Educational resource dynamic recommendation method and system based on multi-dimensional data analysis

Abstract

The invention relates to the technical field of learning behavior data prediction, in particular to an educational resource dynamic recommendation method and system based on multi-dimensional data analysis. Learning behavior data (interest, ability and the like), task completion condition data and auxiliary resource use data of a learner are comprehensively collected, and a multi-dimensional feature system is constructed. The auxiliary resource demand degree of the course is dynamically calculated based on the difference between the behavior data numerical characteristics and the task completion condition data, and the problem of disjunction between recommendation and learning states is solved; furthermore, in combination with the trend difference and numerical value characteristics of the use data of the auxiliary resources and the task completion condition data, the effect contribution degree of the auxiliary resources to the course is quantified, and the supporting ability of the auxiliary resources to learning achievements is objectively reflected. Finally, the demand degree and the effect contribution degree are fused, auxiliary resources with high demands and good effects are preferentially recommended, static recommendation lag is avoided, and the resource utilization efficiency and recommendation accuracy are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of learning behavior data prediction, in particular to an education resource dynamic recommendation method and system based on multi-dimensional data analysis. BACKGROUND

[0002] With the rapid development of information technology, the education field is undergoing profound changes. The traditional education resource allocation method is usually static and single, which is difficult to meet the current personalized, instant and diversified needs of learners. However, the effective use of education resources is of great significance to improve the quality of education, so the education resource recommendation method emerges as the times require, aiming to improve the learning experience and learning effect of learners through intelligent recommendation of education resources.

[0003] In the modern education environment, the needs of learners are increasingly complex, involving different learning goals, learning styles and learning progress. The existing education resource recommendation system often only considers the basic information of users, such as age, grade, etc., while ignoring the multi-dimensional factors of learners' interests, abilities, etc. Moreover, the recommendation system mainly relies on static models, lacking attention to the action logic of learners, which cannot fully understand the behavior patterns and decision-making processes of learners in the learning process, resulting in a disconnection between the recommended results and the actual needs of learners. SUMMARY

[0004] In order to solve the technical problem that the existing education resource recommendation system often only considers the basic information of users, such as age, grade, etc., while ignoring the multi-dimensional factors of learners' interests, abilities, etc., and the recommendation system mainly relies on static models, lacking attention to the action logic of learners, which cannot fully understand the behavior patterns and decision-making processes of learners in the learning process, resulting in a disconnection between the recommended results and the actual needs of learners, the purpose of the present application is to provide an education resource dynamic recommendation method and system based on multi-dimensional data analysis, the technical solution adopted is as follows:

[0005] An education resource dynamic recommendation method based on multi-dimensional data analysis, comprising:

[0006] Obtaining a plurality of learning behavior data of each type of course during learning, a plurality of completion condition data of each learning task, and usage data of each type of auxiliary resource in each learning task;

[0007] Based on the numerical characteristics of the learning behavior data of each type of course during learning and the change difference relationship between the completion condition data of the learning task, determining the auxiliary resource demand degree of each type of course;

[0008] analyze the difference between the use data and the completion data of each auxiliary resource in each type of learning task during the learning period, and determine the contribution degree of each auxiliary resource to each type of course according to the numerical characteristics of the use data;

[0009] According to the auxiliary resource requirement degree of each type of course and the effect contribution degree of the auxiliary resource, the auxiliary resource for the learner to learn each type of course is dynamically recommended.

[0010] Further, the auxiliary resource requirement degree acquisition method comprises:

[0011] According to the numerical characteristics of the learning behavior data of the learner when learning each type of course, the enthusiasm value of the learner for each type of course is determined.

[0012] According to the change difference relationship between the various completion data of the learning task when the learner learns each type of course, the learning hindering degree of the learner for each type of course is determined.

[0013] The sum of the normalized value of the enthusiasm value and the learning hindering degree of the learner for each type of course is taken as the auxiliary resource requirement degree of the learner for each type of course.

[0014] Further, the enthusiasm value acquisition method comprises:

[0015] The learning behavior data includes the number of visits, the visit duration of each visit, the number of questions and the number of replies;

[0016] The product of the mean value of all visit durations and the number of visits of each type of course is taken as the participation degree of the learner for each type of course.

[0017] The sum of the number of questions and the number of replies of each type of course is taken as the positive factor of the learner for each type of course.

[0018] The product of the participation degree and the positive factor of the learner for each type of course is taken as the enthusiasm value of the learner for each type of course.

[0019] Further, the learning hindering degree acquisition method comprises:

[0020] The completion data includes the task time and the task score rate.

