A course selection recommendation method and device based on student intention prediction

Abstract

The application discloses a course selection recommendation method and device based on student intention prediction, and comprises the following steps: obtaining objective characteristic information of a student; performing prediction calculation on the objective characteristic information according to a preset random forest model to obtain an objective prediction result representing student intention; obtaining subjective characteristic information of the student and performing weighted processing on the objective characteristic information and the subjective characteristic information to obtain actual course selection tendency information of the student; and traversing a course information database of a corresponding academic year according to the actual course selection tendency information, outputting course selection information meeting the student intention and recommending the course selection information to the student. The application predicts the course selection intention of the student through the objective characteristic information and the subjective characteristic information of the student, obtains the course selection information meeting the student intention, can avoid the student from blindly selecting some courses that the student is not interested in, reduces the waste of teaching resources, reduces the process time of the student in searching for a favorite course from numerous courses during course selection, and greatly improves the course selection efficiency of the student.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent decision-making, in particular to a course recommendation method and device based on student intention prediction. BACKGROUND

[0002] Universities generally follow the course selection system, which allows students to have some freedom of choice in the courses offered by the school. In each academic year, students need to take a certain number of courses to meet the school's credit rules. Each course has its corresponding fixed credits. Students generally choose courses that interest them or meet the school's credit requirements during the course selection process. By considering various factors, students choose to take appropriate courses. The subject field of university disciplines is constantly expanding, and the courses offered have also increased. The increase in courses brings students more choices, but also easily leads to students blindly choosing courses that they are not interested in, which can easily lead to the occupation of teaching resources.

[0003] In addition, the most headache problem for college students at the beginning of a new semester is course selection. A large number of students concentrate in a time period to select courses, making the course selection system run slowly and extremely inefficient. In this case, students need to choose their favorite courses from a large number of courses after entering the course selection system. In the process of finding their favorite courses, the course selection system is extremely prone to collapse, and students cannot choose the courses they want. If students cannot choose the courses they are interested in, it will often lead to not taking courses seriously, which greatly affects the enthusiasm of students. SUMMARY

[0004] The purpose of the present application is to provide a course recommendation method and device based on student intention prediction to solve one or more technical problems in the prior art and at least provide a beneficial choice or create conditions.

[0005] The solution to the technical problem of the present application is to provide a course recommendation method based on student intention prediction, which comprises the following steps:

[0006] Step 1, obtaining the objective information and professional information of the student, preprocessing the objective information of the student to obtain the objective feature information of the student, and grouping the objective feature information of the student according to a preset subject classification dimension;

[0007] The objective information includes the subject and the corresponding GPA information of the student in the history academic year.

[0008] Step 2, performing prediction calculation on the objective feature information according to a preset random forest model to obtain an objective prediction result representing the student's intention;

[0009] Step 3, obtaining the subjective feature information of the student;

[0010] The subjective characteristic information includes course selection intention information, postgraduate examination intention information and employment intention information.

[0011] Step 4, weighting the objective prediction result and the subjective characteristic information, and outputting the weighted result as actual course selection tendency information of the student;

[0012] Step 5, determining the academic year of the student according to the professional information, traversing the course information database of the corresponding academic year according to the actual course selection tendency information, obtaining target recommended courses and candidate recommended courses meeting the intention of the student, outputting course selection information meeting the intention of the student, and recommending the course selection information meeting the intention of the student to the student.

[0013] As a further improvement of the above technical solution, the preprocessing of the objective information of the student includes:

[0014] Step 1.1, finding missing data of the objective information, calculating the proportion of the missing data in the total data amount of the objective information, deleting the corresponding missing data if the proportion is greater than 70%, and filling in the average value of the grade point information of all subjects if the proportion is not greater than 70%;

[0015] Step 1.2, feature selection is performed on the objective information, the Pearson correlation coefficient is calculated, and the objective information with a correlation coefficient greater than 8% is screened out to generate the objective characteristic information.

[0016] As a further improvement of the above technical solution, the grouping processing of the objective characteristic information of the student according to the preset subject classification dimension includes:

[0017] Each subject of the historical academic year is divided into its corresponding subject field to obtain a plurality of groups.

