CN113935869A - Student subjective and objective factor combined score prediction method and system - Google Patents
Student subjective and objective factor combined score prediction method and system Download PDFInfo
- Publication number
- CN113935869A CN113935869A CN202111050024.4A CN202111050024A CN113935869A CN 113935869 A CN113935869 A CN 113935869A CN 202111050024 A CN202111050024 A CN 202111050024A CN 113935869 A CN113935869 A CN 113935869A
- Authority
- CN
- China
- Prior art keywords
- course
- student
- students
- attribute
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 239000011159 matrix material Substances 0.000 claims abstract description 29
- 230000006399 behavior Effects 0.000 claims abstract description 14
- 238000013528 artificial neural network Methods 0.000 claims abstract description 10
- 238000004140 cleaning Methods 0.000 claims abstract description 8
- 238000004364 calculation method Methods 0.000 claims description 26
- 239000013598 vector Substances 0.000 claims description 22
- 238000012545 processing Methods 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 4
- 238000012360 testing method Methods 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 2
- 230000006870 function Effects 0.000 claims description 2
- 238000013507 mapping Methods 0.000 claims description 2
- 238000007418 data mining Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 230000036651 mood Effects 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/20—Education
- G06Q50/205—Education administration or guidance
- G06Q50/2053—Education institution selection, admissions, or financial aid
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Educational Technology (AREA)
- Economics (AREA)
- Educational Administration (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a student subjective and objective factor combined score prediction method and system. Firstly, constructing a course similarity matrix by using course information in a campus; then calculating the objective scores of the courses to be predicted by the students by using the course similarity matrix and the historical course scores of the students; secondly, collecting daily behavior data and learning data of students, cleaning and repairing the data, calculating the weight of each student attribute through a neural network, and multiplying the weight and the attribute value to obtain the weighted attribute of the students; and finally, taking the weighted attribute and objective result of the student as input, and predicting through a neural network to obtain the final predicted result of the student. The invention provides a score prediction method combining subjective and objective factors of students, which can predict the score of each stage of the students, provide academic early warning for the students and facilitate the teachers to perform personalized guidance for the students.
Description
Technical Field
The invention relates to the field of machine learning technology and data mining, in particular to a student subjective and objective factor combined score prediction method.
Background
With the rapid development of the internet, the amount of information also grows exponentially, which is particularly obvious in the environment of colleges and universities, and the growth of students and education staff and the use of informatization tools make data in campuses become huge and redundant. Therefore, in order to manage and utilize these data, educational data mining has been generated.
As a hotspot in the field of educational data mining research, student performance prediction predicts the performance of students in future learning stages through analysis of course associations, student historical performance, and other background data. In a university environment, without strict time limitation and supervision from teachers, students are easy to form loose learning moods, and the failure of examination at the end of the period is a more serious phenomenon. Therefore, avoiding student hanging has become an urgent need in high schools today. The student score not only reflects the mastery degree of the students on the knowledge, but also reflects the teaching level of teachers. Therefore, the learning attitude of the students can be corrected by predicting the scores of the students in advance, and the results are fed back to teachers to improve the teaching method, so that the school managers can conveniently make decisions which are helpful for the development of the students. Therefore, student performance prediction is becoming one of the important research directions in the smart campus.
The formation of student achievements depends on various factors which are mainly divided into two types, namely the subjective initiative of students, namely the learning condition and the enthusiasm of the students; the other is the objective factor, namely student's subject knowledge reserve and ease of course. In the field of student score prediction, a plurality of models are provided, such as common classification models, regression models and the like, characteristic data which can reflect learning conditions, enthusiasm, consumption conditions, psychological conditions and the like of students are collected, and then a machine learning algorithm is used for predicting the student scores, but the existing knowledge mastering degree of the students is not considered, the different influence degrees of different factors on different students are not considered, and a certain one-sidedness exists. In recent years, the method for converting the student score prediction problem into the recommendation system problem considers the knowledge mastering degree of students and predicts the student score by using the relation among courses, but the fact that the student score changes along with the effort degree of the students is ignored, and the method is comprehensive.