[0021] fitting time spent by the learner on each learning task of all completed learning tasks of each type of course in the order of learning tasks to obtain a time spent curve, fitting score rate of the learner on each learning task of all completed learning tasks of each type of course in the order of learning tasks to obtain a score rate curve, and normalizing the values of the ordinate in the time spent curve and the score rate curve;

[0022] calculating the mean square error between the time spent curve and the score rate curve as a first hindering factor;

[0023] in the score rate fitting curve, obtaining the slope value at each data point, and mapping the sum of the slope values at all data points in a negative correlation to obtain a normalized value as a second hindering factor;

[0024] multiplying the first hindering factor and the second hindering factor of the learner for each type of course to obtain a normalized value as a learning hindering degree of the learner for each type of course.

[0025] Further, the effect contribution degree acquisition method comprises:

[0026] when the learner learns each type of course, analyzing the numerical characteristics of the usage data of each type of auxiliary resource in each completed learning task to determine an auxiliary factor of each type of auxiliary resource for each completed learning task;

[0027] based on the difference in the change trend between the contribution factor of each type of auxiliary resource for the learning task and the completion data of the learning task when the learner learns each type of course, determining a contribution factor of each type of auxiliary resource for each type of course;

[0028] when the learner learns each type of course, normalizing the sum of the average value of the auxiliary factor of each type of auxiliary resource for all completed learning tasks and the contribution factor corresponding to each type of auxiliary resource to obtain a normalized value as an effect contribution degree of each type of auxiliary resource for each type of course.

[0029] Further, the auxiliary factor acquisition method comprises:

[0030] the usage data includes the number of uses and the use duration of each use;

[0031] in each completed learning task of each type of course, normalizing the sum of the number of uses of each type of auxiliary resource and the average value of all use durations to obtain an auxiliary factor of each type of auxiliary resource for each completed learning task.

[0032] Further, the contribution factor acquisition method comprises:

[0033] Fitting the score rate of all completed learning tasks of each type of course by the learner during the learning period in the order of learning tasks to obtain a score rate curve, and fitting the assistance factor of each type of auxiliary resource to the learning tasks in all completed tasks of each type of course in the order of learning tasks to obtain an assistance factor curve;

[0034] Respectively in the score rate curve and the assistance factor curve, the slope value at each data point is obtained;

[0035] The sum of all slope values in the score rate curve is normalized to obtain an effect parameter;

[0036] Under each completed learning task, the absolute value of the difference between the slope values at the data points in the score rate curve and the assistance factor curve is calculated as a trend deviation factor;

[0037] The mean of all trend deviation factors is negatively correlated and normalized, and the product of the effect parameter is multiplied by the normalized value, and the normalized value of the product is obtained as the contribution factor of each type of auxiliary resource to each type of course.

[0038] Further, the dynamic recommendation of the auxiliary resource for the learner to learn each type of course includes:

[0039] For any type of course, the auxiliary resource demand degree of the course and the effect contribution degree of each type of auxiliary resource to the course are fused and analyzed to obtain a behavior dependence index of each type of auxiliary resource for the learner to learn the course.

[0040] All auxiliary resources corresponding to the course are sorted in descending order according to the behavior dependence index to obtain a recommendation sequence;

[0041] When the learner learns the course, all auxiliary resources corresponding to the course are dynamically recommended according to the recommendation sequence.

[0042] Further, the behavior dependence index acquisition method includes:

[0043] For any type of course, the auxiliary resource demand degree of the course and the Euclidean norm of the effect contribution degree of each type of auxiliary resource to the course are normalized to obtain a behavior dependence index of each type of auxiliary resource for the learner to learn the course.

[0044] An education resource dynamic recommendation system based on multi-dimensional data analysis includes a processor and a memory, the memory stores at least one instruction, at least one program, a code set or an instruction set, and the processor loads and executes the at least one instruction, the at least one program, the code set or the instruction set to realize the steps of the education resource dynamic recommendation method based on multi-dimensional data analysis.

[0045] The present application has the following beneficial effects:

[0046] The learning behavior data of the learner for each type of course during learning, the task completion data of each learning task, and the usage data of the auxiliary resources used are obtained, covering the multi-dimensional characteristics of the learner's interest, ability, resource preference, etc. Based on the numerical characteristics of the learning behavior data and the change difference of the learning task completion data, the auxiliary resource demand degree of each type of course is dynamically calculated, solving the problem of "recommended resources being out of touch with the current learning state of the learner", and ensuring the pertinence of resource recommendation. Further, by analyzing the trend difference between the usage data of the auxiliary resources in the learning task and the task completion data, and combining the numerical characteristics of the resource usage data, the effect contribution degree of each auxiliary resource to the course is quantified. This index objectively reflects the support ability of the auxiliary resources to the learning achievement improvement in the actual learning process of the learner. Finally, combining the auxiliary resource demand degree of each type of course and the effect contribution degree of the auxiliary resources, the auxiliary resources with high demand and good effect are preferentially recommended when the learner learns each type of course, improving the acceptance and usage effect of the recommended resources by the learner, and avoiding the problem of "static recommendation lagging behind the learner's demand". Through the technical path of "comprehensive data collection-demand dynamic quantification-effect accurate evaluation-demand and effect fusion-recommendation adjustment", the present application realizes the leap from "extensive recommendation" to "accurate dynamic recommendation", and improves the utilization efficiency and recommendation accuracy of educational resources. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0048] Figure 1 The method flowchart of the educational resource dynamic recommendation method based on multi-dimensional data analysis provided by an embodiment of the present application;

[0049] Figure 2 The method flowchart of the auxiliary resource demand degree acquisition method provided by an embodiment of the present application;

[0050] Figure 3 The schematic diagram of the time consumption curve and the score rate curve provided by an embodiment of the present application;

[0051] Figure 4 The method flowchart of the effect contribution degree acquisition method provided by an embodiment of the present application;

[0052] Figure 5 A score rate curve and an auxiliary factor curve provided by one embodiment of the present application;

[0053] Figure 6 A system block diagram of an education resource dynamic recommendation system based on multi-dimensional data analysis provided by one embodiment of the present application;

[0054] Figure 7 A system structure schematic diagram of an education resource dynamic recommendation system based on multi-dimensional data analysis provided by one embodiment of the present application. DETAILED DESCRIPTION

[0055] In order to further illustrate the technical means and effects taken by the present application to achieve the predetermined purposes, the following describes in detail the specific implementation, structure, features and effects of the education resource dynamic recommendation method and system based on multi-dimensional data analysis according to the present application, with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0056] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0057] The following specifically describes the specific scheme of the education resource dynamic recommendation method and system based on multi-dimensional data analysis provided by the present application, with reference to the accompanying drawings.

[0058] Please refer to Figure 1 which shows a method flowchart of an education resource dynamic recommendation method based on multi-dimensional data analysis provided by one embodiment of the present application. The method includes the following steps:

[0059] Step S1: Obtain multiple learning behavior data of each type of course, multiple completion condition data of each learning task, and usage data of each type of auxiliary resource in each learning task during learning of a learner.

[0060] Online education platforms and various learning management systems (LMS) can use dynamic recommendation systems to improve the learning experience of learners. When a learner registers on the platform and starts learning, the system will collect and analyze his behavior data in real time, including learning progress, learning duration, and selected courses, etc. At the same time, the system will also consider the multi-dimensional background information of the learner, such as personal interests, learning ability, and past learning performance, etc. Based on these data, the dynamic recommendation system can intelligently recommend the most suitable learning resources for the learner.

[0061] Firstly, in the embodiment of the present application, the learning behavior data of the learner during the learning period for each type of course (Chinese, mathematics, English, programming, etc.) is obtained, which can reflect the specific learning state, participation and learning enthusiasm of the learner during the learning of each type of course. At the same time, the completion condition data of each learning task during the learning of each type of course is also obtained, which can reflect the actual learning achievement conversion condition. Finally, the usage condition data of each type of auxiliary resource (document, picture, case, etc.) in each learning task is obtained, which can represent the effect contribution of each type of auxiliary resource to the learning of the learner during the learning process. According to the multi-dimensional data obtained as described above, the learner portrait is constructed, the "one-sided information" is avoided, and the recommendation algorithm is used to generate the personalized auxiliary resource recommendation for each type of course for each learner.

[0062] The learning behavior data of each type of course includes the access times of the learner to the course (such as the number of times of clicking the course button), the access time of each access (such as the time interval from entering the course to exiting the course, in minutes), the number of questions (such as the number of times of publishing questions in the course discussion area and the question system), and the number of replies (such as the number of times of answering other people's questions and sharing learning notes in the course discussion area). The completion condition data includes the task time spent by the learner in completing each learning task when learning each type of course (such as the time interval from clicking the start task to submitting the answer, in minutes) and the task score rate (the ratio of the score in each learning task to the total score, such as 80 points in a 100-point task, the score rate is 80%). The usage condition data includes the number of times of using each type of auxiliary resource (such as the number of times of clicking to view pictures) and the usage time (such as the time interval from viewing pictures to closing pictures, in minutes) by the learner in completing each learning task when learning each type of course. The above data can be obtained through the log record in the online education platform or various learning management systems, and the time period for data acquisition is set as the time when each type of course starts to learn to the current time.

[0063] The collection and acquisition of the learner personal information data in the embodiment of the present application are authorized by the relevant user, which does not violate the relevant laws and regulations and does not violate the public order and good customs.

[0064] Step S2: determining the auxiliary resource demand degree of each type of course based on the numerical characteristics of the learning behavior data of each type of course during the learning period, and the change difference relationship between the completion condition data of the learning task.