[0018] As a further improvement of the above technical solution, the objective prediction result representing the intention of the student is obtained by performing prediction calculation on the objective characteristic information according to the preset random forest model, including:

[0019] Step 2.1, matching the corresponding CART tree for each group in the preset random forest model according to the preset subject classification dimension;

[0020] Step 2.2, performing prediction calculation on the objective characteristic information in the corresponding group by using the CART tree, and generating the student intention factor corresponding to the group;

[0021] Step 2.3, calculating the interest degree of the student for the subject field corresponding to the group according to the student intention factor corresponding to the group, and outputting the objective prediction result representing the intention of the student.

[0022] As a further improvement of the above technical solution, the acquisition process of the preset random forest model is:

[0023] Step 2.1.1, obtaining sample data, wherein the sample data is the subject and the corresponding GPA information of the students in each academic year, and the sample data is grouped according to the preset subject classification dimension to form a sample data set;

[0024] Step 2.1.2, training the sample training set using the random forest algorithm to obtain a plurality of CART trees, and outputting the random forest model when the number of CART trees meets the preset number;

[0025] The implementation process of the CART tree in step 2.1.2 is:

[0026] a. Select a feature i from the N features of the sample data set as the first node of the binary tree;

[0027] b. According to the feature i, select a split point k, and divide the samples with feature i less than the split point k into one class, and divide the samples with feature i greater than the split point k into another class;

[0028] c. Repeat steps a to b until the construction of the CART tree is completed.

[0029] As a further improvement of the above technical solution, the student's academic year is determined according to the professional information, the actual course selection tendency information is used to traverse the course information library, the target recommended course and the candidate recommended course that meet the student's intention are obtained, the course selection information that meets the student's intention is output, and the course selection information that meets the student's intention is recommended to the student, including:

[0030] Step 5.1, determining the student's academic year according to the professional information and the course information library of the corresponding academic year;

[0031] Wherein, the course information library includes the courses available for reading in the corresponding academic year and the corresponding credits;

[0032] Step 5.2, searching for the target recommended course in the plurality of categories of courses according to the actual course selection tendency information;

[0033] Step 5.3, determining whether the target recommended course has a free seat; if the target recommended course has a free seat, go to step 5.3.1;

[0034] Step 5.3.1, when the target recommended course has a free seat, output the target recommended course to the output stack, and execute step 5.4 after outputting to the output stack;

[0035] If the target recommended course has no vacancy, go to step 5.3.2;

[0036] Step 5.3.2, obtain the history courses selected by the student in the history academic year, and find the category of the history courses corresponding to the history courses;

[0037] Among them, the history course is defined as: the course selected by the student in the history academic year and the corresponding GPA is above the preset value;

[0038] Step 5.3.3, according to the actual course selection tendency information, find the candidate recommended course in the category of the history course, output the candidate recommended course to the output stack, and execute step 5.4 after outputting to the output stack;

[0039] Step 5.4, calculate the total credit value of the target recommended course and the candidate recommended course in the output stack;

[0040] Step 5.5, judge whether the total credit value is greater than or equal to the preset expected credit value; if yes, execute step 5.6;

[0041] Step 5.6, sort the target recommended course and the candidate recommended course in the output stack according to the credit from large to small, obtain the course selection information conforming to the student's intention, and recommend the course selection information conforming to the student's intention to the student;

[0042] In step 5.5, if the total credit value is not greater than or equal to the preset expected credit value, steps 5.2 to 5.5 are executed in a loop.

[0043] A course selection recommendation device based on student intention prediction, comprising:

[0044] An acquisition unit is configured to acquire objective information, professional information and subjective characteristic information of a student;

[0045] A prediction unit is configured to perform prediction calculation on the objective characteristic information according to a preset random forest model to obtain an objective prediction result representing the student's intention;

[0046] A weighting processing unit is configured to perform weighting processing on the objective prediction result and the subjective characteristic information, and output the weighting result as actual course selection tendency information of the student;

[0047] A search unit is configured to search for a target recommended course in a plurality of categories of courses according to the actual course selection tendency information;

[0048] A processing unit is configured to determine the academic year of the student according to the professional information, traverse a course information database of the corresponding academic year according to the actual course selection tendency information, and obtain a target recommended course and a candidate recommended course conforming to the student's intention;

[0049] The sorting unit is configured to sort the target recommended course and the candidate recommended course in the output stack in descending order of credits, and output the course selection information meeting the student's intention.

[0050] The recommending unit is configured to recommend the course selection information meeting the student's intention to the student.

[0051] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the course selection recommendation method based on the student's intention prediction when executing the computer program.