The invention describes a student subjective and objective factor combined score prediction method and system, which comprises the steps of firstly, constructing a course similarity matrix by using course information in a campus; then calculating the objective scores of the courses to be predicted by the students by using the course similarity matrix and the historical course scores of the students; secondly, collecting daily behavior data and learning data of students, cleaning and repairing the data, calculating the weight of each student attribute through a neural network, and multiplying the weight and the attribute value to obtain the weighted attribute of the students; and finally, taking the weighted attribute and objective result of the student as input, and predicting through a neural network to obtain the final predicted result of the student.
The invention designs a model combining subjective and objective factors to calculate the objective performance of students and the weighted attribute value of the students, combines the objective performance and the weighted attribute value of the students, and obtains the final predicted performance of the students through a neural network, thereby improving the accuracy of the predicted result.
Disclosure of Invention
The invention aims to overcome the one-sidedness of methods for modeling by collecting student data or predicting student scores by using correlations between student historical scores and courses and the like, and provides a score prediction system combining subjective and objective factors of students.
The technical scheme adopted for solving the technical problem comprises the following steps:
a student subjective and objective factor combined performance prediction system comprises: the system comprises an objective result prediction module, a student weighted attribute calculation module and a final result prediction module;
wherein:
the objective performance prediction module comprises a first data acquisition submodule, a course similarity matrix calculation module and an objective performance calculation module;
the first data acquisition submodule is used for acquiring relevant information of all courses in the smart campus and historical course scores of students;
the course similarity matrix calculation module is used for extracting characteristic words from the course information acquired by the first data acquisition submodule and calculating the similarity between courses to obtain a course similarity matrix;
the objective score calculating module is used for calculating the historical school scores of the students obtained by the first data acquisition submodule through a school similarity matrix to obtain the basic scores of the students;
the student weighted attribute calculation module comprises a second data acquisition submodule, a data processing module and a weighted attribute calculation module;
the second data acquisition submodule is used for acquiring daily behavior data and learning data of students, including basic information of the students, autonomous learning behavior records and the like;
the data processing module is used for cleaning the student data acquired by the second data acquisition submodule to obtain normalized data;
the weighted attribute calculation module is used for vectorizing the normalized data obtained by the data processing module and then bringing the normalized data into the weight calculation model to obtain the weight of the student attributes; and multiplying the calculated weight by the attribute value to obtain the weighted attribute value of the student.
And the final result prediction module is used for taking the weighted attributes and the basic results of the students as input and obtaining the final result prediction value of the to-be-predicted course of the current student through the result prediction model.
The invention also aims to provide a student subjective and objective factor combined achievement prediction method, which comprises the following steps:
step 1: acquiring relevant information of all courses in the smart campus and constructing a course similarity matrix;
step 2: acquiring historical school scores of students, and calculating through a school similarity matrix to obtain objective scores of the students;
and step 3: acquiring daily behavior data and learning data of students, and performing data cleaning on the original data to obtain normalized data serving as attributes of the students;
and 4, step 4: converting discrete variables in the attributes of the students into numerical types, vectorizing all data, bringing vectors into a neural network, constructing and training a weight calculation model, and obtaining the weights of the attributes of the students; multiplying the obtained student attribute weight and the corresponding attribute value to obtain a student weighted attribute value;
and 5: and (4) taking the weighted attribute value of the student obtained in the step (4) and the objective score of the student obtained in the step (2) as input, bringing the input into a neural network, and constructing and training a final score prediction model to obtain a predicted value of the final score of the student.
The invention has the beneficial effects that:
1. the invention provides a score prediction method combining subjective and objective factors of students, which can predict the score of each stage of the students, provide academic early warning for the students and facilitate the teachers to perform personalized guidance for the students.
2. The invention provides a weight calculation model, which is used for calculating attribute weights according to personal attribute data of students, reflecting the difference of different students on the same attribute and realizing the personalized prediction of student scores.
3. According to the method, the basic knowledge quantity of the students is calculated according to the historical scores of the students by constructing the course similarity matrix, and the difference of the knowledge quantities among the individual students is considered.
4. The invention adopts a student achievement prediction model combining subjective and objective factors of students, and the model is used for realizing accurate prediction of learning achievement.
Drawings
Fig. 1 is an overall frame diagram of the present invention.
FIG. 2 is a flow chart of objective performance prediction according to the present invention.
Fig. 3 is a flow chart of student attribute data processing according to the present invention.
Fig. 4 is a general flow chart of student achievement prediction according to the present invention.