[0065] The numerical characteristics of the learning behavior data reflect the interest degree of the learner in the course, that is, the active participation degree, which helps to better identify the preferences of the learner, and the completion data of the learning task can represent the understanding and mastery degree of the learner to the course content, so the change difference relationship between the task time cost and the score rate of the learning task is analyzed to reveal the "obstruction degree" encountered by the learner in the learning process, and the learning behavior data of the numerical characteristics are combined to determine the auxiliary resource demand degree of the learner in learning each type of course. The index can help ensure that the recommended resources match the current learning state of the learner in the subsequent recommendation process, and maintain the synchronization between the needs and behaviors of the learner.

[0066] Preferably, in an embodiment of the present application, the auxiliary resource demand degree acquisition method comprises:

[0067] Please refer to Figure 2 which shows a method flowchart of the auxiliary resource demand degree acquisition method in an embodiment of the present application, and the method comprises the following steps:

[0068] Step S201: Based on the numerical characteristics of the learning behavior data of the learner in learning each type of course, determine the active degree value of the learner to each type of course.

[0069] The product of the mean value of the access time length of each type of course and the access frequency after normalization is taken as the participation degree of the learner to each type of course. The greater the access frequency, the more frequent the learner accesses the course. At the same time, the longer the access time length, the more likely it is to learn deeply, so the greater the participation degree obtained by combining the two, the greater the learning depth of the learner to the course, the higher the participation degree, and then the greater the active degree.

[0070] The sum of the question frequency and the reply frequency of each type of course is taken as the active factor of the learner to each type of course. The question reflects active thinking, and the reply reflects knowledge sharing and interaction. The greater the question frequency and the reply frequency, the higher the learning interaction enthusiasm of the learner, so the greater the active factor obtained by combining the two, the higher the enthusiasm of the learner to learn the course.

[0071] Based on the foregoing analysis, the product of the participation degree and the active factor of the learner to each type of course after normalization is taken as the active degree value of the learner to each type of course. The greater the active degree value, the higher the learning input degree of the learner to the course, and the stronger the enthusiasm.

[0072] It should be noted that normalization is a technical means familiar to those skilled in the art, and the selection of the normalization function can be linear normalization or standard normalization, and the specific normalization method is not limited herein.

[0073] Step S202: According to the change difference relationship between the multiple completion data of the learning task when the learner learns each type of course, the learning hindering degree of the learner for each type of course is determined.

[0074] The learner's enthusiasm for each type of learning course is evaluated in step S201, in order to further evaluate the influence of the enthusiasm on the learning effect, the focus can be turned to the completion of the learning task when learning each type of course, the task completion not only reflects the understanding and mastery degree of the learner to the course content, but also can reveal the action logic in the learning process, providing basis for the dynamic recommendation of subsequent educational resources.

[0075] Generally speaking, in the process of the course, the difficulty of the course will gradually increase, and there are multiple phased learning tasks, the time spent on all completed learning tasks of each type of course during the learning period of the learner is fitted according to the order of the learning task to obtain the time spent curve, the score rate of all completed learning tasks of each type of course during the learning period of the learner is fitted according to the order of the learning task to obtain the score rate curve, and the values of the ordinate in the time spent curve are normalized, the ordinate in the score rate curve is expressed by a decimal, and the time spent curve and the score rate curve respectively represent the change characteristics of the learning efficiency and the knowledge mastery of the learner in the course learning process. Please refer to Figure 3 which shows the schematic diagram of the time spent curve and the score rate curve in an embodiment of the application.

[0076] The mean square error between the time spent curve and the score rate curve is calculated as the first hindering factor, the greater the first hindering factor, the greater the degree of reverse change between the time spent and the score rate of the learning task, that is, the more significant the phenomenon of time-consuming increase but score rate decrease, which can be regarded as the greater learning hindrance encountered by the learner in the learning process, and the index directly reflects the correlation characteristics between efficiency and effect.

[0077] Then in the score rate fitting curve, the slope value at each data point is obtained, the slope value captures the change fluctuation direction of the score rate, if the slope value is mostly negative, it means that the score rate is continuously decreasing, on the contrary, if the slope value is mostly positive, it means that the score rate is stable or increasing, therefore, when the sum of all slope values is positive and greater, it means that the encountered hindrance is smaller, therefore, the sum of the slope values of all data points is negatively correlated, the logical relationship is corrected, and the second hindering factor is obtained, at this time, the greater the second hindering factor, the worse the knowledge point mastery of the learner in the learning process, and the auxiliary resources are needed to help or consolidate. The negative correlation mapping processing here adopts the formula exp(-x), wherein exp() represents the exponential function with natural constant e as the base, and x represents the independent variable.

[0078] Finally, the normalized value of the product of the first hindering factor and the second hindering factor of the learner for each type of course is taken as the learning hindering degree of the learner for each type of course. The learning hindering degree combines the efficiency-effectiveness contradiction and the weak knowledge point accumulation, and thus can reflect the comprehensive hindering strength of the learner for the course. The greater the learning hindering degree is, the worse the mastery degree of the learner for the course is, and the higher the demand degree of the learner for the auxiliary resource is.

[0079] It should be noted that normalization is a technique well known to those skilled in the art, and the selection of the normalization function can be linear normalization or standard normalization, and the specific normalization method is not limited herein. In the embodiment of the present application, curve fitting can use the least square method, which is a well-known technique, and the specific process is not described herein.

[0080] Step S203: The active degree value and the learning hindering degree of the learner for each type of course are fused to obtain the auxiliary resource demand degree of the learner for each type of course.

[0081] Based on the analysis in the foregoing steps S201 and S202, the greater the active degree value of the learner for each type of course is, the stronger the learning enthusiasm of the learner for the course is; the greater the learning hindering degree is, the worse the mastery degree of the learner for the course is. The sum value of the active degree value and the learning hindering degree of the learner for each type of course is normalized to obtain the auxiliary resource demand degree of the learner for each type of course. The greater the auxiliary resource demand degree is, the greater the learning hindering in the course with high participation degree is, which can be caused by poor understanding of the course content and poor mastery of the knowledge points. Therefore, the demand degree of the learner for the auxiliary resource for the course is higher, so as to obtain timely help and understanding through the auxiliary educational resource. The normalization is a technique well known to those skilled in the art, and the selection of the normalization function can be linear normalization or standard normalization, and the specific normalization method is not limited herein.

[0082] Step S3: Analyzing the difference between the change trend of the usage data and the completion data of each type of auxiliary resource in the learning task of each type of course, and combining the numerical characteristics of the usage data to determine the effect contribution degree of each type of auxiliary resource to each type of course.

[0083] In the process of education resource recommendation based on multi-dimensional data analysis, in order to further respond to the goal of personalized education resource recommendation under the action logic, dynamic observation and analysis of the actual learning behavior process of the learner can be introduced. In this step, by analyzing the demand for auxiliary resources of the learner in the actual course learning process, focusing on the behavior trajectory of learning task execution, the use of auxiliary resources and other key features in the learning process, the resource demand portrait of the learner in the learning course is constructed, thereby forming the transition from "static demand judgment" to "dynamic behavior deduction", and the responsiveness and adaptability of the recommendation system to the learning process are strengthened.

[0084] By analyzing the difference in the trend of the use of each auxiliary resource in the learning task and the completion of the learning task during the learning of each type of course by the learner, and combining the numerical characteristics of the use data, the effect contribution degree of each auxiliary resource to each type of course can be quantified. This index objectively reflects the support ability of the auxiliary resource to the improvement of learning achievement in the actual course learning process, and can be used as an effective reference index for subsequent resource recommendation.

[0085] Preferably, in an embodiment of the present application, the method for obtaining the effect contribution degree comprises:

[0086] Please refer to Figure 4 which shows a method flowchart of the method for obtaining the effect contribution degree in an embodiment of the present application, and the method comprises the following steps:

[0087] Step S301: When the learner learns each type of course, analyze the numerical characteristics of the use data of each type of auxiliary resource in each completed learning task, and determine the auxiliary factor of each type of auxiliary resource to each completed learning task.

[0088] In each completed learning task of each type of course, the more the number of uses of a certain type of auxiliary resource, the more frequently the auxiliary resource is called, and the longer the use time, the greater the demand of the learner for the auxiliary resource. Therefore, the sum of the normalized values of the number of uses of each type of auxiliary resource and the average of all use times is used as the auxiliary factor of each type of auxiliary resource to each completed learning task. Based on the foregoing logic, the greater the auxiliary factor, the higher the dependence of the learner on the auxiliary resource if there is an understanding or memory obstacle in the learning of the course. The normalization is a technical means familiar to those skilled in the art, and the selection of the normalization function can be linear normalization or standard normalization, and the specific normalization method is not limited herein.

[0089] Step S302: determining the contribution factor of each auxiliary resource to each type of course based on the difference in the change trend between the contribution factor of each auxiliary resource to the learning task and the completion data of the learning task when the learner learns each type of course.

[0090] The score rate of the learner to all completed learning tasks of each type of course during learning is fitted according to the order of the learning tasks to obtain a score rate curve (the same as the processing in step S202, and the ordinate is expressed by a decimal number), and the auxiliary factor of each auxiliary resource to the learning task is fitted according to the order of the learning tasks in all completed tasks of each type of course to obtain an auxiliary factor curve. The score rate curve and the auxiliary factor curve respectively represent the change characteristics of the knowledge mastery of the learner and the change characteristics of the assistance degree of each auxiliary resource to the learner in the course learning process. Please refer to Figure 5 which shows the schematic diagram of the score rate curve and the auxiliary factor curve in one embodiment of the present application.

[0091] Then, the sum of all slope values in the score rate curve is calculated. If the sum is positive and larger, it indicates that the learning effect of the learner in learning the course is continuously enhanced as a whole, and vice versa. The sum is normalized to obtain an effect parameter. Based on the foregoing logic, the larger the effect parameter is, the better the learning effect is. Since the sum here can be positive or negative, the normalization method adopts a Sigmoid() function.

[0092] Next, the slope values at the data points in the score rate curve and the auxiliary factor curve are obtained respectively. The slope values can capture the dynamic trend of the score rate and the use of the auxiliary resource. The absolute value of the difference between the slope values at the data points in the score rate curve and the auxiliary factor curve is calculated as a trend deviation factor under each completed learning task (under each horizontal coordinate in the curve). The trend deviation factor is used to quantify the deviation degree between the use trend of the auxiliary resource and the learning effect trend. The smaller the value is, the greater the deviation degree is, that is, the higher the change synchronization between them is. Therefore, the mean of all trend deviation factors is negatively correlated and normalized to correct the logical relationship, as the corresponding change consistency of each auxiliary resource. The larger the change consistency is, the closer the change relationship between the auxiliary resource and the score rate is. The negatively correlated and normalized processing here adopts the formula exp(-x), where exp() represents the exponential function with the natural constant e as the base, and x represents the independent variable.

[0093] Finally, when the effect parameter of the learner learning each course is greater, and the consistency of the change between the certain auxiliary resource and the score rate is also higher, it can be considered that the contribution degree of the auxiliary resource to the course learning is higher, therefore, the value obtained by multiplying the effect parameter and the change consistency corresponding to each auxiliary resource after normalization is used as the contribution factor of each auxiliary resource to each course, and the greater the contribution factor, the higher the completion condition of the learning task in the process of completing the learning task, and the higher the contribution degree of the auxiliary resource to the learning effect. The normalization is a technical means familiar to those skilled in the art, and the selection of the normalization function can be linear normalization or standard normalization, and the specific normalization method is not limited here.

[0094] It should be noted that the curve fitting in the embodiment of the application can adopt the least square method, which is a known technology, and the specific process is not described here.

[0095] Step S303: When learning each course, the auxiliary factor of each auxiliary resource to all completed learning tasks is fused with the contribution factor of each auxiliary resource to each course, so as to obtain the effect contribution degree of each auxiliary resource to each course.

[0096] Based on the analysis in steps S301 and S302, the greater the auxiliary factor of a certain auxiliary resource in a certain learning task, the greater the degree of dependence of the learner on the auxiliary resource when learning the course; the greater the contribution factor of a certain auxiliary resource to the course, the higher the contribution degree of the auxiliary resource to the learning effect when the learner learns the course; therefore, the two indexes are positively correlated with the effect contribution degree of the auxiliary resource to the course, so the mean value of the auxiliary factor of each auxiliary resource to all completed learning tasks is normalized with the sum value of the contribution factor corresponding to each auxiliary resource, and the value after normalization is used as the effect contribution degree of each auxiliary resource to each course. Based on the foregoing logic, the greater the effect contribution degree of a certain auxiliary resource, the greater the support ability of the auxiliary resource in the actual learning process to the learning achievement improvement of the learner in a certain course in the process of completing the learning task. The normalization is a technical means familiar to those skilled in the art, and the selection of the normalization function can be linear normalization or standard normalization, and the specific normalization method is not limited here.

[0097] Step S4: According to the auxiliary resource requirement degree of each course and the effect contribution degree corresponding to the auxiliary resource, the auxiliary resource for the learner learning each course is dynamically recommended.

[0098] The purpose of the step is to jointly model the effect contribution of each auxiliary resource to the learning process of each type of course and the auxiliary resource demand of the learner in the learning process of each type of course, to comprehensively evaluate the dependence of the learner on a certain auxiliary resource when learning each type of course, and obtain a behavior dependence index, which not only reflects the positive influence of the auxiliary resource on the learning result, but also integrates the active calling behavior of the learner, and thus can be used as a learning feature expression quantity with more behavior explanation to support the generation of subsequent personalized resource recommendation strategies.

[0099] Preferably, in one embodiment of the present application, the dynamic recommendation of the auxiliary resource for the learner to learn each type of course comprises:

[0100] For any type of course, the auxiliary resource demand of the type of course and the Euclidean norm of the effect contribution of each auxiliary resource to the type of course are normalized as the behavior dependence index of the learner on each auxiliary resource when learning the type of course. The behavior dependence index quantifies the comprehensive dependence strength of the learner on a certain auxiliary resource by comprehensively evaluating the demand of the learner for the auxiliary resource and the contribution of the auxiliary resource to the learning effect. The greater the value, the higher the dependence of the learner on the auxiliary resource in the learning process, and the greater the help of the auxiliary resource to the learner, so the auxiliary resource needs to be recommended first. The normalization is a technical means familiar to those skilled in the art, and the selection of the normalization function can be linear normalization or standard normalization, and the specific normalization method is not limited herein.

[0101] All auxiliary resources corresponding to the type of course are sorted in descending order according to the behavior dependence index to obtain a recommendation sequence, and finally the dynamic recommendation of all auxiliary resources corresponding to the type of course is performed according to the recommendation sequence when the learner learns the type of course.

[0102] In summary, the learning behavior data of the learner on each type of course during learning, the task completion data of each learning task, and the usage data of the auxiliary resources used are obtained, covering the multi-dimensional characteristics of the learner's interest, ability, resource preference, etc. Based on the numerical characteristics of the learning behavior data and the change difference of the learning task completion data, the auxiliary resource demand degree of each type of course is dynamically calculated, the problem of "recommended resources being out of touch with the current learning state of the learner" is solved, and the pertinence of resource recommendation is ensured. Further, by analyzing the trend difference between the usage data of the auxiliary resources in the learning task and the task completion data, and combining the numerical characteristics of the resource usage data, the effect contribution degree of each auxiliary resource to the course is quantified. The index objectively reflects the support ability of the auxiliary resources to the learning achievement improvement in the actual learning process of the learner. Finally, the auxiliary resource demand degree of each type of course and the effect contribution degree of the auxiliary resources are combined, so that the auxiliary resources with high demand and good effect are preferentially recommended when the learner learns each type of course, the acceptance and usage effect of the learner to the recommended resources are improved, and the problem of "static recommendation lagging behind the learner's demand" is avoided. Through the technical path of "comprehensive data collection-demand dynamic quantification-accurate effect evaluation-demand and effect fusion-recommendation adjustment", the invention embodiment realizes the leap from "extensive recommendation" to "accurate dynamic recommendation", and improves the utilization efficiency and recommendation accuracy of the educational resources.

[0103] The embodiment of the present application also provides an educational resource dynamic recommendation system based on multi-dimensional data analysis, please refer to Figure 6 which shows a system block diagram, including a data acquisition module 601 for realizing the step S1 in the above method embodiment; an auxiliary resource demand analysis module 602 for realizing the step S2 in the above method embodiment; an auxiliary resource contribution degree analysis module 603 for realizing the step S3 in the above method embodiment; and a dynamic recommendation module 604 for realizing the step S4 in the above method embodiment.

[0104] It should be noted that the system provided in the above embodiment is only exemplified by the division of the above functional modules. In actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the computer device is divided into different functional modules to complete all or part of the above described functions. In addition, the educational resource dynamic recommendation system based on multi-dimensional data analysis and the educational resource dynamic recommendation method based on multi-dimensional data analysis provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0105] Please refer to Figure 7Fig. 1 shows a system structure schematic diagram of an education resource dynamic recommendation system based on multi-dimensional data analysis according to an embodiment of the present application, which comprises a processor 700, a memory 701, a bus 702 and a communication interface 703, the processor 700, the communication interface 703 and the memory 701 are connected through the bus 702; wherein the memory 701 can contain a high-speed random access memory, the bus 702 can be an ISA bus, a PCI bus or an EISA bus, etc., the processor 700 can be an integrated circuit chip with signal processing capability; the memory 701 stores at least one instruction, at least one program, a code set or an instruction set, which are loaded and executed by the processor to realize the steps of the education resource dynamic recommendation method based on multi-dimensional data analysis.

[0106] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or can be advantageous.

[0107] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the difference from other embodiments.

Claims

1. A method for dynamic recommendation of educational resources based on multi-dimensional data analysis, characterized in that, The method includes: Acquire learners' various learning behaviors during the learning process, various completion data for each learning task, and data on the use of each auxiliary resource in each learning task; Based on the numerical characteristics of learning behavior data for each type of course during the learning period, and the relationship between the changes in the completion data of learning tasks, the demand for auxiliary resources for each type of course is determined. The study analyzes the differences in the trends between the usage and completion data of each auxiliary resource in the learning tasks of each type of course during the learning period, and determines the contribution of each auxiliary resource to the effectiveness of each type of course by combining the numerical characteristics of the usage data. Based on the demand for supplementary resources for each type of course and the corresponding contribution of supplementary resources to the learning effect, supplementary resources are dynamically recommended for learners when learning each type of course.

2. The method for dynamic recommendation of educational resources based on multi-dimensional data analysis according to claim 1, characterized in that, The method for obtaining the auxiliary resource demand includes: Based on the numerical characteristics of learners' learning behavior data when learning each type of course, the learners' level of enthusiasm for each type of course is determined. Based on the differences in the data on the completion of various learning tasks by learners when learning each type of course, the learning barrier level of learners for each type of course is determined. The normalized sum of learners' enthusiasm and learning obstacles for each type of course is used as the learners' demand for supplementary resources for each type of course.

3. The method for dynamic recommendation of educational resources based on multi-dimensional data analysis according to claim 2, characterized in that, The method for obtaining the positivity value includes: The learning behavior data includes the number of visits, the duration of each visit, the number of questions asked, and the number of replies. The normalized value of the product of the average access duration and the number of accesses for each type of course is taken as the learner's level of participation in each type of course. The sum of the number of questions asked and the number of responses for each type of course is used as the learner's positive factor for each type of course. The normalized value of the product of learner participation and positive factor for each type of course is used as the learner's positive level value for each type of course.

4. The method for dynamic recommendation of educational resources based on multi-dimensional data analysis according to claim 2, characterized in that, The methods for obtaining the learning obstacle level include: The completion data includes the task time spent and the task score rate; The time spent by learners on all completed learning tasks for each type of course during the learning period is fitted according to the order of the learning tasks to obtain the time spent curve. The score rate of learners on all completed learning tasks for each type of course during the learning period is fitted according to the order of the learning tasks to obtain the score rate curve. The values ​​of the vertical axis in the time spent curve and the score rate curve are normalized. The mean square error between the time spent curve and the score rate curve is calculated and used as the first obstacle factor. In the score rate fitting curve, the slope value at each data point is obtained, and the sum of the slope values ​​at all data points is normalized by negative correlation mapping, which is used as the second hindering factor. The normalized value of the product of the first and second obstacle factors for each type of course is used as the learning obstacle degree for each type of course.

5. The method for dynamic recommendation of educational resources based on multi-dimensional data analysis according to claim 1, characterized in that, The methods for obtaining the contribution of the effect include: When learners are learning each type of course, analyze the numerical characteristics of the usage data of each auxiliary resource in each completed learning task to determine the auxiliary factor of each auxiliary resource for each completed learning task. Based on the differences in the changing trends between the contribution factor of each auxiliary resource to the learning task and the data on the completion of the learning task when learners are learning each type of course, the contribution factor of each auxiliary resource to each type of course is determined. When learners study each type of course, the mean of the auxiliary factors of each auxiliary resource for all completed learning tasks, and the sum of the contribution factors corresponding to each auxiliary resource are normalized to the value of each auxiliary resource's contribution to the effectiveness of each type of course.

6. The method for dynamic recommendation of educational resources based on multi-dimensional data analysis according to claim 5, characterized in that, The method for obtaining the auxiliary factor includes: The usage data includes the number of times it is used and the duration of each use. In each completed learning task of each course type, the normalized sum of the number of times each auxiliary resource was used and the average duration of all uses will be used as an auxiliary factor for each auxiliary resource in each completed learning task.

7. The method for dynamic recommendation of educational resources based on multi-dimensional data analysis according to claim 5, characterized in that, The method for obtaining the contribution factor includes: The score rate curve is obtained by fitting the score rate of learners for all completed learning tasks of each course type in the order of learning tasks. In all completed tasks of each course type, the auxiliary factor of each auxiliary resource is fitted in the order of learning tasks to obtain the auxiliary factor curve. Obtain the slope value at each data point from the score rate curve and the auxiliary factor curve, respectively; The sum of all slope values ​​in the score rate curve is normalized and used as the effect parameter. For each completed learning task, the absolute value of the difference between the slope values ​​of the score rate curve and the auxiliary factor curve at the data points is calculated as the trend bias factor. The mean of all trend deviation factors is negatively correlated and normalized, then multiplied by the effect parameter, and the resulting product is normalized. This product is then used as the contribution factor of each auxiliary resource to each type of course.

8. The method for dynamic recommendation of educational resources based on multi-dimensional data analysis according to claim 1, characterized in that, The dynamic recommendation of supplementary resources for learners when learning each type of course includes: For any type of course, the demand for auxiliary resources for this type of course and the contribution of each auxiliary resource to the effectiveness of this type of course are analyzed together to obtain the behavioral dependence index of learners on each auxiliary resource when learning this type of course. All supplementary resources corresponding to this type of course are sorted in descending order according to behavioral dependence index to obtain a recommendation sequence; When learners study this type of course, all supplementary resources corresponding to this type of course are dynamically recommended according to the recommended sequence.

9. The method for dynamic recommendation of educational resources based on multi-dimensional data analysis according to claim 8, characterized in that, The methods for obtaining the behavior dependency index include: For any type of course, the demand for auxiliary resources for this type of course and the Euclidean norm of the contribution of each auxiliary resource to the effectiveness of this type of course are normalized to serve as the behavioral dependence index of learners on each auxiliary resource when learning this type of course.

10. A dynamic recommendation system for educational resources based on multi-dimensional data analysis, characterized in that, The method includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, code set, or instruction set, and the steps of the method for dynamically recommending educational resources based on multi-dimensional data analysis as described in any one of claims 1-9 are implemented when the at least one instruction, at least one program, code set, or instruction set is loaded and executed by the processor.

Images