[0052] A computer readable storage medium stores a computer program, and the computer program implements the steps of the course selection recommendation method based on the student's intention prediction when executed by a processor.

[0053] The application has the beneficial effects that a course selection recommendation method and device based on a student's intention prediction are provided, which include: obtaining objective information and professional information of a student, preprocessing the objective information of the student to obtain objective feature information of the student, and performing grouping processing on the objective feature information of the student according to a preset subject classification dimension; performing prediction calculation on the objective feature information according to a preset random forest model to obtain an objective prediction result representing the student's intention; obtaining subjective feature information of the student; performing weighting processing on the objective prediction result and the subjective feature information to obtain actual course selection tendency information of the student; determining a school year in which the student is located according to the professional information, traversing a course information database of a corresponding school year according to the actual course selection tendency information to obtain a target recommended course and a candidate recommended course meeting the student's intention, outputting course selection information meeting the student's intention, and recommending the course selection information meeting the student's intention to the student. The application predicts the student's course selection intention through the objective feature information and the subjective feature information of the student, obtains the actual course selection tendency information of the student, and selects a recommended course according to the actual course selection tendency information, which can avoid the student from blindly selecting some courses that the student is not interested in, reduces the waste of teaching resources, predicts a recommended course that the student is interested in according to the student's intention, reduces the process time of the student in searching for a desired course in a large number of courses when selecting a course, and greatly improves the efficiency of the student's course selection. BRIEF DESCRIPTION OF DRAWINGS

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly described. Obviously, the described drawings are only a part of the embodiments of the present application, and those skilled in the art can obtain other design schemes and drawings according to these drawings without creative labor.

[0055] Figure 1 is a method flowchart of a course selection recommendation method based on student intention prediction according to an embodiment of the present application;

[0056] Figure 2 is a method flowchart of obtaining an objective prediction result representing student intention according to a course selection recommendation method based on student intention prediction according to an embodiment of the present application;

[0057] Figure 3 is a method flowchart of obtaining a preset random forest model according to a course selection recommendation method based on student intention prediction according to an embodiment of the present application;

[0058] Figure 4 is a method flowchart of obtaining course selection information meeting student intention according to a course selection recommendation method based on student intention prediction according to an embodiment of the present application. DETAILED DESCRIPTION

[0059] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0060] It should be noted that although the functional modules are divided in the system schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a manner different from the module division in the system or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and above-described drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence.

[0061] In order to illustrate the technical solutions described in the present application, the following specific embodiments are described and introduced.

[0062] The present application aims to provide a course selection recommendation method based on student intention prediction. Based on the multidimensionality of student course selection intention and the multidimensionality of course information, the student's interest in the subject category can be accurately captured by combining the student's grade point information in the subject in the historical academic year, the accuracy of student course selection intention prediction is improved, the student is prevented from blindly selecting some courses that he is not interested in, and the efficiency of student course selection is greatly improved.

[0063] Referring to Figure 1 The present application discloses a course selection recommendation method based on student intention prediction, which comprises the following steps:

[0064] Step 1, obtaining the objective information and professional information of the student, preprocessing the objective information of the student to obtain the objective feature information of the student, and grouping the objective feature information of the student according to the preset subject classification dimension;

[0065] The objective information includes subject information and corresponding GPA information of the student in a historical academic year.

[0066] Step 2, performing prediction calculation on the objective characteristic information according to a preset random forest model to obtain an objective prediction result representing the student's intention;

[0067] Step 3, obtaining subjective characteristic information of the student;

[0068] The subjective characteristic information includes course selection intention information, postgraduate examination intention information and employment intention information.

[0069] Step 4, performing weighted processing on the objective prediction result and the subjective characteristic information, and outputting a weighted result as actual course selection tendency information of the student;

[0070] Step 5, determining an academic year of the student according to the major information, traversing a course information database of the corresponding academic year according to the actual course selection tendency information, obtaining target recommended courses and candidate recommended courses meeting the student's intention, outputting course selection information meeting the student's intention, and recommending the course selection information meeting the student's intention to the student.

[0071] In step 1, objective information and major information of the student are obtained, the objective information of the student is preprocessed to obtain objective characteristic information of the student, and the objective characteristic information of the student is grouped according to a preset subject classification dimension, wherein the objective information includes subject information and corresponding GPA information of the student in a historical academic year.

[0072] Further, the grouping of the objective characteristic information of the student according to the preset subject classification dimension includes: dividing each subject in the historical academic year into a corresponding subject field to obtain a plurality of groups.

