Learning resource recommendation method and system based on big data

Abstract

The present application relates to the field of information and communication technology in the field of education, and in particular to a learning resource recommendation method and system based on big data. The learning resource recommendation method based on big data includes: dividing the student's learning course into multiple learning nodes; after the student completes the learning of the first learning node, using a benchmark test group consisting of multiple benchmark test questions to conduct an initial test on each student to obtain the score distribution of the current learning node; looping the following steps so that the student's score distribution in the test of subsequent learning nodes gradually tends to a predetermined distribution: generating learning resources for each student's future learning node based on the position of each student in the current score distribution; adjusting the benchmark test group of the subsequent learning node based on the use of the generated learning resources by each student to test the student at the subsequent learning node. The technical solution provided by this application can improve the teaching effect.

Description

Technical Field

[0001] The present application relates to the field of information and communication technology in the field of education, and in particular to a learning resource recommendation method and system based on big data. Background Art

[0002] In the field of educational technology, the application of big data technology to achieve personalized learning and data-driven teaching adjustments are two major research directions. Existing technologies generally adjust teaching content and difficulty based on students' learning behavior and performance. The core of these systems is to collect a large amount of student learning data, understand students' learning needs and progress through machine learning and data analysis technology, and provide corresponding learning resources and support. However, these systems often rely on standardized tests and fixed learning paths, which are not enough to cope with the rapid changes in student performance and individual differences, and the teaching effect is not good. Summary of the invention

[0003] In order to solve the above technical problems or at least partially solve the above technical problems, the present application provides a learning resource recommendation method and system based on big data, which can improve the teaching effect.

[0004] In a first aspect, the present application provides a learning resource recommendation method based on big data, and the learning resource recommendation method based on big data includes:

[0005] Divide students' learning courses into multiple learning nodes;

[0006] After the students have completed the study of the first learning node, each student is initially tested using a benchmark test group consisting of multiple benchmark test questions to obtain the score distribution of the current learning node;

[0007] The following steps are executed in a loop so that the distribution of students' scores in the tests of subsequent learning nodes gradually approaches the predetermined distribution:

[0008] Generate learning resources for each student's future learning nodes based on their position in the current score distribution;

[0009] According to each student's use of the generated learning resources, the benchmark test question group of the subsequent learning node is adjusted to test the student at the subsequent learning node.

[0010] Optionally, generating learning resources for each student's future learning node according to each student's position in the current score distribution includes the following steps:

[0011] Obtain each student's learning resource recommendation score based on the learning resource benchmark score and each student's score distribution;

[0012] The similarity benchmark and difficulty benchmark for recommending after-class exercises are generated based on each student's learning resource recommendation score. In the next learning node, for each student, based on the student's similarity benchmark and difficulty benchmark, questions with similarity and difficulty closest to the benchmark questions are selected from the question bank and recommended to the student as after-class exercises.

[0013] Optionally, the learning resource benchmark score Obtained by the following steps:

[0014]

[0015] in:

[0016]

[0017]

[0018]

[0019] Wherein, k is the preset first adjustment coefficient, is the preset second adjustment coefficient, is the standard deviation of the current learning node score distribution, is the mean of the score distribution of the current learning node.

[0020] Optionally, obtain the recommended learning resource score for each student based on the learning resource benchmark score and each student's grade distribution , including the following steps:

[0021]

[0022] in, is the preset third adjustment coefficient, is the student's score in this learning node. is the mean of students’ scores at this learning node.

[0023] Optionally, generating a similarity benchmark and a difficulty benchmark for after-class exercise recommendation based on each student's learning resource recommendation score includes the following steps:

[0024]

[0025]

[0026] Wherein, d is the preset fourth adjustment coefficient, and l is the preset fifth adjustment coefficient.

[0027] Optionally, adjusting the benchmark test questions of future learning nodes according to the usage of the generated learning resources by each student includes the following steps:

[0028] In the next learning node, the scores of each student completing the recommended homework exercises are obtained;

[0029] Calculate the reasonable value of the difficulty setting of basic test questions :

[0030]

[0031] N is the total number of students, is the score of the i-th student completing the recommended homework exercises, is the average score of each student completing the recommended homework exercises, is the standard deviation of each student's scores on the recommended homework exercises;

[0032] when When the value is between [-0.5, 0.5], no adjustment is made to the benchmark test set;

[0033] when When <-0.5, add difficult questions to the baseline test set;

[0034] when >0.5, add easy questions to the benchmark test set.

[0035] In a second aspect, the present application provides a learning resource recommendation system based on big data, wherein the big data-based auxiliary learning system includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set are loaded and executed by the processor to implement a learning resource recommendation method based on big data as described in any one of the first aspects.

[0036] Compared with the prior art, the technical solution provided by this application has the following advantages:

[0037] On the one hand, the big data-based learning resource recommendation system provided by this application first uses benchmark test questions to conduct an initial test on students to obtain the score distribution of each learning node. Subsequently, based on the specific performance of each student, the system will generate personalized learning resources for future learning nodes. These resources are finely adjusted in similarity and difficulty to match the learning ability of each student, thereby ensuring that each student can learn on the basis of their current ability.

[0038] Secondly, the system also adjusts the difficulty and similarity of subsequent students' learning resources according to the speed of change in the distribution of students' grades after using the learning resources. This dynamic adjustment mechanism not only takes into account the uneven changes in students' learning abilities, but also overcomes the problem of students' rigid thinking or lack of thinking that may be caused by improper application of learning resources. By analyzing a large amount of students' performance data, continuously tracking students' performance and adjusting teaching resources and methods according to actual conditions, the system maximizes the advantages of big data analysis and improves teaching effectiveness.

[0039] Thirdly, the benchmark test set consisting of benchmark test questions is often designed based on the teacher's experience. Therefore, for different student groups, this benchmark test set may be too simple or too difficult, and the design of the benchmark test set itself is flawed and cannot reasonably test the level of the student group. This system also adjusts the benchmark test set of the current learning node according to the actual situation of the student group and the use of learning resources by students at the current learning node, avoiding the technical problem that the level of the student group cannot be reasonably tested due to the initial design of the benchmark test set itself being too simple or too difficult.

[0040] Therefore, the technical solution provided by this application effectively improves the pertinence of teaching resources, thereby improving teaching effectiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A flowchart of a learning resource recommendation method based on big data provided in an embodiment of the present application. DETAILED DESCRIPTION

[0042] The technical solution in this application will be described below in conjunction with the accompanying drawings.

[0043] In the following description, many specific details are set forth to facilitate a full understanding of the present application, but the present application can also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only part of the embodiments of the present application, not all of the embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0044] In a first aspect, the present application provides a learning resource recommendation method based on big data, and the learning resource recommendation method based on big data includes:

[0045] S101: Divide the student's learning courses into multiple learning nodes.

[0046] Specifically, in the embodiment of the present application, the learning course is divided into multiple learning nodes according to the division of chapters of knowledge points in specific subjects.

[0047] S102: After the student completes the learning of the first learning node, an initial test is performed on each student using a benchmark test question group consisting of a plurality of benchmark test questions to obtain the score distribution of the current learning node.

[0048] Specifically, the benchmark test questions here are set by teachers based on their experience. Generally speaking, the benchmark test questions for each learning node are questions that the teacher has formulated or used in the past to test the knowledge learned in the past when teaching the learning node.

[0049] The following steps are executed in a loop so that the distribution of students' scores in the tests of subsequent learning nodes gradually approaches the predetermined distribution:

[0050] S103: Generate learning resources for each student's future learning nodes according to each student's position in the current score distribution.

[0051] Specifically, generating learning resources for each student's future learning node according to each student's position in the current score distribution includes the following steps:

[0052] Obtain each student's learning resource recommendation score based on the learning resource benchmark score and each student's score distribution;

[0053] Specifically, the learning resource benchmark score Obtained by the following steps:

[0054]

[0055] in:

[0056]

[0057]

[0058]

[0059] Wherein, k is the preset first adjustment coefficient, is the preset second adjustment coefficient, is the standard deviation of the current learning node score distribution, is the mean of the score distribution of the current learning node.

[0060] Specifically, the first adjustment coefficient and the second adjustment coefficient are manually set empirical coefficients.

[0061] The target mean high cutoff and target mean low cutoff need to be set according to the actual test paper scores. Generally speaking, 70% of the test paper scores are the target mean low cutoff, and 90% of the test paper scores are the target mean high cutoff. t is the current learning node, and t-1 is the previous learning node.

[0062] Specifically, the learning resource recommendation score for each student is obtained based on the learning resource benchmark score and the score distribution of each student. , including the following steps:

[0063]

[0064] in, is the preset third adjustment coefficient, is the student's score in this learning node. is the mean of students’ scores at this learning node.

[0065] Specifically, the third adjustment coefficient is an empirical parameter set artificially.

[0066] The similarity benchmark and difficulty benchmark for recommending after-class exercises are generated based on each student's learning resource recommendation score. In the next learning node, for each student, based on the student's similarity benchmark and difficulty benchmark, questions with similarity and difficulty closest to the benchmark questions are selected from the question bank and recommended to the student as after-class exercises.

[0067] Specifically, the similarity and difficulty of the questions in the question bank compared to each benchmark question need to be manually marked based on experience. Generally speaking, the similarity benchmark is divided from 0 to 2, and the difficulty benchmark is divided from 0 to 10. It is only necessary to adjust the fourth adjustment coefficient, the fifth adjustment coefficient and the corresponding basic constant according to the actual test paper score, so that the final similarity benchmark and difficulty benchmark are mapped to the corresponding marking interval.

[0068] Specifically, generating a similarity benchmark and a difficulty benchmark for after-class exercise recommendation based on each student's learning resource recommendation score includes the following steps:

[0069]

[0070]

[0071] Wherein, d is the preset fourth adjustment coefficient, and l is the preset fifth adjustment coefficient.

[0072] Specifically, the fourth adjustment coefficient and the fifth adjustment coefficient are artificially set empirical parameters.

[0073] The technical problem to be solved is that, first of all, technicians in this field need to be aware that the adjustment process needs to carefully consider the non-uniform changes in students' learning abilities. On the one hand, students' learning abilities need to be stimulated after a certain foundation, so for some students, their scores may be relatively low at the beginning, but they will gradually increase later. Therefore, since the distribution of students' scores affects the similarity of the homework exercises recommended to students, the lower the student's scores, the more similar their homework exercises are to the benchmark test questions used in the exam. If the student's learning ability is solidified at the beginning, it will lead to problems such as rigid thinking or lack of thinking of students due to improper application of learning resources. This solidification is manifested in a reduction in the difficulty of the exam, resulting in a large concentration of student score distribution.

[0074] On the other hand, if the test questions are too easy or too difficult, this will be reflected in the fact that the students' scores will suddenly gather together. When the scores are generally concentrated in the lower position, it may indicate that the questions are too difficult, while when the scores are generally concentrated in the higher position, it may indicate that the questions are too easy.

[0075] The specific working principle is as follows: First, the calculation of the learning resource benchmark score B(t) of this application mainly consists of two parts. The first part consists of This reflects the change in the speed of change in the standard deviation of the distribution of student scores. If this value is large, it indicates that the concentration of the distribution of scores is accelerating or decelerating. The bigger, The bigger the will become larger, eventually leading to each student The smaller it is, the lower the similarity of the recommended test questions obtained by even low-scoring students will be, thus avoiding the problem of students' rigid thinking or lack of thinking caused by improper application of learning resources.

[0076] On the other hand, the second part of the learning resource benchmark score B(t) is composed of P. The present application can adjust the benchmark score according to the position of the mean of the student's score relative to the set high score and low score boundaries. If the overall score is too low, the system can reduce the difficulty of subsequent tests. If the overall score is too high, the system can increase the difficulty of subsequent tests to make the score distribution more even.

[0077] S104: adjusting the benchmark test question group of the subsequent learning node according to the use of the generated learning resources by each student to test the students at the subsequent learning nodes.

[0078] Specifically, adjusting the benchmark test questions of future learning nodes according to the use of the generated learning resources by each student includes the following steps:

[0079] In the next learning node, the scores of each student completing the recommended homework exercises are obtained;

[0080] Calculate the reasonable value of the difficulty setting of basic test questions :

[0081]

[0082] N is the total number of students, is the score of the i-th student completing the recommended homework exercises, is the average score of each student completing the recommended homework exercises, is the standard deviation of each student's scores on the recommended homework exercises;

[0083] when When the value is between [-0.5, 0.5], no adjustment is made to the benchmark test set;

[0084] when When <-0.5, add difficult questions to the baseline test set;

[0085] when >0.5, add easy questions to the benchmark test set.

[0086] Specifically, difficult or simple questions are determined based on the median ranking of students in their scores, and the difficulty benchmark of the median student is obtained as the median difficulty benchmark. Questions in the question bank that are lower than the median difficulty benchmark are simple questions, and questions that are higher than the median difficulty benchmark are difficult questions. The method of adding can be that the teacher replaces a part of the benchmark test questions that are more difficult than the median difficulty benchmark with the simple questions, or replaces a part of the benchmark test questions that are less difficult than the median difficulty benchmark with the difficult questions.

[0087] In summary, the learning resource recommendation system based on big data provided by the present application has the beneficial effect that the learning resource recommendation system based on big data provided by the present application first uses benchmark test questions to conduct an initial test on students to obtain the score distribution of each learning node. Subsequently, based on the specific performance of each student, the system will generate personalized learning resources for future learning nodes, which are finely adjusted in similarity and difficulty to match the learning ability of each student, thereby ensuring that each student can learn on the basis of their current ability.

[0088] In addition, the system also adjusts the difficulty and similarity of subsequent students' learning resources according to the speed of change in the distribution of students' grades after using the learning resources. This dynamic adjustment mechanism not only takes into account the uneven changes in students' learning abilities, but also overcomes the problem of students' rigid thinking or lack of thinking that may be caused by improper application of learning resources. By analyzing a large amount of students' performance data, continuously tracking students' performance and adjusting teaching resources and methods according to actual conditions, the system maximizes the advantages of big data analysis and improves teaching effectiveness.

[0089] At the same time, it should be realized that the benchmark test set composed of benchmark test questions is often designed based on the teacher's experience. Therefore, for different student groups, this benchmark test set may be too simple or too difficult, and the design of the benchmark test set itself is flawed and cannot reasonably test the level of the student group. This system also adjusts the benchmark test set of the current learning node according to the actual situation of the student group based on the situation of students using learning resources at the current learning node, avoiding the technical problem that the level of the student group cannot be reasonably tested due to the initial design of the benchmark test set itself being too simple or too difficult.

[0090] Therefore, the technical solution provided by this application effectively improves the pertinence of teaching resources, thereby improving teaching effectiveness.

[0091] In a second aspect, the present application provides a learning resource recommendation system based on big data, wherein the big data-based auxiliary learning system includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement a learning resource recommendation method based on big data as described in any of the above embodiments.

[0092] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. In addition, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "including one..." do not exclude the existence of other identical elements in the process, method, article or device including the elements. Moreover, in the description of the embodiments of the present application, unless otherwise specified, " / " means or, for example, A / B can represent A or B; "and / or" in this article is only a kind of association relationship describing the associated objects, indicating that there can be three relationships, for example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. Furthermore, in the description of the embodiments of the present application, “plurality” refers to two or more than two.

[0093] The above description is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments described herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A learning resource recommendation method based on big data, characterized in that: The learning resource recommendation method based on big data includes: Divide students' learning courses into multiple learning nodes; After the students have completed the study of the first learning node, each student is initially tested using a benchmark test group consisting of multiple benchmark test questions to obtain the score distribution of the current learning node; The following steps are executed in a loop so that the distribution of students' scores in the tests of subsequent learning nodes gradually approaches the predetermined distribution: Generate learning resources for each student's future learning nodes based on their position in the current score distribution; According to each student's use of the generated learning resources, the benchmark test question group of the subsequent learning node is adjusted to test the student at the subsequent learning node; Generate learning resources for each student's future learning node according to each student's position in the current score distribution, including the following steps: Obtain each student's learning resource recommendation score based on the learning resource benchmark score and each student's score distribution; Generate a similarity benchmark and a difficulty benchmark for after-class exercises based on each student's learning resource recommendation score. In the next learning node, for each student, based on the student's similarity benchmark and difficulty benchmark, select from the question bank the questions that are closest to the student's similarity benchmark and difficulty benchmark compared to the benchmark questions, and recommend them to the student as after-class exercises. The learning resource benchmark score Obtained by the following steps: in: Wherein, k is the preset first adjustment coefficient, is the preset second adjustment coefficient, is the standard deviation of the current learning node score distribution, is the mean of the score distribution of the current learning node.

2. The learning resource recommendation method based on big data according to claim 1, characterized in that: Obtain each student's recommended learning resource score based on the learning resource benchmark score and each student's score distribution , including the following steps: in, is the preset third adjustment coefficient, is the student's score in this learning node. is the mean of students’ scores at this learning node.

3. The learning resource recommendation method based on big data according to claim 2 is characterized in that: The similarity benchmark and difficulty benchmark for after-class exercise recommendation are generated based on each student's learning resource recommendation score, including the following steps: Wherein, d is the preset fourth adjustment coefficient, and l is the preset fifth adjustment coefficient.

4. The learning resource recommendation method based on big data according to claim 1, characterized in that: Adjusting the benchmark test questions of future learning nodes according to each student's use of the generated learning resources includes the following steps: In the next learning node, the scores of each student completing the recommended homework exercises are obtained; Calculate the reasonable value of the difficulty setting of basic test questions : N is the total number of students, is the score of the i-th student completing the recommended homework exercises, is the average score of each student completing the recommended homework exercises, is the standard deviation of each student's scores on the recommended homework exercises; when When the value is between [-0.5, 0.5], no adjustment is made to the benchmark test set; when When <-0.5, add difficult questions to the baseline test set; when >0.5, add easy questions to the benchmark test set.

5. The learning resource recommendation system based on big data is characterized by: The big data-based auxiliary learning system includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the big data-based learning resource recommendation method as described in any one of claims 1-4.

Images