LU100314B1 - Method and system for predicting academic achievements of students based on naive bayesian model - Google Patents

Method and system for predicting academic achievements of students based on naive bayesian model Download PDF

Info

Publication number
LU100314B1
LU100314B1 LU100314A LU100314A LU100314B1 LU 100314 B1 LU100314 B1 LU 100314B1 LU 100314 A LU100314 A LU 100314A LU 100314 A LU100314 A LU 100314A LU 100314 B1 LU100314 B1 LU 100314B1
Authority
LU
Luxembourg
Prior art keywords
student
students
academic
data
achievement
Prior art date
Application number
LU100314A
Other languages
French (fr)
Inventor
Ruojuan Xue
Original Assignee
Univ Shandong
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Univ Shandong filed Critical Univ Shandong
Application granted granted Critical
Publication of LU100314B1 publication Critical patent/LU100314B1/en

Links

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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/20—Education
    • G06Q50/205—Education administration or guidance
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N7/00—Computing arrangements based on specific mathematical models
    • G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
    • G06F18/24155—Bayesian classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching

Abstract

The present invention discloses a method and a system for predicting academic achievements of students based on a naive Bayesian model. The method includes: obtaining learning data of students, and transmitting the obtained data to a database of a first server; performing data conversion on the learning data of the students stored in the database to obtain a normalized student learning status data table; for the normalized student learning status data table, calculating conditional probabilities of attributes in different categories and prior probabilities of the different categories through a calculating unit in the server; and performing data conversion on to-be-predicted student data, and inputting the to-be-predicted student data in the trained naive Bayesian model to perform classification prediction of the student data. The present invention adopts the naive Bayesian model that is applied to the accurate prediction of the academic achievements of the students with high reliability. FIG. 1