Detailed Description
Embodiments of the present invention will be described in further detail below with reference to the accompanying drawings. The overall framework diagram of the invention is shown in fig. 1, and the specific steps are described as follows, wherein:
step 1: acquiring relevant information of all courses in the smart campus, and acquiring a feature word set of each course; then converting the feature word set of each course into a course vector of each course; finally, calculating the course similarity according to the course vector of each course to obtain a course similarity matrix;
in the step 1, relevant information of the school courses comprises course names, subject of the courses, the superior-subordinate relation between the courses and the course description;
the feature words of the course refer to keywords and labels of the course, such as a computer aided geometric design course, and the feature word set of the course is { "computer", "geometric design", "mathematics", "programming" };
the course similarity calculation is converted into cosine similarity calculation of course vectors in a multidimensional space, and the cosine similarity calculation of the course vectors is as follows:
suppose course vectors for course i and course j are respectively
And
the calculation formula of the cosine similarity between the two course vectors is as follows (1):
and (3) calculating the similarity among all courses by using a formula (1) to finally obtain a course similarity matrix.
Step 2: acquiring historical school scores of students, and acquiring objective scores of the students through a school similarity matrix; the method comprises the following specific steps:
2-1, according to the course similarity matrix (1), satisfying the formula (2), selecting a course set D related to the course C to be predicted;
simlar(i,j)≥S (2)
wherein S is an artificially defined threshold;
2-2, judging whether the corrected courses exist in the relevant course set D, and if not, taking the school entering score or the initial value set by the school side as the basic score of the course C to be predicted; if yes, screening out relevant repaired courses of the current course C to be predicted of the student from the relevant course set D, constructing a relevant repaired course E, sequencing all courses in the relevant repaired course set E from large to small according to the similarity value of the course C to be predicted and the relevant repaired courses, then obtaining N relevant repaired courses sequenced in the front, wherein N is set manually according to experience, and finally obtaining the basic score of the course C to be predicted by adopting a formula (3).
The academic achievement can be an academic test achievement, a college entrance achievement or a middle school achievement and the like.
Wherein P-scoreA,CIndicating the basic achievement of the student in the course C to be predicted,
indicating the related revised course E of the kth doorkThe performance of (a) is determined,
simlar (C, E) represents the average score of the student's N related classes takenk) Representing courses C to be predicted and associated taken courses EkThe similarity between them.
The basic achievement prediction process of the students is shown in figure 2.
And step 3: and acquiring daily behavior data and learning data of the students, and performing data cleaning on the original data to obtain normalized data serving as attributes of the students.
The student daily behavior data and the learning data comprise student basic information, student scores and student independent learning behavior records; the basic information of the students refers to the study numbers, names, sexes, ages, grades and belonged professions of the students; the student score represents the classroom performance score, the post-classroom work score, the classroom test and the mid-term examination score of the student; the self-learning behavior records comprise data of the student such as the number of times of library visit, the number of times of book borrowing, the access time of the library, the class late arrival rate, the number of times of leave, the completion condition of post-class work and the like; if the data can not be directly acquired, the data can be acquired in a data statistical mode.
The data statistics mode can be that data records of student library access, book borrowing, classroom check-in, leave asking, post-class homework submission and the like are collected from the system, then data are divided according to a schooling period, data in the same period are classified and counted according to a schooling number or a course, and finally the counted data are used as student attribute data.
For example, obtaining access records of library personnel within a period of time, screening out records of related school numbers, and counting the number of data to obtain library access frequency data of related students;
and when the students enter the library and leave the library, subtracting the time of the students leaving the library to obtain the access duration data of the library of the students.
Other data can also be obtained by this method.
The data cleaning is to process repeated values and missing values in the data. Wherein, one piece of data is reserved for the repeated value, and the rest data are deleted; deleting or filling the missing value, and deleting the field when the missing value of one attribute field reaches one third; if only a small amount of data is missing, a repair value can be obtained by taking the average number or the median of all data of the field, and the data missing part is filled in.
The whole process flow is shown in fig. 3.