[0073] Each course has a corresponding fixed credit, and the student generally selects a course of interest or a course meeting the school credit requirement in the course selection process, and selects a suitable course by comprehensively considering various factors. Each subject studied by the student belongs to a corresponding subject category, and the GPA information of the subject of the student in the historical academic year can reflect the interest degree of the student in the subject field corresponding to the subject. The lower the GPA information of a subject, the lower the interest degree of the student in the subject field to which the subject belongs. Therefore, the preprocessed objective characteristic information is grouped according to the subject field to which each subject belongs to obtain a plurality of groups, and each group corresponds to each subject field.

[0074] In this embodiment, the major information of the student is defined as the major to which the student belongs.

[0075] Further, the pre-processing of the objective information of the student to obtain the objective feature information includes the following steps:

[0076] Step 1.1, find the missing data in the objective information, calculate the proportion of the missing data in the total data amount of the objective information, if the proportion is greater than 70%, delete the corresponding missing data, if the proportion is not greater than 70%, fill in the average value of the objective information;

[0077] Step 1.2, feature selection is performed on the objective information, Pearson correlation coefficient is calculated, and objective information with a correlation coefficient greater than 8% is screened out to generate objective feature information.

[0078] It should be noted that the present application calculates the similarity between the objective information and the student's course selection intention by Pearson coefficient, and screens out objective information with a correlation coefficient greater than 8% to obtain objective feature information for predicting the student's course selection intention.

[0079] In step 2, the objective feature information is predicted and calculated according to the preset random forest model to obtain an objective prediction result representing the student's intention.

[0080] It can be understood that the preset random forest model includes a plurality of CART trees, i.e. classification and regression trees, which can solve classification and regression problems. When the dependent variable of the input data set is continuous numerical value, the classification and regression tree is a regression tree, and the mean value of the leaf node observation is used as the prediction value. When the dependent variable of the input data set is discrete numerical value, the classification and regression tree is a classification tree, which is used to solve the classification problem.

[0081] Further, please refer to Figure 2 , the objective prediction result representing the student's intention is obtained by predicting and calculating the objective feature information according to the preset random forest model, which includes:

[0082] Step 2.1, according to the preset subject classification dimension, a corresponding CART tree is matched for each group in the preset random forest model;

[0083] Step 2.2, the CART tree is used to predict and calculate the objective feature information in the corresponding group, and a student intention factor corresponding to the group is generated;

[0084] Step 2.3, according to the student intention factor corresponding to the group, the interest degree of the student for the subject field corresponding to the group is calculated respectively, and the objective prediction result representing the student's intention is output.

[0085] In the present application, the objective feature information obtained in step 1 is taken as the input of the random forest model, and the output of the random forest model is the objective prediction result representing the student intention. For each group, a corresponding classification and regression tree is configured, and the objective feature information in each group is input into the corresponding CART tree for prediction calculation to analyze the student's course selection tendency for each group, thereby obtaining the objective prediction result representing the student intention. Each CART tree generates a corresponding student intention factor after prediction calculation, and the above objective prediction result representing the student intention is an integrated data set of all student intention factors, which predicts the student's course selection intention from different dimensions.

[0086] Further, please refer to Figure 3 The acquisition process of the preset random forest model is as follows:

[0087] Step 2.1.1, obtaining sample data, wherein the sample data is the subject and the corresponding GPA information of the students in each academic year, and the sample data is grouped according to the preset subject classification dimension to form a sample data set.

[0088] It should be noted that the sample data set includes multiple groups of sample data, each group of sample data includes multiple subjects belonging to the same subject category and the corresponding GPA information of the subjects, and these subjects are the subjects read by the students in each academic year.

[0089] Step 2.1.2, training the sample training set by using the random forest algorithm to obtain multiple CART trees, and outputting the random forest model when the number of CART trees meets the preset number.

[0090] It can be understood that after obtaining the sample data set, each group of sample data is trained by using the random forest algorithm to obtain the corresponding classification and regression tree, and the random forest model is output after integrating all the classification and regression trees.

[0091] Further, the implementation process of the CART tree in step 2.1.2 is as follows:

[0092] a. Select a feature i from the N features of the sample data set as the first node of the binary tree;

[0093] b. According to the feature i, select a split point k, and divide the samples with the value of the feature i less than the split point k into one class, and divide the samples with the value of the feature i greater than the split point k into another class;

[0094] Wherein, the split point k corresponding to the feature i is obtained by the mean square deviation calculation formula;

[0095] c. Repeat steps a to b until the construction of the CART tree is completed.

[0096] In step 3, the subjective characteristic information of students is obtained, including course selection intention information, postgraduate entrance examination intention information, and employment intention information.

[0097] It should be noted that the subjective characteristic information is obtained from three dimensions: course selection intention, postgraduate entrance examination intention, and employment intention. Course selection intention directly reflects a student's level of interest in the courses they intend to study; postgraduate entrance examination intention is usually a single academic field, reflecting a student's level of interest in that field; employment intention is usually a specific work field, which can be mapped to an academic field, similarly reflecting a student's level of interest in that field. Therefore, this application obtains students' subjective characteristic information to compensate for errors in the objective prediction results representing students' intentions, thereby improving the accuracy of predicting students' course selection intentions.

[0098] In step 4, the objective prediction results and the subjective feature information are weighted and processed, and the weighted results are output as the students' actual course selection preference information.

[0099] In step 5, please refer to Figure 4 Based on the student's major information, the academic year is determined. Then, based on the actual course selection preference information, the course information database for the corresponding academic year is traversed to obtain target recommended courses and candidate recommended courses that match the student's preferences. Finally, course selection information matching the student's preferences is output and recommended to the student. Specifically, this includes:

[0100] Step 5.1: Determine the student's academic year and the corresponding course information database based on the major information. The course information database includes the courses that can be taken in the corresponding academic year and their corresponding credits.

[0101] Understandably, depending on the specific requirements of each university, each major has its corresponding courses, and the required courses vary from academic year to academic year. First, based on the student's major information, the academic year and the corresponding course information database are determined. This database includes the courses available for that academic year and their corresponding credits.

[0102] Step 5.2: Based on the actual course selection preference information, search for target recommended courses in multiple categories of courses;

[0103] Step 5.3: Determine whether the target recommended courses have available spots.

[0104] In this embodiment, the target recommended course that meets the student's course selection intention is found in the course information library according to the actual course selection tendency information. It should be noted that each course usually has a certain number of enrollment quotas, so it is necessary to confirm whether the target recommended course found has a free quota, and when the target recommended course has a free quota, the target recommended course is selected for the student; if the target recommended course does not have a free quota, other recommended courses are selected for the student.

[0105] If the target recommended course has a free quota, go to step 5.3.1;

[0106] Step 5.3.1, when the target recommended course has a free quota, the target recommended course is output to the to-be-output stack, and after output to the to-be-output stack, step 5.4 is executed.

[0107] If the target recommended course does not have a free quota, go to step 5.3.2;

[0108] Step 5.3.2, the historical courses selected by the student in the historical academic years are obtained, and the categories of the historical courses corresponding to the historical courses are found;

[0109] Among them, the historical course is defined as: the course selected by the student in the historical academic years, and the grade point corresponding to the course is above the preset value;

[0110] Step 5.3.3, according to the actual course selection tendency information, find the candidate recommended course in the category of the historical course, output the candidate recommended course to the to-be-output stack, and after output to the to-be-output stack, execute step 5.4.

[0111] It should be noted that when the target recommended course does not have a free quota, the recommended course similar to the historical course is selected for the student. The selection of the historical course in this application is subject to a grade point condition, because the higher the grade point of the course, the higher the degree of interest of the student in the subject field to which the course belongs, therefore, the historical course selected must have a grade point information reaching a preset value, so as to ensure that the student is interested in the subject field corresponding to the historical course, and the candidate recommended course obtained can better meet the student's learning interest.

[0112] Step 5.4, calculate the total credit value of the target recommended course and the candidate recommended course in the to-be-output stack.

[0113] In this embodiment, the course information library obtained in step 5.1 includes the courses available for enrollment in the corresponding academic year and the credits corresponding to the courses, and the total credit value of the target recommended course and the candidate recommended course is calculated to determine whether the total credit value of the selected recommended course meets the student's enrollment requirement in the current academic year.

[0114] Step 5.5, judging whether the total score value is greater than or equal to the preset expected score value; if yes, executing step 5.6; if the total score value is not greater than or equal to the preset expected score value, executing steps 5.2 to 5.5 cyclically.