Description

METHOD AND SYSTEM FOR PREDICTING ACADEMIC ACHIEVEMENTS OF STUDENTS BASED ON NAIVE BAYESIAN MODEL
Field of the Invention
The present invention relates to the field of educational technology and computer application discipline, and in particular to a method and a system for predicting academic achievements of students based on a naive Bayesian model.
Background of the Invention
At present, with rapid development of the higher education in China, the number of ordinary institutions of higher learning has reached more than 2,500. In recent years, the continuous decline of student source has brought a huge crisis of survival to the ordinary institutions of higher learning. How to improve the training quality of students and improve the employment competitiveness of the students has become an urgent problem to be solved by many colleges and universities. The academic achievements of students, as an important core index for the training quality of the students, have earned much attention of managers of the colleges and universities.
In the colleges and universities in China, counselors in the Youth League Committee (class sponsors) are usually responsible for the daily management of the students, and teachers are responsible for the curriculum theory and professional skills teaching of the students. There is often a lack of effective communication between the counselors and the teachers, which may cause some students to go astray because of lack of discipline and to be forced to postpone graduation or drop out due to poor academic achievements. The academic achievements of the students are often affected by many factors, including previous academic achievements of the students, learning ability, teacher guidance and many other factors. If the academic achievements of the students can be predicted according to the historical academic achievements, the comprehensive performance of all aspects and the quality conditions of the students, for the students who may be in trouble, the management and education are enhanced in time, then they are urged to study hard so as to avoid the consequences of failing to pass their academic examinations according to prediction results of the academic achievements, which greatly facilitates the education and management of the students by the counselors, and plays an important role to improve the training quality of the students.
At present, all kinds of teaching management systems have been very common in the institutions of higher learning and can effectively manage the learning achievements of the students. However, researches on the prediction and analysis of the academic achievements of the students are still very rare and are not widely implemented neither. 1. The existing teaching management systems focus only on the management of learning achievement data of the students, but ignore the management of other behavior data of the students. The collection of student data is incomplete, and it is difficult to carry out comprehensive analysis and evaluation on the students. 2. With respect of the data of student achievements, only the student achievements are recorded in the teaching management systems at present; and historical data of the student achievements are stored in the teaching management systems. The evaluation of the existing ability of the students is only obtained through the analysis of the historical data of the student achievements, corresponding data processing models are not adopted, and intelligent prediction of the academic achievements of the students cannot be realized.
The patent of the present invention applies the data mining technology to the prediction of the academic achievements in view of these technical problems in the prediction of the academic achievement of the students, realizes a method and a system for predicting academic achievements of students based on the naive Bayesian model, and strives to promote the development of this research.
Summary of the Invention
In order to solve the shortcomings of the prior art, the present invention discloses a method and a system for predicting academic achievements of students based on the naive Bayesian model. The present invention realizes the intelligent prediction of the academic achievements of the students by adopting corresponding data acquisition and analysis technology.
To achieve the above purpose, the specific solution of the present invention is as follows: A method for predicting academic achievements of students based on a naive Bayesian model includes the following steps: step 1: obtaining learning data of the students, and transmitting the obtained data to a database of a first server; step 2: performing data conversion on the learning data of students stored in the database to obtain a normalized student learning status data table; step 3: for the normalized student learning status data table, calculating conditional probabilities'of attributes in different categories and prior probabilities of the different categories through a calculating unit in the first server, and learning parameters of the naive Bayesian model to obtain the naive Bayesian model; and step 4: performing data conversion on to-be-predicted student data, inputting the to-be-predicted student data in the trained naive Bayesian model to perform classification prediction of the academic achievements to obtain prediction results of the academic achievements of the students, and displaying the prediction results of the academic achievements of the students through a display unit.
Further, in step 1, the learning data of the students include academic achievement information of the students, wherein the academic achievements are information stored in a database server of a teaching management system, and the database server of the teaching management system communicates with the first server to transmit the academic achievements of the students to the first server.
Further, the academic achievement information includes academic achievements of two adjacent semesters and admission academic achievements of the students, wherein the academic achievement conditions of the previous semester and the admission academic achievement conditions will be used as historical academic achievement attributes of student individuals; and the academic achievement conditions of the later semester are used as classification results of the academic achievements of the students.
Further, in the step 1, the learning data of students further include learning behavior information, the learning behavior information is obtained by a data collection terminal, and the data collection terminal can be a computer or a mobile intelligent device.
Further, the learning behavior information specifically includes learning times, online entertainment times, library usage frequency and borrowed book types, etc.
Further, when the academic achievements of the students are specifically obtained, achievement data of the students and class achievement tables of the classes are extracted from the database server of the teaching management system by using student numbers of the students as search terms.
Further, the student data are converted in the first server, and the obtained continuous data are converted into grade data in intervals where the student information data are located in a segmentation manner.
Further, the academic achievement information includes the academic achievements of two adjacent semesters and the admission academic achievements of the students, and the data need to be converted, wherein the specific processing flows are as follows: obtaining the class achievement tables, calculating average achievements of the students according to the numbers of examination subjects of the students, sorting according to the average achievements of the students, and outputting class ranking tables; and outputting total numbers of students in the classes; inquiring the rankings of the students according to the class ranking tables and the achievement data of the students, and outputting the rankings of the students; judging the overall positions of the rankings of the students in the classes according to the rankings of the students and the total numbers of the students in the classes; and if the rankings of the students belong to the top 20%, then outputting the grades of the academic achievements of the students as A; if the rankings of the students belong to an interval between 20% and 40%, then outputting the grades of the academic achievements of the students as B; if the rankings of the students belong to an interval between 40% and 60%, then outputting the grades of the academic achievements of the students as C; if the rankings of the students belong to an interval between 60%-80%, then outputting the grades of the academic achievements of the students as D; and if the rankings of the students belong to the bottom 20%, then outputting the grades of the academic achievements of the students as E.
Further, in step 3, the conditional probability parameters are specifically calculated as follows: 3-1) traversing the normalized student learning status data table to count the number of students of category G, and outputting Count(G); 3-2) traversing the normalized student learning status data table to count the number of students of category G and with attribute values of the rth attribute being xr, and outputting Count(GG); 3-3) traversing the normalized student learning status data table to count the number of categories of the academic achievements of the students, and outputting a number value K; and 3-4) calculating /YGG) according to the Count(G) obtained in step 3-1), the Count(xr|G) obtained in step 3-2) and the K obtained in step 3-3), and outputting Pfxr|G9, wherein the computational formula is:
wherein, λ is 0.1.
Further, in step 3, the prior probability parameters are specifically calculated as follows: 1) traversing the normalized student learning status data table to count the number of students of category G, and outputting Count(G); 2) traversing the normalized student learning status data table to count the number of categories of the academic achievements of the students, and outputting a number value K\ 3) traversing the normalized student learning status data table to count the total number of students, and outputting a number value A; and 4) calculating P(Ci) through a formula according to the Count(G) obtained in step 1), the K obtained in step 2) and the N obtained in step 3), and outputting P(Ci), wherein the computational formula is:
wherein, λ is 0.1.
Further, in step 4, the specific steps are as follows: 4-1) calculating P(X\Ci)P(Ci) values of the category G according to the obtained probability parameters P(xr\Ci) and P(Ci)·, and 4-2) comparing the P(T|G9PO values obtained in step 4-1) and corresponding to the category G, and predicating the category of a student sample Xas G with the maximum P(X\Ci)P(Ci) value. Further, in step 4-1), the calculating process is as follows: 4-1-1) for each attribute xrof the student sample X, obtaining the calculated conditional probability parameter Pfxr|C() values in sequence; then, multiplying all P(xr\Ci) according to a formula P(X|C,) = P(x, |C,)xP(x2 |C()x......xP(xJC,) to obtain P(XC,) values, and outputting the P(X\Ci) values; and 4-1-2) multiplying the P(X\Ct) values calculated in step 4-1-1) with the prior probability parameter P(Ci) values to obtain the P(X\Ci)P(Ci) values, and outputting the P(X\Ci)P(Ci) values. A system for predicting academic achievements of students based on a naive Bayesian model includes: a data obtaining module, used for obtaining learning data of the students, and transmitting the
obtained data to a database of a first server; a data conversion module, used for performing data conversion on the learning data of the students stored in the database to obtain a normalized student learning status data table; a naive Bayesian model establishment module, used for, with respect to the normalized student learning status data table, calculating conditional probabilities of attributes in different categories and prior probabilities of the different categories through a calculating unit in the server to obtain the naive Bayesian model; and a student academic achievement prediction module, used for performing data conversion on to-be-predicted student data, inputting the to-be-predicted student data in the trained naive Bayesian model to perform classification prediction of the student data to obtain a student academic achievement prediction result, and displaying the student academic achievement prediction result through a display unit.
The present invention has the following beneficial effects: 1. The present invention provides the method for predicting the academic achievements of the students based on the naive Bayesian model to predict the academic achievements of the students in the future so as to enable the institutions of higher learning to enhance the educational management of the students. 2. The present invention describes student samples through 14 attributes of the students such as the historical academic achievements, the learning behavior information and the like, the sample data can be obtained by the teaching management system and the data collection terminal of the school, the data sources are simple, convenient and accurate, and thus the present invention can be widely popularized in the institutions of higher learning. 3. The present invention performs data conversion on the obtained data to obtain the normalized student learning status data table, and converts the student information data into grade data in a segmentation manner according to the intervals thereof to reduce the number of the attribute values and facilitate the use of the data in successive modeling. 4. The present invention stores the obtained student data information in the database of the first server so as to facilitate the successive data scheduling and processing and ensure the security and the stability of the data. 5. At present, the student management tasks of the counselors of the institutions of higher learning are heavy, it is difficult to take each student into account, the present invention will be able to LU I U’ effectively predict the students with upcoming academic problems for the counselors, which is conductive to enhancing the working pertinence of the counselors and improving the training quality of the students of the institutions of higher learning. 6. The present invention adopts the naive Bayesian model that is applied to the accurate prediction of the academic achievements of the students with high reliability.
Brief Description of the Drawings
Fig.l is an overall prediction flowchart of the present invention;
Fig.2 is a specific acquisition flowchart of academic achievements of students of the present invention;
Fig.3 is a specific processing flowchart of conversion processing of academic achievement information data of the present invention;
Fig.4 is a specific flowchart of calculation of probability parameters of attributes in different categories of the present invention; and
Fig. 5 is a specific flowchart of calculation of products of conditional probabilities and prior probabilities of categories of the present invention.
Detailed Description of the Embodiments
The present invention will be illustrated below in detail in combination with the drawings:
As shown in Fig.l, the method for predicting academic achievements of students based on a naive Bayesian model of the present invention mainly includes the following steps: stepl: collecting student data comprising academic achievements and learning behavior information of the students; step 2: converting the student data to obtain a normalized student learning status data table; step 3: constructing a student academic achievement prediction model based on the naive Bayesian model through the normalized student learning status data table, and learning the parameters of the prediction model; step 4: predicting academic achievement categories of the students by using the naive Bayesian model according to obtained model probability parameters.
By adoption of the present invention, the academic achievements of the students in the future can be predicted according to daily learning statuses of the students, so that the institutions of higher learning can conveniently enhance the educational management of the students.
In step 1, the academic achievements of the students can be derived directly from a teaching management system database, other learning behavior information is obtained by carrying out questionnaires of the students (electronic questionnaires can also be released through the network) or obtained by using a data collection terminal, preferably, the other learning behavior information is obtained by using the data collection terminal, the data collection terminal is a computer terminal or a mobile device, student numbers are used as the IDs of the students, each student has one and only one student number, and each student number has a corresponding data record, this is because the data obtained in such a manner can avoid the problem of data distortion caused by possible counterfeit of the questionnaires.
When the electronic questionnaires are released through the network to obtain the data, the electronic questionnaires are in one-to-one correspondence with the student numbers of the students, and contents filled by each student are collectively summarized and processed.
The other learning behavior information can include: learning times, attendance on time, online entertainment times, library usage frequency, borrowed book types, time management ability, learning ability, extracurricular activities, teacher guidance, family guidance and professional interests.
For example, at the beginning of the learning times of the students, a computer starts timing, and the computer stops timing when the learning times of the students are ended, and thus the learning times of the students are obtained; and with respect to the attendance on time, each student performs fingerprint recognition on classes in a fingerprint recognition manner, and the students performing no fingerprint recognition are deemed as absence, and the attendance on time of the students is obtained in this way.
The way of obtaining the online entertainment times is similar to the way of obtaining the learning times of the students.
The library usage frequency and the borrowed book types can be obtained from a database server of a book management system of the school, and the data stored in the database server of the book management system are transmitted to the computer.
The time management ability, the learning ability, the extracurricular activities, the teacher guidance, the family guidance, the professional interests and other data can be determined with combination of self evaluation and the evaluation of teachers and other students.
The attributes of all categories of student information are as shown in Table 1.
Table 1
With respect to the academic achievement information, academic achievements of two adjacent semesters and admission academic achievements of the students are collected, wherein the academic achievement conditions of the previous semester and the admission academic achievement conditions will be used as historical academic achievement attributes of student individuals; and the academic achievement conditions of the later semester are used as classification results of the academic achievements of the students. Specific obtaining steps of the all categories of academic achievements of the students are as shown in Fig.2. According to the student numbers of the students and a student academic achievement database of the school, student academic achievement data and class achievement tables of the classes of the students are extracted.
In step 2, the academic achievement information includes the academic achievements of two adjacent semesters and admission academic achievements of the students, and the data require conversion processing. The specific processing flows are as shown in Fig.3.
For the obtained class achievement tables, average achievements of the students are calculated according to the numbers of examination subjects of the students, sorting is performed according to the average achievements of the students, and class ranking tables are output; and total numbers of students in the classes are output.
The rankings of the students are inquired according to the class ranking tables and the achievement data of the students, and the rankings of the students are output.
The overall positions of the rankings of the students in the classes are judged according to the rankings of the students and the total numbers of the students in the classes. If the rankings of the students belong to the top 20%, then the grades of the academic achievements of the students are output as A; if the rankings of the students belong to an interval between 20% and 40%, then the grades of the academic achievements of the students are output as B; if the rankings of the students belong to an interval between 40% and 60%, then the grades of the academic achievements of the students are output as C; if the rankings of the students belong to an interval between 60%-80%, then the grades of the academic achievements of the students are output as D; and if the rankings of the students belong to the bottom 20%, then the grades of the academic achievements of the students are output as E.
Other learning behavior information and the converted academic achievement information are combined together to obtain a student learning status data table. In the embodiment, it is assumed that the student learning status data table as shown in table 2 is obtained.
Table 2
In step 3, the student academic achievement prediction model based on the naive Bayesian model is constructed through the normalized student learning status data table, and the parameters of the prediction model are learnt.
For ease of presentation, in the present invention, the fields “academic achievements in the later semester” in table 2 are marked as C, the five categories A, B, C, D and E are respectively marked as Ci, C2, C3, C4 and C5, and the other 13 fields in table 2 are marked as R1-R13 in sequence. According to the naive Bayesian model, the category C of the academic achievements in the later semester of the student X can be predicted by formula (1). 1
In the naive Bayesian model, the attributes are deemed as being conditionally independent from
each other. Therefore, P(X\Ci) in the formula (1) can be calculated by formula (2).
In order to avoid the situation of zero probability, the present invention performs smooth processing on the probabilities in formulas (1) and (2) by adopting the Laplasse algorithm, as shown in formulas (3) and (4).
wherein, Λ is 0.1, K represents the number of categories of the academic achievements, and N represents the total number of students; Count(xr\C,) represents the number of students with values of the rth attribute being xr in the samples of the category G; and Count(Ci) represents the number of students of the category G.
With the student X as an example, assuming that after the data of X are converted, the learning times, the attendance on time, the online entertainment times, the library usage frequency, the borrowed book types, the time management ability, the learning ability, the extracurricular activities, the teacher guidance, the family guidance, the professional interests, the academic achievement conditions of the previous semester and the admission academic achievement conditions are respectively: 4> C, 2, <2, novels, Poor, General, Relatively good, General, Relatively good, yes, C and Bo
In order to predict the academic achievement of the student X in the current semester, all probability parameters of the student academic achievement prediction model based on the naive Bayesian model need to be leamt at first according to the student learning status data table. The specific implementation steps are as follows.
According to the student learning status data table, the probability parameters P(xr\Ci) of the attributes of different categories are calculated, and smooth processing is performed by using the Laplasse algorithm. The specific implementation flows are as shown in Fig.4.
The student learning status data table is traversed to count the number of students of the category G, and Count(G) is output.
The student learning status data table is traversed to count the number of students of the category G with the attribute value of the rth attribute being xr, and Count(xr|G) is output.
The student learning status data table is traversed to count the number of categories of the achievements of the students, and a number value K is output. P(xr\Ci) is calculated according to the Count(C,), the Count(xrjC;) and the Kobtained, and P(xr\Ci) is output.
In the embodiment, according to Table 2, the calculating results of the probability parameters P(xr\Ci) of the attributes of different categories are as follows: for P(xi\Ci), Count(C,= ”A ”)=4, Count(7?/= ”4”|C;=”?1”)=1, and K=5 are substituted in formula (3) to obtain:
similarly,
According to the student learning status data table, the probability parameters P(Ci) of samples of different categories are calculated, and smooth processing is performed by using the Laplasse algorithm. The specific implementation flows are as shown in Fig.4.
The student learning status data table is traversed to count the number of students of the category G, and Count(G) is output.
The student learning status data table is traversed to count the number of categories of the academic achievements of the students, and a number value K is output.
The student learning status data table is traversed to count the total number of the students, and a number value N is output. P(Ci) is calculated according to Count(G), K and N obtained through formula (4), and P(C,) is output.
In the embodiment, according to Table 2, the calculating results of P(Ct) of the attributes of different categories are as follows: for P(Ci), Count(G= ”A ”)=4, K=5 and A=20 are substituted in formula (4) to obtain:
Similarly, it can be obtained that P(C2), P(Ci), P(C4) and P(Cs) are all 0.2.
Step 4, according to the model probability parameters obtained in step 3, the categories of the academic achievements of the students are predicted by using the naive Bayesian model.
It can be seen from the formula (1) of the naive Bayesian model that, the category G ensuring the
maximum value of P(X\Ci)P(Ci) is the prediction results of the academic achievements, specific implementation steps are as follows.
Step 4-1: P(X\Cj)P(Ci) values of the category G are calculated according to the probability parameters P(xr\Ct) and P(Q) obtained in step 3.
The specific implementation flows of step 4.1 are as shown in Fig.5.
Firstly, for each attribute xrof the student sample X, the P(xr\Ci) values are obtained in sequence; then, the P(Xr|G7 are multiplied according to formula (2) to obtain P(Y]G^ values, and the P/X]G> values are output.
The calculated P(%|G9 values are multiplied with the P(Ct) values to obtain the P(X\Ci)P(Ct) values, and the P(X}Cj)P(Ci) values are output.
In the embodiment, taking the student X as an example, the specific calculating method of the step is illustrated.
Similarly,
Step 4.2: the P(X\Ci)P(Ci) values obtained in step 4-1 and corresponding to the categories G , and the category of the student sample X is predicted as G with the maximum P(X\ Ci)P(Ct) value.
In the embodiment, by comparison of the P(X\Ci)P(Ci) values of the categories, it can be seen that
the value of the Cj (“E”) is the maximum. Therefore, the academic achievement of the sample X in the current semester is predicted as Cj (“E”) .
According to the predicted academic achievement of the student sample, the academic achievement belongs to the “E” class, namely the ranking will be behind 80%, and obviously the educational management of the student should be enhanced. The counselor can interpose the student in time based on the prediction result, performs critical education on the student so as to correct the bad habits and correct the student’s attitude, in order to avoid a serious problem of the academic achievement of the student.
It should be noted that the student academic achievement prediction method and system based on the naive Bayesian model in the present application are implemented on the basis of the existing hardware products such as computers and servers, and the obtained prediction results can be displayed by the corresponding display unit.
Although the specific embodiments of the present invention have been described above in combination with the drawings, the protection scope of the present invention is not limited thereto, and those skilled in the art to which the art pertains should be aware that various modifications or variations made by those skilled in the art on the basis of the technical solutions of the present invention without creative effects shall still fall within the protection scope of the present invention.

Claims (10)

1. Méthode pour prédire les réussites académiques ou scolaires des élèves sur la base d'un modèle bayésien naïf, comprenant les étapes suivantes : étape 1 : obtenir des données d'apprentissage des élèves, et transmettre des données obtenues à une base de données d'un premier serveur ; étape 2 : effectuer une conversion de données sur les données d’apprentissage des élèves stockées dans la base de données, pour obtenir un tableau normalisé de données d'état de l'apprentissage des élèves ; étape 3 : pour le tableau normalisé des données d'état de l'apprentissage des élèves, le calcul des probabilités conditionnelles des caractéristiques dans différentes catégories et des probabilités antérieures des différentes catégories à l'aide d'une unité de calcul dans le premier serveur, apprendre les paramètres du modèle bayésien naïf pour obtenir le modèle bayésien naïf ; et étape 4 : effectuer une conversion de données sur les données à prédire des élèves, entrer les données à prédire des élèves dans le modèle bayésien naïf formé pour effectuer une prédiction de classification de la réussite académique afin d'obtenir un résultat de prédiction de la réussite académique des élèves, et afficher le résultat la prédiction de la réussite académique des élèves par une unité d'affichage.1. Method for predicting student academic or academic successes based on a naive Bayesian model, including the following steps: Step 1: Obtain student learning data, and transmit data to a student database. a first server; Step 2: Perform a data conversion on the student learning data stored in the database, to obtain a standardized table of student learning status data; Step 3: For the standardized table of student learning status data, the calculation of the conditional probabilities of the characteristics in different categories and the prior probabilities of the different categories using a calculation unit in the first server , learn the parameters of the naive Bayesian model to obtain the naive Bayesian model; and Step 4: Perform a data conversion on the data to predict students, enter the data to predict students into the Bayesian naive model formed to perform a prediction of classification of academic achievement to obtain a prediction result of the Students' academic achievement, and display the result the prediction of academic achievement of students by a display unit. 2. Méthode pour prédire les réussites académiques des élèves sur la base d'un modèle bayésien naïf selon la revendication 1, dans laquelle, à l'étape 1, les données d'apprentissage des élèves comprennent l'information sur les résultats académiques des élèves, dans laquelle les réussites académiques sont des informations stockées dans un serveur de base de données d'un système de gestion d'enseignement, et le serveur de base de données du système de gestion de l'enseignement communique avec le premier serveur pour transmettre les résultats académiques des élèves au premier serveur.A method for predicting student academic successes based on a naïve Bayesian model according to claim 1, wherein in step 1, the student learning data includes information about student academic achievement. , in which the academic successes are information stored in a database server of a teaching management system, and the database server of the teaching management system communicates with the first server to transmit the students' academic results at the first server. 3. Méthode pour prédire les réussites académiques des élèves sur la base d'un modèle bayésien naïf selon la revendication 2, dans laquelle l'information sur les résultats académiques comprend les résultats académiques de deux semestres voisons et les résultats académiques d'admission des élèves, dans laquelle les conditions de réussite académique du semestre précédent et les conditions de réussite académique d'admission sont utili- sées comme caractéristiques historiques de réussite académique des élèves ; et les conditions de réussite académique du dernier semestre sont utilisées comme résultats de classification des réussites académiques des élèves.3. A method for predicting academic achievement of students based on a naive Bayesian model according to claim 2, wherein the academic outcome information includes the academic results of two semesters and the academic admission results of the students. , in which the academic success conditions of the previous semester and the conditions for academic admission success are used as historical characteristics of students' academic success; and the conditions for academic success in the last semester are used as the results of classifying students' academic achievements. 4. Méthode pour prédire les réussites académiques des élèves sur la base d'un modèle bayésien naïf selon la revendication 1, dans laquelle à l'étape 1, les données d'apprentissage des élèves comprennent en outre une information de comportement d'apprentissage, l'information de comportement d'apprentissage étant obtenue par un terminal de collecte de données et le terminal de collecte de données est un ordinateur ou un dispositif mobile intelligent.A method for predicting student academic successes based on a naive Bayes model according to claim 1, wherein in step 1, the student learning data further comprises a learning behavior information, the learning behavior information being obtained by a data collection terminal and the data collection terminal is a computer or a smart mobile device. 5. Méthode pour prédire les réussites académiques des élèves sur la base d'un modèle bayésien naïf selon la revendication 3, dans laquelle l'étape spécifique pour obtenir les résultats académiques des élèves, les données de réussite des élèves et les tableaux de réussite de classe des classes des élèves sont extraits du serveur de la base de données du système de gestion de l'enseignement avec les nombres d'étudiants des étudiants, en tant que termes de recherche.5. A method for predicting academic achievement of students based on a naive Bayesian model according to claim 3, wherein the specific step for obtaining student academic results, student success data, and student achievement charts are provided. class of students' classes are extracted from the server of the education management system database with student student numbers, as search terms. 6. Méthode pour prédire les réussites académiques des élèves sur la base d'un modèle bayésien naïf selon la revendication 2, dans laquelle les données des élèves sont converties dans le premier serveur, les données continues obtenues sont converties en données de qualité dans des intervalles où les données d'information des élèves se situent de manière segmentée.A method for predicting student academic successes based on a naïve Bayesian model according to claim 2, wherein the student data is converted in the first server, the resulting continuous data is converted to quality data in intervals. where students' information data is segmented. 7. Méthode pour prédire les réussites académiques des élèves sur la base d'un modèle bayésien naïf selon la revendication 3, dans laquelle l'information sur les résultats académiques comprend les réussites académiques de deux semestres voisins et les réussites académiques d'admission des élèves, et les données doivent être converties, les flux de traitement spécifiques étant les suivants : l'obtention de tableaux de réussite de la classe, le calcul des réussites moyennes des élèves en fonction du nombre d'examens des élèves, le tri selon les réussites moyennes des élèves, et la génération de tableaux de classement ; et la génération du nombre total d'élèves dans les classes ; l'interrogation sur les classements des élèves selon les tableaux de classement et les données de réussite des élèves, et la génération des classements des élèves ; l'évaluation des positions générales des classements des élèves dans les classes en fonction des classements des élèves et du nombre total d'élèves dans les classes; et si les classements des élèves appartiennent aux 20% les plus élevés, alors générer les notes des réussites académiques des élèves comme A ; si les classements des élèves appartiennent à un intervalle entre 20% et 40%, alors générer les notes des réussites académiques des élèves comme B ; si les classements des élèves appartiennent à un intervalle compris entre 40% et 60%, alors générer les notes des réussites académiques des élèves comme C ; si les classements des élèves appartiennent à un intervalle compris entre 60% et 80%, alors générer les notes des réussites académiques des élèves en tant que D ; et si les classements des élèves appartiennent à un niveau inférieur de 20%, alors générer les notes des réussites académiques des élèves comme E.7. A method for predicting academic achievement of students based on a naive Bayesian model according to claim 3, wherein the academic outcome information includes academic successes of two semesters adjacent to and academic achievement of student admissions , and the data must be converted, the specific treatment flows being as follows: obtaining class success charts, calculating students' average achievement based on the number of student exams, sorting by successes student averages, and the generation of league tables; and the generation of the total number of students in the classes; the questioning of the rankings of the students according to the classification tables and the student success data, and the generation of the student rankings; the evaluation of the general positions of the classifications of pupils in the classes according to the classifications of the pupils and the total number of pupils in the classes; and if the student rankings belong to the highest 20%, then generate student achievement grades such as A; if student rankings fall within the range of 20% to 40%, then generate grades for student achievement such as B; if student rankings fall within a range of 40% to 60%, then generate students' academic achievement scores such as C; if the student rankings fall within 60% to 80%, then generate students' academic achievement scores as D; and if the student rankings belong to a lower level of 20%, then generate student achievement grades such as E. 8. Méthode pour prédire les réussites académiques des élèves sur la base d'un modèle bayésien naïf selon la revendication 1, dans laquelle à l'étape 3, les paramètres de probabilité conditionnelle sont spécifiquement calculés comme suit : 3-1) parcourir le tableau normalisé des données d'état de l'apprentissage des élèves pour compter le nombre d'élèves de la catégorie Ci, et générer le nombre (Ci) ; 3-2) parcourir le tableau normalisé des données d'état de l'apprentissage des élèves pour compter le nombre d'élèves de la catégorie Ci et avec la valeur de caractéristique de la caractéristique rth étant xr, et générer le nombre (xr| Ci) ; 3-3) parcourir le tableau des données d'état de l'apprentissage des élèves normalisé pour compter le nombre de catégories de réussites académiques des élèves, et générer une valeur numérique K ; et 3-4) calculer P(xr|Ci) en fonction du nombre (Ci) obtenu à l'étape 3-1), le nombre (xr | Ci) obtenu à l’étape 3-2) et le K obtenu à l'étape 3-3), et générer P(xr | Ci), dans laquelle la formule de calcul est :8. A method for predicting academic achievement of students based on a naive Bayes model according to claim 1, wherein in step 3, the conditional probability parameters are specifically calculated as follows: 3-1) browse the table standardized state student learning data to count the number of pupils in category Ci, and generate the number (Ci); 3-2) browsing the standardized table of student learning status data to count the number of pupils in category Ci and with the characteristic value of the characteristic rth being xr, and generating the number (xr | This) ; 3-3) browse the standardized student learning state data table to count the number of student academic achievement categories, and generate a numerical value K; and 3-4) calculating P (xr | Ci) as a function of the number (Ci) obtained in step 3-1), the number (xr | Ci) obtained in step 3-2) and the K obtained at step 3-3), and generate P (xr | Ci), in which the calculation formula is: dans laquelle λ vaut 0,1 ; à l'étape 3, les paramètres de probabilité antérieurs sont spécifiquement calculés comme suit : 1) parcourir le tableau normalisé des données d'état de l'apprentissage des élèves pour compter le nombre d'élèves de la catégorie Ci, et générer le nombre (Ci) ; 2) parcourir le tableau normalisé des données d'état de l'apprentissage des élèves pour compter le nombre de catégories des réussites académiques des élèves, et générer une valeur numérique K ;in which λ is 0.1; in step 3, the previous probability parameters are specifically calculated as follows: 1) browse the standardized table of student learning status data to count the number of students in category Ci, and generate the number (This) ; 2) browse the standardized table of student learning status data to count the number of categories of students' academic achievements, and generate a numerical value K; 3) parcourir le tableau normalisé de données d'état de l'apprentissage des élèves pour compter le nombre total des élèves, et générer une valeur numérique N ; et 4) calculer P(Ci) par une formule en fonction du nombre (Ci) obtenu à l'étape 1), le K obtenu à l'étape 2) et le N obtenu à l'étape 3), et générer P(Ci), dans lequel la formule de calcul est :3) scan the standardized table of student learning status data to count the total number of students, and generate a numerical value N; and 4) calculating P (Ci) by a formula as a function of the number (Ci) obtained in step 1), the K obtained in step 2) and the N obtained in step 3), and generating P ( Ci), in which the calculation formula is: dans laquelle, λ vaut 0,1.in which, λ is 0.1. 9. Méthode pour prédire les réussites académiques des élèves sur la base d'un modèle bayésien naïf selon la revendication 8, dans laquelle à l'étape 4, les étapes spécifiques sont les suivantes : 4-1) calculer les valeurs P(X| Ci)P(Ci) de la catégorie Ci en fonction des paramètres de probabilité obtenus P (xr| Ci) et P(Ci); et 4-2) comparer les valeurs P(X | Ci)P(Ci) obtenues à l'étape 4-1) et correspondant à la catégorie Ci, et prédire la catégorie d'un échantillon d'élèves X comme Ci avec la valeur de P(X | Ci)P(Ci) maximale ; à l'étape 4-1), le processus de calcul est le suivant: 4-1-1) pour chaque caractéristique xr de l'échantillon d'élèves X, obtenir successivement le paramètre de probabilité conditionnelle calculé P(xr|Ci) ; puis multiplier P(xr|Ci) selon une formule P(X|Ci) = P(xi | Ci) x P(x21 Ci) x......x P(xn | Ci) pour obtenir des valeurs P (X | Ci), et générer les valeurs P(X | Ci) ; et 4-1-2) multiplier les valeurs P(X | Ci) calculées à l'étape 4-1-1) avec les valeurs de probabilité antérieure P(Ci) pour obtenir les valeurs P(X|Ci)P(Ci), et générer les valeurs P(X | Ci)P(Ci).9. A method for predicting academic achievement of students based on a naive Bayes model according to claim 8, wherein in step 4, the specific steps are as follows: 4-1) calculating the values P (X | Ci) P (Ci) of category Ci as a function of the probability parameters obtained P (xr | Ci) and P (Ci); and 4-2) to compare the values P (X | Ci) P (Ci) obtained in step 4-1) and corresponding to the category Ci, and to predict the category of a sample of pupils X as Ci with the value of P (X | Ci) P (Ci) maximum; in step 4-1), the calculation process is as follows: 4-1-1) for each characteristic xr of the student sample X, successively obtain the calculated conditional probability parameter P (xr | Ci) ; then multiply P (xr | Ci) according to a formula P (X | Ci) = P (xi | Ci) x P (x21 Ci) x ...... x P (xn | Ci) to obtain values P ( X | Ci), and generate the values P (X | Ci); and 4-1-2) multiplying the values P (X | Ci) calculated in step 4-1-1) with the prior probability values P (Ci) to obtain the values P (X | Ci) P (Ci ), and generate the values P (X | Ci) P (Ci). 10. Système pour prédire les réussités académiques des élèves sur la base d'un modèle bayésien naïf, comprenant: un module d'obtention de données, utilisé pour obtenir des données d'apprentissage des élèves, et transmettre les données obtenues à une base de données d'un premier serveur; un module de conversion de données, utilisé pour effectuer une conversion de données sur les données d'apprentissage des élèves stockées dans la base de données pour obtenir un tableau normalisé de données d'état de l'apprentissage des élèves ;10. A system for predicting student academic successes based on a naive Bayesian model, comprising: a data obtaining module, used to obtain student learning data, and transmitting the obtained data to a student database; data from a first server; a data conversion module, used to perform a data conversion on the student learning data stored in the database to obtain a standardized table of student learning status data; un module d'établissement de modèle bayésien naïf, utilisé pour, en ce qui concerne le tableau normalisé des données d'état de l'apprentissage des élèves, le calcul des probabilités conditionnelles des caractéristiques dans différentes catégories et des probabilités antérieures des différentes catégories par l'intermédiaire d'une unité de calcul dans le serveur pour obtenir le modèle bayésien naïf ; et un module de prédiction de la réussite académique des élèves, utilisé pour effectuer une conversion de données sur les données des élèves à prédire, en saisissant les données des élèves à prédire dans le modèle bayésien naïf formé, pour effectuer une prédiction de la classification des données des élèves afin d'obtenir un résultat de prédiction de la réussite académique des élèves , et en affichant le résultat de la prédiction de la réussite académique des élèves par le biais d'une unité d'affichage.a naive Bayesian modeling module, used for the standardized table of student learning status data, the calculation of the conditional probabilities of the characteristics in different categories and the prior probabilities of the different categories by via a computing unit in the server to obtain the naive Bayes model; and a module for predicting students' academic success, used to perform a data conversion on the student data to be predicted, by capturing the student data to be predicted in the naive Bayes model formed, to predict the classification of the students. students' data in order to obtain a prediction result of students' academic success, and by displaying the result of the prediction of students' academic success through a display unit.
LU100314A 2016-06-20 2017-06-20 Method and system for predicting academic achievements of students based on naive bayesian model LU100314B1 (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610447484.3A CN106127634B (en) 2016-06-20 2016-06-20 Student academic achievement prediction method and system based on naive Bayes model

Publications (1)

Publication Number Publication Date
LU100314B1 true LU100314B1 (en) 2017-10-31

Family

ID=57471111

Family Applications (1)

Application Number Title Priority Date Filing Date
LU100314A LU100314B1 (en) 2016-06-20 2017-06-20 Method and system for predicting academic achievements of students based on naive bayesian model

Country Status (2)

Country Link
CN (1) CN106127634B (en)
LU (1) LU100314B1 (en)

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108241625B (en) * 2016-12-23 2021-10-01 科大讯飞股份有限公司 Method and system for predicting student achievement change trend
CN106951079A (en) * 2017-03-17 2017-07-14 北京康邦科技有限公司 A kind of self adaptation course control method and system
CN107274191A (en) * 2017-05-12 2017-10-20 南京邮电大学 A kind of shopping at network return of goods forecasting system based on seller
CN107274020B (en) * 2017-06-15 2020-05-22 北京师范大学 Learner subject total measured result prediction system and method based on collaborative filtering thought
CN108280531B (en) * 2017-07-28 2021-07-09 淮阴工学院 Student class score ranking prediction method based on Lasso regression
CN107967540A (en) * 2017-12-21 2018-04-27 中国海洋大学 Student's academic warning system and method
CN108228779B (en) * 2017-12-28 2021-03-23 华中师范大学 Score prediction method based on learning community conversation flow
CN109598995B (en) * 2019-01-08 2020-11-17 上海健坤教育科技有限公司 Intelligent teaching system based on Bayesian knowledge tracking model
CN111160444B (en) * 2019-12-25 2023-09-08 武汉美和易思数字科技有限公司 Test paper difficulty degree analysis method and system based on Bayesian principle
CN111428963B (en) * 2020-02-21 2023-12-19 贝壳技术有限公司 Data processing method and device
CN112396105B (en) * 2020-11-18 2023-11-07 沈阳航空航天大学 Intelligent generation method of flight training subjects based on Bayesian network

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050080663A1 (en) * 2003-10-08 2005-04-14 Kef.Software Ag Management tool
CN105677854A (en) * 2016-01-04 2016-06-15 安徽大学 Student single-subject grade decision analysis method based on rough set and analyzer

Also Published As

Publication number Publication date
CN106127634B (en) 2022-02-08
CN106127634A (en) 2016-11-16

Similar Documents

Publication Publication Date Title
LU100314B1 (en) Method and system for predicting academic achievements of students based on naive bayesian model
US10922991B2 (en) Cluster analysis of participant responses for test generation or teaching
Balahadia et al. Teacher's performance evaluation tool using opinion mining with sentiment analysis
Gottipati et al. A conceptual framework for analyzing students' feedback
Farhan et al. Multimedia based qualitative assessment methodology in eLearning: student teacher engagement analysis
US11449681B2 (en) Data processing system and method for dynamic assessment, classification, and delivery of adaptive personalized recommendations
CN109754349B (en) Intelligent teacher-student matching system for online education
Dewua et al. The accounting curriculum and the emergence of Big Data
CN111524578B (en) Psychological assessment device, method and system based on electronic psychological sand table
WO2022170985A1 (en) Exercise selection method and apparatus, and computer device and storage medium
Lauder et al. Pupil composition and accountability: An analysis in English primary schools
Prinsloo Context matters: An African perspective on institutionalizing learning analytics
Zhao et al. What are students thinking and feeling? Understanding them from social data mining
CN115081965B (en) Big data analysis system of condition of learning and condition of learning server
CN111274779A (en) Legal document generation system and method based on user experience prediction
Davies et al. Choosing in schools: locating the benefits of specialisation
JP5628151B2 (en) Optimal learning item automatic generation system and method for controlling the system
CN113822774A (en) Financial teaching system, using method thereof and computer readable storage medium
Singh et al. Concepts of data mining and process mining
CN111105138A (en) Human resource analysis and evaluation system based on task completion data
Yan Conceptualization and operationalization of accountability in Canadian education: A systematic review
Yu et al. College student management system based on k-means clustering algorithm
Oyekunle et al. Sentiment Analysis of Students' Tweets on Unilorin CBT Examinations
Widyahastuti The relationship between online tutorials and academic performance in distance education: a predictive framework for Open University, Indonesia
Chen et al. Application and Research of Media Information Retrieval Technology in Student Work Management Systems

Legal Events

Date Code Title Description
FG Patent granted

Effective date: 20171031

PD Change of ownership

Owner name: SHANDONG EDUCATION EQUIPMENT CENTER CO., LTD.; CN

Free format text: FORMER OWNER: SHANDONG NORMAL UNIVERSITY

Effective date: 20210608