Finally, the student attributes are collected into a student attribute table, as shown in table 1:
TABLE 1 student Attribute Table
And 4, step 4: converting the discrete variables in the attributes of the students in the step 3 into numerical types, vectorizing all data, bringing vectors into a neural network, constructing and training a weight calculation model, and obtaining the weight of the attributes of the students; multiplying the obtained student attribute weight and the corresponding attribute value to obtain a student weighted attribute value;
4-1, converting the discrete variables in the attributes of the students in the step 3 into numerical value types, and obtaining attribute set X of the students { X ═ X }1,x2,...,xnThe concrete steps are as follows:
if the current discrete variables have obvious sequential relation or size relation, the data can be converted into numerical data in a mapping mode, for example, the average night bedtime can be given the following numerical values according to the time period: {21:00 before: 1,21:00-21:30: 2,21:30-22:00: 3,22:00-22:30: 4,22:30-23:00: after 5, 23: 00: 6 }; if the current discrete variable is unordered data, encoding is carried out according to the data type, for example, students belong to the professions, a unique encoding numerical value is given to each profession, and the parallel relation among the data is reflected.
4-2 vectorizing all data as follows:
converting all student attribute values into attribute matrix through randomly generated conversion matrix
Wherein m is the vector dimension of each attribute, and n is the number of attributes.
Wherein each column vector A in the attribute matrixiVectors representing corresponding attributes, Ai＝(a1i,a2i,...,ami)T，amiAnd m-th dimension data representing the ith attribute.
4-3 construction and training of Multi-layer perceptron (MLP) vector AiAs input, outputs the corresponding weight uiFinally, the weight U ═ of all student attributes is obtained (U)1,u2,...,un) (ii) a The multi-layer perceptron (MLP) is an existing conventional model and is not explained in detail;
ui＝MLP(Ai)i＝1,2,...,n (5)
4-4 all weights U are processed by softmax function and mapped into (0,1) interval, and the formula is as follows:
wherein v isiRepresenting the value of an attribute xiThe weight of (c).
Thus, the final set of attribute weights is V ═ V (V)1,v2,...,vn)。
4-5, multiplying each attribute weight by the attribute value respectively to obtain the weighted attribute value, wherein the calculation formula is as follows:
fi＝vi×xi (7)
thus, the current student's weighted attribute set F ═ F1,f2......fn}。
And 5: and (3) constructing a result prediction model, and taking the student weighted attribute set F obtained in the step (4) and the student objective result P-score obtained in the step (2) as input to obtain a final result prediction value of the course C to be predicted by the student.
The achievement prediction model adopts a three-layer fully-connected neural network. The whole flow is shown in fig. 4.
Finally, experiments were performed on the UCI public data set Student Performance. The five common regression algorithms of experiments and decision trees, random forests, linear regression, support vector machines and gradient lifting trees are compared, and Mean Absolute Error (MAE), Root Mean Square Error (RMSE) and coefficient of solution (R) are adopted2R Squared) as an evaluation index. Where MAE represents the mean of the absolute values of the differences between the predicted and true values, RMSE represents the root of the ratio of the sum of the squares of the differences between the predicted and true values to the number of samples, R2Reflecting the extent of interpretation of the variability of the dependent variable y by the independent variable x. The smaller the MAE and RMSE, the better R2Closer to 1 indicates better model fit.
Wherein n represents the number of test samples, yiThe actual value is represented by the value of,
the mean value of the true values is represented,
indicating the predicted value.
The results are shown in table 2:
TABLE 2 model comparison experiment prediction results
Compared with other 5 common regression models, all indexes of the student subjective and objective factor combined achievement prediction method provided by the invention reach the best.