[0115] Step 5.6, sorting the target recommended courses and the candidate recommended courses in the output stack according to the score from large to small, obtaining the course selection information meeting the student's intention, and recommending the course selection information meeting the student's intention to the student.

[0116] In the present application, after obtaining the target recommended courses and the candidate recommended courses, the target recommended courses and the candidate recommended courses meeting the student's intention can also be sorted according to the score value from large to small, so as to recommend the courses to the student in the logic of the score from large to small, which helps the student to select courses.

[0117] The present application also discloses a course selection recommendation method based on student intention prediction.

[0118] The obtaining unit is configured to obtain objective information, professional information and subjective characteristic information of the student.

[0119] The prediction unit is configured to perform prediction calculation on the objective characteristic information according to a preset random forest model, to obtain an objective prediction result representing the student's intention.

[0120] The weighting processing unit is configured to perform weighting processing on the objective prediction result and the subjective characteristic information, and output a weighting result as actual course selection tendency information of the student.

[0121] The searching unit is configured to search for target recommended courses from courses of multiple categories according to the actual course selection tendency information.

[0122] The processing unit is configured to determine a school year in which the student is located according to the professional information, and traverse a course information database of the corresponding school year according to the actual course selection tendency information, to obtain target recommended courses and candidate recommended courses meeting the student's intention.

[0123] The sorting unit is configured to sort the target recommended courses and the candidate recommended courses in the output stack according to the score from large to small, and output course selection information meeting the student's intention.

[0124] The recommending unit is configured to recommend the course selection information meeting the student's intention to the student.

[0125] The present application also discloses an electronic device, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, and the processor implements the steps of the course selection recommendation method based on student intention prediction when executing the computer program.

[0126] The application further discloses a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize steps of the course recommendation method based on student intention prediction.

[0127] The above describes the preferred embodiments of the application in detail, but the application is not limited to the embodiments, and those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the application, and the equivalent modifications or replacements are all included in the scope defined by the claims of the application.

Claims

1. A course selection recommendation method based on student intention prediction, characterized by, The method comprises the following steps: Step 1, obtaining objective information and professional information of students, preprocessing the objective information of students to obtain objective characteristic information of students, and grouping the objective characteristic information of students according to a preset subject classification dimension; Wherein, the objective information includes the subject and the corresponding GPA information of the student in the history year; Step 2, performing prediction calculation on the objective characteristic information according to a preset random forest model to obtain an objective prediction result representing the intention of the student; Step 3, obtaining the subjective characteristic information of the student; Wherein, the subjective characteristic information includes the course selection intention information, the postgraduate examination intention information and the employment intention information; Step 4, performing weighted processing on the objective prediction result and the subjective characteristic information, and outputting the weighted result as the actual course selection tendency information of the student; Step 5, determining the school year of the student according to the professional information, traversing the course information database of the corresponding school year according to the actual course selection tendency information, obtaining the target recommended course and the candidate recommended course meeting the intention of the student, outputting the course selection information meeting the intention of the student, and recommending the course selection information meeting the intention of the student to the student; Wherein, the grouping processing of the objective characteristic information of the student according to the preset subject classification dimension comprises: Dividing each subject of the history year into the corresponding subject field to obtain a plurality of groups; Wherein, the prediction calculation on the objective characteristic information according to the preset random forest model to obtain the objective prediction result representing the intention of the student comprises: Step 2.1, matching the corresponding CART tree for each group in the preset random forest model according to the preset subject classification dimension; Step 2.2, performing prediction calculation on the objective characteristic information in the corresponding group by using the CART tree, and generating the student intention factor corresponding to the group; Step 2.3, calculating the interest degree of the student for the subject field corresponding to the group according to the student intention factor corresponding to the group, and outputting the objective prediction result representing the intention of the student. 2.The course recommendation method based on student intention prediction of claim 1, wherein, The preprocessing of the objective information of the student comprises: Step 1.1, finding the missing data of the objective information, calculating the proportion of the missing data in the total data amount of the objective information, deleting the corresponding missing data if the proportion is greater than 70%, and filling the missing data with the average value of the GPA information of all subjects if the proportion is not greater than 70%; Step 1.2, performing feature selection on the objective information, calculating the Pearson correlation coefficient, and screening out the objective information with a correlation coefficient greater than 8% to generate the objective characteristic information. 3.The course recommendation method based on student intention prediction of claim 1, wherein, The acquisition process of the preset random forest model comprises: Step 2.1.1, obtaining sample data, wherein the sample data is the subject and the corresponding GPA information of the students in each school year, grouping the sample data according to the preset subject classification dimension to form a sample data set; Step 2.1.2, training the sample training set by using the random forest algorithm to obtain a plurality of CART trees, and outputting the random forest model when the number of CART trees meets the preset number; The implementation process of the CART tree in step 2.1.2 comprises: a. select a feature i from N features of the sample dataset as the first node of the binary tree; b. select a split point k according to the feature i, divide the samples with values less than the split point k into one class, and divide the samples with values greater than the split point k into another class; c. repeat steps a to b until the construction of the CART tree is completed. 4.The course recommendation method based on student intention prediction of claim 1, wherein, The method comprises the following steps of: Step 5.1, determining the academic year of the student according to the professional information and the course information library of the corresponding academic year; The course information library comprises courses available for selection in the corresponding academic year and corresponding credits; Step 5.2, searching for a target recommended course in the multiple categories of courses according to the actual course selection tendency information; Step 5.3, judging whether the target recommended course has a vacant seat; if the target recommended course has a vacant seat, proceed to step 5.3.1; Step 5.3.1, when the target recommended course has a vacant seat, output the target recommended course to the to-be-output stack, and then execute step 5.4; If the target recommended course does not have a vacant seat, proceed to step 5.3.2; Step 5.3.2, obtaining the historical courses selected by the student in the historical academic years, and searching for the categories of the historical courses corresponding to the historical courses; The historical course is defined as a course selected by the student in the historical academic years and having a performance score above a preset value; Step 5.3.3, searching for a candidate recommended course in the category of the historical courses according to the actual course selection tendency information, outputting the candidate recommended course to the to-be-output stack, and then executing step 5.4; Step 5.4, calculating the total credit value of the target recommended course and the candidate recommended course in the to-be-output stack; Step 5.5, judging whether the total credit value is greater than or equal to a preset expected credit value; if yes, execute step 5.6; Step 5.6, sorting the target recommended course and the candidate recommended course in the to-be-output stack in descending order of credits to obtain the course selection information meeting the intention of the student, and recommending the course selection information meeting the intention of the student to the student; In step 5.5, if the total credit value is not greater than or equal to the preset expected credit value, steps 5.2 to 5.5 are executed cyclically.

5. A course recommendation device based on student intention prediction, characterized by, The method comprises the following steps of: An acquisition unit is configured to acquire objective information, professional information and subjective characteristic information of a student, preprocess the objective information of the student to obtain objective characteristic information of the student, and group process the objective characteristic information of the student according to a preset subject classification dimension; A prediction unit is configured to perform prediction calculation on the objective characteristic information according to a preset random forest model to obtain an objective prediction result representing the intention of the student; A weighting processing unit is configured to perform weighting processing on the objective prediction result and the subjective characteristic information, and output a weighting result as actual course selection tendency information of the student. The searching unit is configured to search for target recommended courses from the courses in multiple categories according to the actual course selection tendency information; The processing unit is configured to determine the academic year of the student according to the professional information, and traverse the course information database of the corresponding academic year according to the actual course selection tendency information to obtain target recommended courses and candidate recommended courses that meet the intention of the student; The sorting unit is configured to sort the target recommended courses and the candidate recommended courses in the output stack in descending order of credits, and output the course selection information that meets the intention of the student; The recommending unit is configured to recommend the course selection information that meets the intention of the student to the student; The grouping processing of the objective characteristic information of the student according to the preset subject classification dimension comprises: Each subject of the historical academic year is divided into a corresponding subject field to obtain multiple groups; The objective characteristic information is predicted and calculated according to the preset random forest model to obtain an objective prediction result representing the intention of the student, which comprises: A corresponding CART tree is matched for each group in the preset random forest model according to the preset subject classification dimension; The objective characteristic information in the corresponding group is predicted and calculated by using the CART tree, and a student intention factor corresponding to the group is generated; The interest degree of the student in the subject field corresponding to the group is calculated according to the student intention factor corresponding to the group, and an objective prediction result representing the intention of the student is output.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the computer program to realize the steps of the course selection recommendation method based on the intention prediction of the student according to any one of claims 1-4.

7. A computer-readable storage medium storing a computer program, wherein the computer program comprises the following steps of: The computer program is executed by the processor to realize the steps of the course selection recommendation method based on the intention prediction of the student according to any one of claims 1-4.

Images