progress to the learning time CA this semester, we can get the learning efficiency. The larger the ratio is, the greater the user's progress per unit time is, and the higher the learning progress index XA is. The correct rate of the user's questions reflects the user's initial grasp of knowledge. The user's wrong question repetition rate is used to measure the depth of the user's understanding of the knowledge and whether the user has truly mastered it. By combining the two, we can get the firmness of knowledge mastery. The larger the value, the more firmly the user has mastered the knowledge, and the higher the learning progress index XA. This formula uses the geometric mean method. By taking the cube root of the product of the three indicators of score stability, learning efficiency and firmness of knowledge mastery, it can balance the contribution of the three indicators and finally obtain the learning progress index XA.

[0043] Step 203: The method for calculating the learning initiative index XB is: The study time data also includes the continuous study time DA of this semester and the total study time DB of this semester.

[0044] It should be noted that the continuous learning time DA of this semester refers to the number of days that the user studies uninterruptedly within the time range of a semester. It is an important indicator to measure the user's learning habits and self-discipline. The method of obtaining it is as follows: the online learning platform will record the user's learning time and learning date, check the user's learning date, mark the user's continuous learning time period and calculate the number of days it contains. For example, if the user studies every day from January 1 to January 7, but does not study on January 8, then January 1 to January 7 is the user's continuous learning time, which contains 7 days. Count all the user's continuous learning time periods, and calculate the number of days contained in these time periods respectively, and take the maximum number of days as the continuous learning time DA of this semester. The total duration DB of this semester refers to the length of the entire time period from the beginning of this semester to the end of this semester. The method of obtaining it is as follows: query the start date and end date of this semester through the school calendar or the academic affairs management system, calculate the difference between the two, and obtain the total duration DB of this semester.

[0045] The interaction data includes the number of active speeches EA and the number of correct answers EB.

[0046] It should be noted that the number of active speeches EA refers to the total number of times users actively start communication, express their own views, ideas, questions or share knowledge in learning scenarios such as classrooms, study discussion groups, and Q&A sessions. It reflects the user's enthusiasm, participation and willingness to actively learn. The method of obtaining it is as follows: for live courses, the platform can record the number of times users actively send messages in the chat box. For courses with voice interaction functions, the platform can count the number of times users actively start voice speaking. At the same time, by viewing the user's personal homepage, the number of times users post learning-related content on social platforms can be obtained. The three times are added together as the number of active speeches EA. The number of correct answers to questions EB refers to the total number of times users' answers to questions raised by teachers are consistent with the correct answers. It reflects the user's mastery of knowledge and ability to apply it. The method of obtaining it is as follows: the online teaching platform has the function of automatically recording and counting the user's answers. After the teacher posts exercises, test questions or classroom questions, the platform can record the user's answers in real time and automatically determine whether the answers are correct, and then count the number of correct answers to questions EB.

[0047] The learning initiative index XB is calculated based on the continuous learning time DA, the total learning time DB and the interaction data of this semester. The formula is:

[0048] Among them, ɑ1 is the weight coefficient of the learning persistence index, which ranges from 0.3 to 0.7 and is determined according to the importance of the learning persistence index to the learning initiative index XB; ɑ2 is the weight coefficient of the interactive participation index, which ranges from 0.3 to 0.7 and is determined according to the importance of the interactive participation index to the learning initiative index XB; and ɑ1+ɑ2=1.

[0049] It should be noted that the operating principle of this formula is: It is used to measure the user's persistence in the learning process. By calculating the ratio of the continuous learning time DA in this semester to the total learning time DB in this semester and multiplying it by the total score 100, the learning persistence index is obtained. By calculating the average of the number of active speeches EA and the number of correct answers EB, the enthusiasm of users in participating in the interaction in the course is measured by comprehensively considering the two aspects of students' active participation in communication and correct answers to questions. Finally, by calculating the influence of persistence and enthusiasm, the learning initiative index XB is obtained.

[0050] Step 204: The method for calculating the family resource support index WD is: Family data include the quietness index of the learning space FA, the number of books owned FB, the average number of books owned FC, the number of learning devices FD and the parent education index FE.

[0051] It should be noted that the learning space quietness index FA is a quantitative assessment of the quietness of the user's learning environment. It is used to measure whether the learning space is conducive to the user's concentration on learning. The method of obtaining it is: obtaining the noise decibel value of the user's learning space through a questionnaire. When the noise decibel value is less than 40 decibels, the learning space quietness index FA is 5 points. When the noise decibel value is between 40 and 55 decibels, the learning space quietness index FA is 3 points. When the noise decibel value is greater than 55 decibels, the learning space quietness index FA is 1 point. The number of books owned FB refers to the total number of books available for reading and learning owned by the user. These books can cover various types of subject knowledge, literary works, reference materials, etc. It reflects the richness of learning resources. The method of obtaining it is: through user authorization or signing a cooperation agreement with the e-commerce platform, query the detailed list of books purchased by the user, including the title, quantity and price of the purchased books, and accumulate the number of books in the purchase record as the number of books owned FB. The average number of books owned FC refers to the average number of books owned by other users in the same region and age group as the user. It is obtained by using Internet big data to collect book purchase records of other users in the same region and age group as the user on the e-commerce platform, calculate the number of books purchased by each user and add them up, and then divide it by the number of users to get the average number of books owned FC. The number of learning devices FD refers to the number of devices that can be used to assist learning in the learning environment. These devices include but are not limited to computers, tablets, electronic dictionaries, learning machines, etc. It reflects the hardware support level of learning resources. It is obtained by making a list of learning devices, and users check whether they have the device at home according to the list, and count the number of devices owned as the number of learning devices FD. The Parent Education Index FE is a quantitative assessment of the parents' educational level. It is used to measure the possible impact of parents on the user's learning in terms of knowledge and cultural level. The method of obtaining it is: obtain the highest educational level of the user's parents through a questionnaire. When the highest educational level is primary school, the score is 1 point; when the highest educational level is junior high school, the score is 2 points; when the highest educational level is high school, the score is 3 points; when the highest educational level is university, the score is 4 points; when the highest educational level is a master's degree or above, the score is 5 points. Add up the scores of the user's parents to obtain the Parent Education Index FE.

[0052] The family resource support index WD is calculated based on family data according to the following formula:

[0053] It should be noted that this formula obtains the family resource support index WD by calculating the geometric mean of the learning space quietness index FA, the book ownership ratio, the learning equipment index and the parent education index.

[0054] Step 205: The method for calculating the family-habit synergy index XC is: Study habit data include preview frequency score GA, review frequency score GB and wrong question sorting frequency score GC.

[0055] It should be noted that the preview frequency score GA is a quantitative assessment of the frequency of user preview behavior. The higher the preview frequency score, the more frequently the user actively previews during the learning process, and the stronger the user's initiative and foresight in learning. The acquisition method is as follows: the online course platform will automatically record the number of times each user watches preview videos and reads preview documents. The number of times each user watches preview videos and reads preview documents within a month is obtained by querying the platform records. The number of times for all users is added up and divided by the number of users to obtain the average number of times. When the number of times the user watches preview videos and reads preview documents within a month is greater than the average number of times, the preview frequency score GA is 5 points. When the number of times the user watches preview videos and reads preview documents within a month is equal to the average number of times, the preview frequency score GA is 3 points. When the number of times the user watches preview videos and reads preview documents within a month is less than the average number of times, the preview frequency score GA is 1 point. The review frequency score GB is used to measure the frequency of users reviewing their lessons, reflecting the users' awareness of knowledge consolidation and self-discipline in learning. It is obtained as follows: the number of times different users review within three months is obtained through a questionnaire, the number is accumulated and divided by the number of users to obtain the average number of reviews, and the user's questionnaire is checked to obtain the number of times the user reviews within three months. When the number is greater than the average number of reviews, the review frequency score GB is 5 points; when the number is equal to the average number of reviews, the review frequency score GB is 3 points; when the number is less than the average number of reviews, the review frequency score GB is 1 point. The wrong question sorting frequency score GC is a quantitative reflection of the user's learning behavior of sorting out wrong questions, reflecting whether the user can reflect on his or her learning problems in a timely manner. The method of obtaining it is: obtain the number of times different users sort out wrong questions within three months through a questionnaire, add up the number and divide it by the number of users to get the average value of wrong question sorting, check the user's questionnaire to obtain the number of times the user sorts out wrong questions within three months. When the number is greater than the average value of wrong question sorting, the wrong question sorting frequency score GC is 5 points. When the number is equal to the average value of wrong question sorting, the wrong question sorting frequency score GC is 3 points. When the number is less than the average value of wrong question sorting, the wrong question sorting frequency score GC is 1 point.

[0056] The formula for calculating the learning habit index WE based on the learning habit data is:

[0057] It should be noted that the operating principle of this formula is: It is used to normalize the preview frequency score GA, review frequency score GB and wrong question sorting frequency score GC so that they are between [0,1], and then use the sin function to adjust them to obtain the learning habit index WE.

[0058] The family-habit synergy index XC is calculated based on the family resource support index WD and the learning habit index WE, and the formula is as follows:

[0059] It should be noted that the operating principle of this formula is: this formula comprehensively considers the influence of the family resource support index WD and the learning habit index WE to obtain the family-habit synergy index XC. The higher the family resource support index WD and the learning habit index WE, the more conducive it is to improving the user's learning ability, and the higher the family-habit synergy index XC.

[0060] Step 206: The method for calculating the learning effectiveness index XD is: The learning time data also includes the non-learning application usage time HB.

[0061] It should be noted that the non-learning application usage time HB refers to the time users spend on non-learning applications such as social software and entertainment software, which reflects the degree of students' investment in leisure and entertainment. It is obtained by using software with application usage time statistics function such as Tencent Mobile Manager and Maimiao Guardian to obtain the usage time of each application of users during the semester and summarize it as the non-learning application usage time HB. Among them, social software includes WeChat, QQ, etc., and entertainment software includes Douyin, iQiyi, etc.

[0062] Video viewing data includes the number of videos played at a speed higher than 1x (HC) and the number of video playbacks (HE). m and the average number of video playbacks HF m .

[0063] It should be noted that the number of videos played at speeds higher than 1x (HC) refers to the total number of videos played at speeds higher than 1x, such as 1.25x, 1.5x, 2x, etc., during the semester. It reflects the user's familiarity with the video content, their need to quickly obtain information, and their ability to adapt to the rhythm of the video. The method of obtaining the data is as follows: the background of the online course platform will record the playback behavior data of each user, including playback speed, number of playbacks, etc. The playback behavior data of the user in the semester is obtained through the data analysis tools or interfaces provided by the platform, and the videos with playback speeds higher than 1x are screened out and counted as the number of videos played at speeds higher than 1x (HC). Video playback times HE mIt refers to the number of times a user repeatedly plays a single video, reflecting the user's attention, interest, and understanding of the video content. The method of obtaining it is: obtain the number of times the user plays each video through the data analysis tools or interfaces provided by the platform, and subtract 1 from the number to obtain the number of video playbacks. m . Video playback times average HF m It refers to the average number of times each video is played back, reflecting the quality and difficulty of the video content. It is obtained by: for a specific video, traversing the playback behavior data recorded in the platform server, and obtaining the total number of users who have watched the video by identifying the user ID, and the total number of times each user has watched the video, Q r , according to the formula (Q r -1) / tCalculate the average number of video playback times HF m , where Q r The cumulative number of times the rth user has watched the video.

[0064] Resource access data include the number of visits to academic websites HG, the number of disciplines included in the collection resources HI, the number of academic journal subscriptions HJ and the proportion of academic journal reading HK.

[0065] It should be noted that the number of academic website visits HG refers to the total number of times users visit academic websites such as CNKI and ResearchGate in this semester, reflecting the frequency of users' active acquisition of academic resources. The acquisition method is as follows: the server of each academic website will automatically record the relevant information of each access request, including access time, IP address, visited pages, etc. By analyzing the server log files through log analysis tools such as AWStats and Webalizer, the number of times users visit each academic website in this semester can be counted. The number of times users visit each academic website is summarized, and the summarized values ​​are normalized to obtain the number of academic website visits HG. The number of disciplines included in the collection resources HI refers to the number of different disciplines covered in the user's collection resources, reflecting the user's academic interest breadth and knowledge integration tendency. The acquisition method is as follows: the online education platform will provide statistical information about the total number of collection resources, the number of disciplines included, the number of collection resources for each discipline, etc. in the user's personal center or personal profile page. By querying the statistical information, the number of disciplines included in the collection resources is obtained, and the normalized data is processed to obtain the number of disciplines HI. The number of academic journal subscriptions HJ refers to the total number of academic journals subscribed by users. It is obtained in the following way: the academic journal subscription platform will record the subscription information of each user in the database, including the name of each journal subscribed, the number of journals, the subscription time, etc., and obtain the number of journals subscribed by the user each time by querying the database and accumulating them. The accumulated results are normalized and processed as the number of academic journal subscriptions HJ. The academic journal reading ratio HK refers to the ratio of the number of academic journal articles actually read by users in this semester to the total number of articles published in the journals subscribed by users in the same period of time. It is used to measure the user's utilization and reading efficiency of subscribed journals. It is obtained in the following way: the number of journals actually read by users in this semester RA is obtained through questionnaires, and the number of articles published in the journals subscribed by users in this semester RB is obtained by visiting the academic journal subscription platform. The ratio of RA to RB is calculated as the academic journal reading ratio HK.

[0066] The learning effectiveness index XD is calculated based on the non-learning application usage time HB, the learning time CA this semester, the video viewing data and the resource access data. The formula is:

[0067] Among them, HE m is the number of times the mth video is played back, m is the serial number corresponding to different videos, and its value is [1, n]; n is the total number of videos watched by the user, and its value is a positive integer; HF m is the average number of playback times of the mth video.

[0068] It should be noted that the operating principle of this formula is: It reflects the proportion of time users spend on non-learning applications. The lower the proportion, the higher the user's self-discipline and the greater the contribution to the learning effectiveness indicator XD. It is used to measure the speed at which users understand the video. The larger the value is, the faster the user understands the video, and the greater the contribution to the learning efficiency indicator XD. It is used to measure the user's attention to cutting-edge knowledge. The larger the value is, the higher the user's attention to cutting-edge knowledge is, and the greater the contribution to the learning effectiveness index XD is. This formula uses the natural logarithm function ln to comprehensively consider the influence of the above three aspects and obtain the learning effectiveness index XD.

[0069] Combined with the contents of step 201 to step 206: The learning progress index XA is calculated based on the learning performance data, learning time data and exercise data, which can accurately evaluate the user's learning outcomes and discover the bottlenecks and advantages of the user in the learning process. The learning initiative index XB is calculated based on the learning time data and interaction data, which can understand the user's investment in learning and measure the user's interest in the learning content. The family-habit synergy index XC based on family data and learning habit data can reflect the impact of the family environment on the user's learning habits, which is conducive to discovering family factors that may affect user learning in advance. The learning effectiveness index XD is calculated based on the learning time data, video viewing data and resource access data, which can help users objectively understand their own learning situation and provide a basis for users to adjust their learning strategies.

[0070] Step 3: Calculate the learning ability index X based on the learning progress index XA, the learning initiative index XB, the family-habit synergy index XC and the learning effectiveness index XD; preset a learning ability level threshold set; determine the user's learning ability level based on the learning ability index X and the learning ability level threshold set, and select a course recommendation method based on the user's learning ability level.

[0071] Step 301: The method for calculating the learning ability index X is: The learning ability index X is calculated based on the learning progress index XA, the learning initiative index XB, the family-habit synergy index XC and the learning effectiveness index XD, and the formula is:

[0072] Among them, β1 is the weight coefficient of the learning progress indicator XA, which takes a value of 0.2~0.4, and is determined according to the influence of the learning progress indicator XA on the learning ability indicator X; β2 is the weight coefficient of the learning initiative indicator XB, which takes a value of 0.2~0.3, and is determined according to the influence of the learning initiative indicator XB on the learning ability indicator X; β3 is the weight coefficient of the family-habit synergy indicator XC, which takes a value of 0.2~0.3, and is determined according to the influence of the family-habit synergy indicator XC on the learning ability indicator X; β4 is the weight coefficient of the learning effectiveness indicator XD, which takes a value of 0.3~0.4, and is determined according to the influence of the learning effectiveness indicator XD on the learning ability indicator X; and β1+β2+β3+β4=1.

[0073] It should be noted that the operating principle of this formula is: by summing the squares of the four indicator values, it is possible to avoid excessive influence of a certain indicator being too large or too small on the result; by multiplying each indicator by the corresponding weight coefficient and then summing them up, the contribution of different indicators can be adjusted; finally, the two parts are added, the square root is taken and divided by 4 to obtain the learning ability indicator X.

[0074] Step 302: The method for determining the user's learning ability level is: The learning ability level threshold set includes a high learning ability threshold YG and a low learning ability threshold YD.

[0075] It should be noted that the method for determining the higher learning ability threshold YG and the lower learning ability threshold YD is: collect historical data of different users, calculate the learning ability index according to the above method, screen out the higher 30% of the data, calculate their average value as the reference value of the higher learning ability threshold YG, and screen out the lower 40% of the data, calculate their average value as the reference value of the lower learning ability threshold YD.

[0076] The user's learning ability level is determined based on the learning ability indicator X and the learning ability level threshold set. The formula is:

[0077] When the learning ability is high, select the course recommendation method 1, specifically: for users with higher learning ability, some interdisciplinary learning courses can be recommended, so that they can integrate and apply the knowledge and methods of different disciplines. At the same time, some competition, independent learning and academic exploration courses can be recommended to encourage users to conduct independent exploration and learning according to their own interests and specialties.

[0078] When the learning ability is medium, choose the second course recommendation method, which is: this type of user already has certain basic knowledge, and some courses that expand and extend the basic knowledge can be recommended. For example, in Chinese learning, some classic literary works appreciation courses can be recommended to help them deepen their understanding of literary works and improve their literary literacy. At the same time, we should focus on training and improving the learning methods of this type of users. We can recommend courses on learning methods and thinking methods, such as the application of mind maps, memory method training, etc., to help them master more effective learning skills and improve learning efficiency.

[0079] When the learning ability is low, choose the course recommendation method three, which is: this type of user has not yet mastered the basic knowledge solidly and needs to consolidate the basics first. Some interesting basic consolidation courses can be recommended to stimulate the user's interest in learning. At the same time, some learning method guidance courses can be recommended to help users establish learning habits that suit them.

[0080] Combined with the contents of step 301 to step 302: The learning ability index X is calculated based on the learning progress index XA, the learning initiative index XB, the family-habit synergy index XC and the learning effectiveness index XD, which can comprehensively evaluate the user's learning ability and solve the problem that when making course recommendations, courses are recommended only based on the initial grade, age and other information, while the differences between the learning ability indicators such as the user's learning progress and learning initiative are ignored, which may make the difficulty of the recommended courses mismatch with the user's actual ability, and then cause the recommended courses to gradually become out of touch with the user's actual needs. The user's learning ability level is judged based on the learning ability index X and the learning ability level threshold set, and the course recommendation method is selected according to the user's learning ability level, which can improve the accuracy of course recommendations and make more effective use of educational resources.

[0081] On the other hand, the present invention also discloses an Internet-based online teaching course recommendation system, which is used to implement the above-mentioned Internet-based online teaching course recommendation method, comprising: Data collection module, used to collect users' learning performance data, learning time data, exercise data, interaction data, family data, learning habit data, video viewing data and resource access data; The indicator calculation module can calculate the performance stability WA based on the learning performance data, calculate the learning progress index XA based on the performance stability WA, the learning performance data, the learning time data and the exercise data; calculate the learning initiative index XB based on the learning time data and the interaction data; calculate the family resource support index WD based on the family data, calculate the learning habit index WE based on the learning habit data, and calculate the family-habit synergy index XC based on the family resource support index WD and the learning habit index WE; calculate the learning effectiveness index XD based on the learning time data, the video viewing data and the resource access data; The judgment module can calculate the learning ability index X according to the learning progress index XA, the learning initiative index XB, the family-habit synergy index XC and the learning effectiveness index XD; preset a learning ability level threshold set; judge the user's learning ability level according to the learning ability index X and the learning ability level threshold set, and select a course recommendation method according to the user's learning ability level.

[0082] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product. A person of ordinary skill in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented in electronic hardware or in combination with computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0083] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0084] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be covered by the protection scope of the present application.

Claims

1. An online teaching course recommendation method based on the Internet, characterized by: The following steps are involved: Collect users' learning performance data, study time data, exercise practice data, interaction data, family data, study habit data, video viewing data and resource access data; Calculate the achievement stability WA based on the learning achievement data, and calculate the learning progress index XA based on the achievement stability WA, the learning achievement data, the learning time data and the exercise practice data; The learning initiative index XB is calculated based on the learning time data and the interaction data; the family resource support index WD is calculated based on the family data, the learning habit index WE is calculated based on the learning habit data, and the family-habit synergy index XC is calculated based on the family resource support index WD and the learning habit index WE; the learning effectiveness index XD is calculated based on the learning time data, video viewing data, and resource access data; Calculate the learning ability index X according to the learning progress index XA, the learning initiative index XB, the family-habit synergy index XC and the learning efficacy index XD; A learning ability level threshold set is preset; the user's learning ability level is judged according to the learning ability indicator X and the learning ability level threshold set, and a course recommendation method is selected according to the user's learning ability level.

2. The method for recommending online teaching courses based on the Internet according to claim 1, characterized in that: The method for calculating the performance stability WA is: Academic performance data includes the subject grade AA this semester ac ; The performance stability WA is calculated based on the learning performance data according to the following formula: Among them, AA ac It is the score of the user in the cth test of the ath subject. a is the serial number corresponding to different subjects, and its value is [1, b]; b is the total number of subjects, and its value is a positive integer; c is the serial number corresponding to different tests, and its value is [1, d]; d is the total number of tests, and its value is a positive integer.

3. The method for recommending online teaching courses based on the Internet according to claim 2, characterized in that: The method for calculating the learning progress indicator XA is: Academic performance data also includes the comprehensive grades AB for this semester and AC for the previous semester; Study time data include study time CA for this semester; The exercise practice data include the number of correct exercises BA, the total number of exercises practice BB, and the number of repeated incorrect questions BC; The learning progress index XA is calculated based on the score stability WA, the comprehensive score AB of this semester, the comprehensive score AC of the previous semester, the study time data and the exercise practice data. The formula is: 。 4. The method for recommending online teaching courses based on the Internet according to claim 3, characterized in that: The method for calculating the learning initiative index XB is: The study time data also includes the continuous study time DA and the total study time DB of this semester; The interaction data include the number of active speeches, EA, and the number of correct answers to questions, EB; The learning initiative index XB is calculated based on the continuous learning time DA, the total learning time DB and the interaction data of this semester. The formula is: Among them, ɑ1 is the weight coefficient of the learning persistence index, which ranges from 0.3 to 0.7; ɑ2 is the weight coefficient of the interactive participation index, which ranges from 0.3 to 0.7; and ɑ1+ɑ2=1.

5. The method for recommending online teaching courses based on the Internet according to claim 4, characterized in that: The method for calculating the family resource support index WD is: Family data include the quietness index of learning space FA, the number of books owned FB, the average number of books owned FC, the number of learning devices FD and the parent education index FE; The family resource support index WD is calculated based on family data according to the following formula: 。 6. The method for recommending online teaching courses based on the Internet according to claim 5, characterized in that: The method for calculating the family-habit synergy index XC is: Study habit data include preview frequency score GA, review frequency score GB and wrong question sorting frequency score GC; The formula for calculating the learning habit index WE based on the learning habit data is: The family-habit synergy index XC is calculated based on the family resource support index WD and the learning habit index WE, and the formula is as follows: 。 7. The method for recommending online teaching courses based on the Internet according to claim 6, characterized in that: The method for calculating the learning effectiveness index XD is: The learning time data also includes the time spent using non-learning applications, HB; Video viewing data includes the number of videos played at a speed higher than 1x (HC) and the number of video playbacks (HE). m and the average number of video playbacks HF m ; Resource access data include the number of visits to academic websites HG, the number of disciplines included in the collection resources HI, the number of academic journal subscriptions HJ, and the proportion of academic journal reading HK; The learning effectiveness index XD is calculated based on the non-learning application usage time HB, the learning time CA this semester, the video viewing data and the resource access data. The formula is: Among them, HE m is the number of times the mth video is played back, m is the serial number corresponding to different videos, and its value is [1, n]; n is the total number of videos watched by the user, and its value is a positive integer; HF m is the average number of playback times of the mth video.

8. The method for recommending online teaching courses based on the Internet according to claim 7, characterized in that: The method for calculating the learning ability index X is: The learning ability index X is calculated based on the learning progress index XA, the learning initiative index XB, the family-habit synergy index XC and the learning effectiveness index XD, and the formula is: Among them, β1 is the weight coefficient of the learning progress index XA, and its value is 0.2~0.4; β2 is the weight coefficient of the learning initiative index XB, and its value is 0.2~0.3; β3 is the weight coefficient of the family-habit synergy index XC, and its value is 0.2~0.3; β4 is the weight coefficient of the learning effectiveness index XD, and its value is 0.3~0.4; and β1+β2+β3+β4=1.

9. The method for recommending online teaching courses based on the Internet according to claim 8, characterized in that: The method for judging the user's learning ability level is: The learning ability level threshold set includes a high learning ability threshold YG and a low learning ability threshold YD; The user's learning ability level is determined based on the learning ability indicator X and the learning ability level threshold set. The formula is: When the learning ability is high, select the course recommendation method one; when the learning ability is medium, select the course recommendation method two; when the learning ability is low, select the course recommendation method three.

10. An online teaching course recommendation system based on the Internet, characterized by: include: Data collection module, used to collect users' learning performance data, learning time data, exercise data, interaction data, family data, learning habit data, video viewing data and resource access data; The indicator calculation module can calculate the performance stability WA based on the learning performance data, calculate the learning progress index XA based on the performance stability WA, the learning performance data, the learning time data and the exercise data; calculate the learning initiative index XB based on the learning time data and the interaction data; calculate the family resource support index WD based on the family data, calculate the learning habit index WE based on the learning habit data, and calculate the family-habit synergy index XC based on the family resource support index WD and the learning habit index WE; calculate the learning effectiveness index XD based on the learning time data, the video viewing data and the resource access data; The judgment module can calculate the learning ability index X according to the learning progress index XA, the learning initiative index XB, the family-habit synergy index XC and the learning effectiveness index XD; A learning ability level threshold set is preset; the user's learning ability level is judged according to the learning ability indicator X and the learning ability level threshold set, and a course recommendation method is selected according to the user's learning ability level.

Images