Claims (8)
1. A student subjective and objective factor combined score prediction method is characterized by comprising the following steps:
step 1: acquiring relevant information of all courses in the smart campus, and acquiring a feature word set of each course; then converting the feature word set of each course into a course vector of each course; finally, calculating the course similarity according to the course vector of each course to obtain a course similarity matrix;
step 2: acquiring historical school scores of students, and acquiring objective scores of the students through a school similarity matrix; the method comprises the following specific steps:
2-1, screening out a relevant course set D of the course C to be predicted according to the course similarity matrix;
2-2, judging whether the corrected courses exist in the relevant course set D, and if not, taking the school entering score or the initial value set by the school side as the basic score of the course C to be predicted; if yes, screening out relevant repaired courses of the current course C to be predicted of the student from the relevant course set D, and constructing a relevant repaired course set E; according to the similarity value of the curriculum C to be predicted and the relevant curriculum to be revised, all curriculums in the relevant curriculum to be revised set E are sorted from big to small, then the relevant revised curriculum with the front N times of the sorting is obtained, N is set manually according to experience, and finally the basic score of the curriculum C to be predicted is obtained by adopting a formula (3);
wherein P-scoreA,CIndicating the basic achievement of the student in the course C to be predicted,
indicating the relevant revised class at the kth doorEkThe performance of (a) is determined,
simlar (C, E) represents the average score of the student's N related classes takenk) Representing courses C to be predicted and associated taken courses EkThe similarity between them;
and step 3: acquiring daily behavior data and learning data of students, and cleaning the data to be used as attributes of the students;
and 4, step 4: constructing and training a weight calculation model MLP, and taking student attribute vectorization as input to obtain the weight of the student attribute; multiplying the obtained student attribute weight and the corresponding attribute value to obtain a student weighted attribute value;
and 5: building a result prediction model, and taking the student weighted attribute set F obtained in the step 4 and the student objective result P-score obtained in the step 2 as input to obtain a final result prediction value of the course C to be predicted by the student;
the achievement prediction model adopts a three-layer fully-connected neural network.
2. The method as claimed in claim 1, wherein the information related to the courses in step 1 includes the names of the courses, the subjects to which the courses belong, the relationship between the courses, and the descriptions of the courses.
3. The method as claimed in claim 1, wherein the feature words of the lesson in step 1 refer to keywords and tags of the lesson.
4. The method as claimed in claim 1, wherein the course similarity in step 1 is a cosine similarity of course vectors in a multidimensional space, and the cosine similarity of course vectors is calculated as follows:
suppose course vectors for course i and course j are respectively
And
the calculation formula of the cosine similarity between the two course vectors is as follows (1):
5. the method for predicting the achievement combined with the subjective and objective factors of the student as claimed in claim 1, wherein the step 2-1 screens out a relevant course set D related to the course C to be predicted according to the course similarity matrix and satisfying the formula (2);
simlar(i,j)≥S (2)
where S is an artificially defined threshold.
6. The method for predicting the performance of the combination of subjective and objective factors of the students as claimed in claim 1, wherein the daily behavior data and learning data of the students in the step 3 comprise basic information of the students, the performance of the students and the independent learning behavior records of the students; the basic information of the students refers to the study numbers, names, sexes, ages, grades and belonged professions of the students; the student scores comprise classroom performance scores, classroom post-work scores, classroom tests and mid-term examination scores of students; the self-study behavior records comprise the access times of the student in the library, the borrowing times of the book, the access time of the library, the late arrival rate of the classroom, the number of times of asking for vacation and the completion condition of post-class work.
7. The method for predicting the achievement of the combination of the subjective and objective factors of the students according to claim 1, wherein the step 4 is as follows:
4-1, converting the discrete variables in the attributes of the students in the step 3 into numerical value types, and obtaining attribute set X of the students { X ═ X }1,x2,...,xn}；
4-2 vectorizing all data as follows:
converting all student attribute values into attribute matrix through randomly generated conversion matrix
Wherein m is the vector dimension of each attribute, and n is the number of attributes;
wherein each column vector A in the attribute matrixiVectors representing corresponding attributes, Ai＝(a1i,a2i,...,ami)T，amiM-dimensional data representing an i-th attribute;
4-3 constructing and training a multi-layer perceptron MLP, and converting a vector AiAs input, outputs the corresponding weight uiFinally, the weight U ═ of all student attributes is obtained (U)1,u2,...,un)；
ui＝MLP(Ai)i＝1,2,...,n (5)
4-4, processing all the weights U through a softmax function, and mapping the weights U into a (0,1) interval;
wherein v isiRepresenting the value of an attribute xiThe weight of (c);
thus, the final set of attribute weights is V ═ V (V)1,v2,...,vn)；
4-5, multiplying each attribute weight by the attribute value respectively to obtain weighted attribute values;
fi＝vi×xi (7)
thus, the current student's weighted attribute set F ═ F1,f2......fn}。
8. A student subjective and objective factor combined score prediction system is characterized by comprising an objective score prediction module, a student weighted attribute calculation module and a final score prediction module;
wherein:
the objective performance prediction module comprises a first data acquisition submodule, a course similarity matrix calculation module and an objective performance calculation module;
the first data acquisition submodule is used for acquiring relevant information of all courses in the smart campus and historical course scores of students;
the course similarity matrix calculation module is used for extracting characteristic words from the course information acquired by the first data acquisition submodule and calculating the similarity between courses to obtain a course similarity matrix;
the objective score calculating module is used for calculating the historical school scores of the students obtained by the first data acquisition submodule through a school similarity matrix to obtain the basic scores of the students;
the student weighted attribute calculation module comprises a second data acquisition submodule, a data processing module and a weighted attribute calculation module;
the second data acquisition submodule is used for acquiring daily behavior data and learning data of students;
the data processing module is used for cleaning the student data acquired by the second data acquisition submodule to obtain normalized data;
the weighted attribute calculation module is used for vectorizing the normalized data obtained by the data processing module and then bringing the normalized data into the weight calculation model to obtain the weight of the student attributes; multiplying the calculated weight by the attribute value to obtain a weighted attribute value of the student;
and the final result prediction module is used for taking the weighted attribute value and the basic result of the student as input and obtaining the final result prediction value of the course to be predicted of the current student through the result prediction model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111050024.4A CN113935869A (en) | 2021-09-08 | 2021-09-08 | Student subjective and objective factor combined score prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111050024.4A CN113935869A (en) | 2021-09-08 | 2021-09-08 | Student subjective and objective factor combined score prediction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113935869A true CN113935869A (en) | 2022-01-14 |
Family
ID=79275568
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111050024.4A Pending CN113935869A (en) | 2021-09-08 | 2021-09-08 | Student subjective and objective factor combined score prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113935869A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114912027A (en) * | 2022-05-31 | 2022-08-16 | 济南大学 | Learning scheme recommendation method and system based on learning outcome prediction |
-
2021
- 2021-09-08 CN CN202111050024.4A patent/CN113935869A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114912027A (en) * | 2022-05-31 | 2022-08-16 | 济南大学 | Learning scheme recommendation method and system based on learning outcome prediction |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107230174B (en) | Online interactive learning system and method based on network | |
| Stapel et al. | An Ensemble Method to Predict Student Performance in an Online Math Learning Environment. | |
| CN106127634B (en) | Student academic achievement prediction method and system based on naive Bayes model | |
| CN112446591A (en) | Evaluation system for student comprehensive capacity evaluation and zero sample evaluation method | |
| CN111709575A (en) | Academic achievement prediction method based on C-LSTM | |
| WO2022170985A1 (en) | Exercise selection method and apparatus, and computer device and storage medium | |
| CN111950708A (en) | Neural network structure and method for discovering daily life habits of college students | |
| CN111126552A (en) | Intelligent learning content pushing method and system | |
| Fong et al. | Applying a hybrid model of neural network and decision tree classifier for predicting university admission | |
| CN114021722A (en) | Attention knowledge tracking method integrating cognitive portrayal | |
| CN114943275A (en) | Student performance prediction method based on multi-source sparse attention convolution network | |
| CN113935869A (en) | Student subjective and objective factor combined score prediction method and system | |
| CN110533992A (en) | A kind of learning effect prediction technique and system | |
| Wu et al. | Student achievement analysis and prediction based on the whole learning process | |
| Diao et al. | Personalized exercise recommendation driven by learning objective within e-learning systems | |
| Asril et al. | Prediction of students study period using K-Nearest Neighbor algorithm | |
| CN116361541A (en) | Test question recommendation method based on knowledge tracking and similarity analysis | |
| Zhang et al. | Research on student Big Data portrait method based on improved K-means algorithm | |
| CN113919983A (en) | Test question portrait method, device, electronic equipment and storage medium | |
| Arafiyah et al. | MONITORING LEARNERS’PERFORMANCE BY MODELING LEARNING PROGRESS USING MACHINE LEARNING | |
| CN112766765A (en) | Professional learning ability evaluation method and system based on interval middle intelligence theory | |
| Zhou | Research on teaching resource recommendation algorithm based on deep learning and cognitive diagnosis | |
| Su | Design of intelligent classroom teaching scheme using artificial intelligence | |
| Phanniphong et al. | The Distinction Learning Style in Learning Outcomes of the Secondary School Learner | |
| Jianyun | Big data assisted online teaching platform for ideological and political theory course in universities |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |