KR102304683B1 - Learning management system - Google Patents

Abstract

A learning management system according to an embodiment of the present invention includes: an academic achievement calculation module for calculating an academic achievement based on a student's academic achievement test details; a learning time measuring module for measuring the actual learning time of the student; and an academic grade calculation module for calculating the student's academic grade based on the first characteristic data including the student's academic achievement and learning time.

Description

Learning management system {LEARNING MANAGEMENT SYSTEM}

The present invention relates to a learning management system, and to a learning management system capable of calculating a student's academic grade based on the student's academic achievement and learning time, and managing the student based on the calculated academic grade.

For candidates preparing for the exam, an effective exam strategy should be established in order to obtain effective grades within a limited time.

However, most of the test takers tend to establish the test strategy emotionally rather than rationally. There is a limit to

Therefore, there is a need for a sophisticated and specific learning management system to establish effective test strategies for students.

On the other hand, the following prior art document discloses only the contents of the school management and textbook ordering system using a wireless terminal, but does not disclose the technical gist of the present invention.

Korean Patent Publication No. 2003-0075065

The learning management system according to an embodiment of the present invention aims to solve the following problems in order to solve the above problems.

It is to provide a learning management system that can calculate the academic grade for each student in consideration of the academic achievement and study time for each student, and establish a customized examination strategy according to the corresponding academic grade.

The problems to be solved of the present invention are not limited to those mentioned above, and other problems not mentioned will be clearly understood by those of ordinary skill in the art from the following description.

A learning management system according to an embodiment of the present invention includes: an academic achievement calculation module for calculating an academic achievement based on a student's academic achievement test details; a learning time measuring module for measuring the actual learning time of the student; and an academic grade calculation module for calculating the student's academic grade based on first characteristic data including the student's academic achievement and learning time; and calculates the academic achievement based on the academic achievement test score, and the academic grade calculation module uses an algorithm set based on the correlation between the academic achievement and learning time and the academic grade to the student's academic grade to decide

A plurality of students' incorrect answer notes are stored, and a test generation module for generating the test question based on the academic grade and the problem described in the incorrect answer note; further comprising, the test generation module, written by a plurality of students Incorrect note storage unit in which the incorrect notes are stored; a problem extracting unit for extracting customized problems for the student from among the problems listed in the incorrect note storage unit in consideration of the academic grade; a problem arranging unit for randomly arranging the problems extracted by the problem extracting unit to generate a test; and a reward unit for giving a predetermined point to a student who provides a problem constituting the test.

The algorithm is implemented as a deep learning algorithm composed of a neural network, wherein the neural network includes: an input layer for receiving the first characteristic data; at least one hidden layer configured to receive the first feature data from the input layer and include a plurality of nodes; and an output layer that determines the student's academic grade based on the operation result of the hidden layer, wherein a function that is a correlation between an input factor transmitted to the node and an output factor output from the node is defined in the node, The function preferably includes a weight and a bias.

A big data storage module storing big data including the number of times of taking the academic achievement test of a plurality of students, the academic achievement test scores, learning time, and actual entrance exam scores; further comprising, wherein the weights and biases are stored in the big data storage module It is preferable to use big data as training data to determine it through machine learning performed on a neural network.

Further comprising; a training data generation module for generating training data for determining the weight and bias, wherein the training data generation module receives the first characteristic data and receives the first characteristic data and based on a previously learned discrimination criterion, the student's a judging unit that determines the student's academic grade; and a generator for generating second feature data corresponding to the first feature data, wherein the discriminator generates the discrimination criterion through learning using a generative adversarial network (GAN), and The discrimination unit is trained to distinguish the first characteristic data and the second characteristic data, and the generating device is trained so that the discrimination device cannot distinguish the first characteristic data and the second characteristic data to generate the second characteristic data. It is preferable to do

The learning time measuring module measures the actual study time of the student by the study time measuring device, wherein the study time measuring device includes: a transfer rail formed on one side of the desk used by the student; a seating member formed to move in a left and right direction along the transfer rail, the seating member being rotatable up and down based on the transfer rail; and a terminal formed to be seated on the seating member and having a seating sensor and a camera, wherein the seating member includes a seating part for seating the terminal, and a first for moving the seating member along the transfer rail It is preferable to include a driving unit and a second driving unit for rotating the seating member up and down, and when the terminal is mounted on the seating part, the terminal starts counting the learning time of the student.

The seating part drives the first driving unit and the second driving unit so that the camera can detect the pupil's eyes, and the camera is driven by the pupil despite the driving of the first driving unit and the second driving unit. If the eye is not detected, the terminal preferably stops counting the learning time.

The camera tracks the blinking and gaze state of the student's eyes, and the terminal indicates that the blinking of the student's eyes does not last for more than a preset time or the student's gaze exceeds a preset time in a preset study area (Study Area) It is preferable to stop counting the learning time when it is determined that it is out of range.

The learning management system according to an embodiment of the present invention can calculate a student's academic grade based on the student's academic achievement and learning time, and establish a customized learning strategy for the student in consideration of the calculated academic grade. It works.

Furthermore, it is possible to establish a more sophisticated and effective customized learning strategy by calculating sophisticated weights for academic achievement and learning time when calculating academic grades.

Effects of the present invention are not limited to those mentioned above, and other effects not mentioned will be clearly understood by those of ordinary skill in the art from the following description.

1 is a block diagram schematically illustrating a learning management system according to an embodiment of the present invention.
2 is a diagram schematically illustrating a neural network used in a learning management system according to an embodiment of the present invention.
3 and 4 are diagrams for explaining a training data generation module in the learning management system according to an embodiment of the present invention.
5 is a block diagram illustrating a specific configuration of a test generation module among configurations of a learning management system according to an embodiment of the present invention.
6 is a diagram schematically illustrating a study time measuring device applied to a learning time measuring module among the configuration of a learning management system according to an embodiment of the present invention.

A preferred embodiment according to the present invention will be described in detail with reference to the accompanying drawings, but the same or similar components are assigned the same reference numerals regardless of reference numerals, and overlapping descriptions thereof will be omitted.

In addition, in the description of the present invention, if it is determined that a detailed description of a related known technology may obscure the gist of the present invention, the detailed description thereof will be omitted. In addition, it should be noted that the accompanying drawings are only for easy understanding of the spirit of the present invention, and should not be construed as limiting the spirit of the present invention by the accompanying drawings.

As shown in FIG. 1, the learning management system according to an embodiment of the present invention includes an academic achievement calculation module 100, a learning time measurement module 200, an academic grade calculation module 300, a test generation module 400, It is configured to include a big data storage module 500 and a training data generation module 600 .

The academic achievement calculation module 100 performs a function of calculating the academic achievement based on the student's academic achievement test details. In particular, the academic achievement calculation module 100 includes the student's academic achievement test scores, for example, affiliated school and surrounding areas. The academic achievement is calculated by considering the relative evaluation score considering the hagwon.

In addition, it is preferable that the academic achievement calculation module 100 calculates the student's academic achievement by considering not only the above-described student's academic achievement test score but also the number of times taken in the academic achievement test.

On the other hand, the above-described test may be configured based on the problems described in a plurality of incorrect answer notes written by students, and the above-described test generating module 400 performs this role.

Specifically, the test generation module 400 stores a plurality of student's incorrect answer notes, and generates a test question based on the calculated student's academic grade and the problem described in the incorrect answer note.

For example, if student A's academic grade is higher than that of student B, student B can identify their weaknesses by encountering the problems included in student A's incorrect note, and furthermore, student B's ability improves Even if it is upgraded to Student A's academic grade, it is highly likely that the questions included in Student A's incorrect note will be wrong again, so it is desirable for Student B to refer to Student A's incorrect note.

The test generating module 400 includes an incorrect answer note storage unit 410, a problem extracting unit 420, a problem arranging unit 430, and a compensating unit 440 as shown in FIG. 5 to perform the above-described functions. is configured to include

The incorrect note storage unit 410 performs a function of storing the incorrect notes written by a plurality of students, and the problem extraction unit 420 considers the student's academic grade among the problems described in the incorrect note storage unit 410. It performs the function of extracting customized problems for the student.

The problem arranging unit 430 performs a function of generating a test by randomly arranging the problems extracted by the problem extracting unit 420, and the compensation unit 440 provides a predetermined point to a student who provides a problem constituting the test, etc. It performs the function of giving

The learning time measurement module 200 performs a function for grasping the student's attendance and attendance, and in order to calculate an accurate academic grade, it is preferable to measure the actual learning time of the student.

To this end, the learning time measuring module 200 of the learning management system according to an embodiment of the present invention is configured to measure the student's actual study time by the study time measuring device 700, and details thereof will be described later. .

The academic grade calculation module 300 performs a function of calculating the student's academic grade based on the first characteristic data including the student's academic achievement and actual learning time.

The academic grade calculation module 300 determines the student's academic grade using an algorithm set based on the correlation between the input factor, academic achievement and learning time, and the output factor, academic grade, at this time, the academic achievement and learning time. It is important to precisely determine each weight for the reflection of

For this purpose, the algorithm used in the academic grade calculation module 300 according to an embodiment of the present invention is preferably implemented as a deep learning algorithm composed of the neural network 10 .

The neural network 10 is configured to include an input layer 11 , a hidden layer 12 , and an output layer 13 as shown in FIG. 2 .

The input layer 11 performs a function of receiving the above-described first characteristic data, that is, an input variable, and the hidden layer 12 receives the first characteristic data from the input layer 11 and performs an operation.

The hidden layer 12 is preferably composed of a plurality of layers, and each hidden layer 12 includes a plurality of nodes as shown in FIG. 2 .

A linear or nonlinear function that is a correlation between an input factor transmitted to a corresponding node and an output factor output from the node is defined at each node, and the function is configured to include a weight and a bias.

For example, when the input factor at the node is defined as x and the output factor at the node is defined as y, the function defined at the node can be defined as “y=w*x+b” and , where w denotes a weight, and b denotes a bias.

In the case of such a deep learning algorithm, it is possible to generate an optimal weight and bias through machine learning, and use this to calculate an accurate academic grade.

The output layer 13 performs a function of finally determining the student's academic grade based on the operation result of the hidden layer.

On the other hand, the big data storage module 500 in the learning management system according to an embodiment of the present invention stores big data including the number of times a plurality of students take an academic achievement test, academic achievement test scores, learning time, and actual entrance exam scores. Save function.

That is, the weight and bias of each node included in the above-described neural network 10 is machine learning performed on the neural network 10 by using the big data stored in the big data storage module 500 as training data. is determined through

On the other hand, what is required in machine learning is to perform learning using a lot of data above all else.

Deep learning is a technology based on such learning, and there are convolutional neural networks (CNNs) and recurrent neural networks (RNNs) as supervised learning methods based on numerous data.

However, in the case of the supervised learning method described above, it is not easy to apply because it requires information data for numerous correct answers. Otherwise, there is a problem that the calculation of sophisticated weights and biases may be difficult because machine learning is not performed properly.

Therefore, interest in unsupervised learning technology that proceeds without labeling data is increasing, and in order to overcome the above-mentioned problems, the learning management system according to an embodiment of the present invention provides real data without providing information on correct answers. Through learning using a generative adversarial network (GAN), which is a generating technology, it is possible to calculate academic achievement and determine classes.

In the case of such a generative adversarial network, in order to generate data and make it indistinguishable from the real data, a role of making the data a imitation (generator) and a role of distinguishing whether the data is a counterfeit or not (discriminator) must be given.

Looking at this generative adversarial network in more detail, as shown in Fig. 3, it is composed of a generator (1), a database (DB) (2), and a discriminator (3). The generator (1) and the discriminator (2) are networks composed of neural networks.

The discriminator 3 receives data from the database 2 or the generator 1, and learns to discriminate between the data (data_real) provided by the database 2 and the data (data_fake) provided from the generator 10 among these data. do.

On the other hand, the generator 1 learns that the data data_fake provided to the discriminator 3 is discriminated as data_real, that is, the generator 1 learns that the discriminator 3 determines data_fake as data_real.

Summarizing the learning directions of the two neural network learners (discriminator and generator), the discriminator 3 learns to distinguish the real data from the data generated by the generator 1, and the generator 1 is the discriminator 3 learns to generate indistinguishable data.

By repeated learning as described above, the generator 1 can generate similar data to the actual data stored in the database 2, and thereby the discriminator 1 generates a large amount of training data (Training Data) This makes it possible to produce sophisticated results.

The learning management system according to an embodiment of the present invention may further include a training data generation module 600 that generates training data for determining the above-described weights and biases, and the training data generation module 600 generates An adversarial network is applied, which will be described in more detail with reference to FIG. 4 .

As shown in FIG. 4, the training data generating module 600 receives random noise data (Input_Z) to the generator 620, and initially noise data (Input_Z) for faster and faster reinforcement learning. and the conventional first characteristic data (Real_Data) stored in the big data storage module 500 may be combined and input to the generation unit 620, and through this, the generation unit 620 may generate a plurality of diversified second characteristic data. You can create feature data.

The determining unit 610 may be the above-described neural network 10, and receives the input first characteristic data (Real_Data) and the second characteristic data (Fake_Data) and determines the authenticity of the received Fake/Real results. is calculated, and the result is provided to the generator 620 so that the generator 620 can learn.

That is, the generating unit 620 improves the neural network so that the determining unit 610 determines the real data based on the authenticity determination of the determining unit 610 , and as a result, as the learning of the generating unit 620 continues, the generating unit 620 ) is to improve the neural network so that the determining unit 610 determines that it is real data.

To summarize the above, the determining unit 610 is trained to distinguish the first characteristic data and the second characteristic data, and the generating unit 620 is the determining unit 610 to distinguish the first characteristic data and the second characteristic data. It is learned so as not to be able to generate the second characteristic data.

Through the process as described above, the determination unit 610 can perform academic achievement calculation and class determination more accurately, and through reinforcement learning using the generative confrontation network of the determination unit 610 and the generation unit 620 . After sufficient learning of the determining unit 610 is completed, the generating unit 620 may be removed.

On the other hand, as described above, in calculating the student's academic grade, the student's actual learning time should be measured, and a study time measuring device 700 for measuring the student's actual study time may be provided.

This study time measuring device 700 is configured to include a transfer rail 720, a seating member 730 and a terminal 740, as shown in FIG.

The transfer rail 720 is configured to be formed on one side of the desk 710 used by the student, and the seating member 730 can move in the left and right directions on the transfer rail 720, and further rotates in the vertical direction. it is preferable

The terminal 740 is a personal management terminal in which an individual student ID is stored, and is configured to include a display, a camera, and a seating sensor.

In particular, the seating member 730 is provided with a seating portion for seating the terminal 740, and further includes a first driving unit that moves the seating member 730 in the left and right directions along the transfer rail 720, and the seating member 730. It includes a second driving unit for rotating up and down.

When explaining the specific operation of the above-described study time measuring device, first, when the student seats his/her terminal 740 on the seating part of the seating member 730, the terminal 740 by the seating sensor of the lower part of the terminal 740. It is confirmed whether or not the terminal 740 is seated, and therefore the terminal 740 starts counting the learning time of the corresponding student.

Thereafter, the camera of the terminal 740 continuously detects the image in the direction of the student, and communicates with the seating member 730 to obtain an image of the student's eyes to move the seating member 730 . and rotation control.

Eventually, the terminal 740 controls the first and second driving units of the seating member 730 based on the image acquired by the camera. When the camera of the terminal 740 first detects the pupil's eye image, Thereafter, the camera of the terminal 740 continuously tracks the pupil's eyes, and continuously controls the first driving unit and the second driving unit based on the tracking.

On the other hand, when the camera of the terminal 740 does not detect the pupil's eyes despite the driving of the first driving unit and the second driving unit by the terminal 740, the pupil has left the desk or is near the desk. However, it is determined that the study is not in progress, and the terminal 740 stops counting the learning time.

In addition, the camera of the terminal 740 tracks the blinking and gaze state of the target student's eyes, and when the blinking of the student's eyes does not last for more than a preset time, it is determined that the student is in a drowsy state, and counting of the learning time is stopped.

In addition, the terminal 740 stops counting the learning time even when the student's gaze exceeds the preset time and deviates from the preset study area.

For example, if the student's gaze is directed toward the blackboard or on the desk, the counting of the learning time continues, but if the student's gaze is directed toward an area other than the blackboard and the desk, it means that the student is not studying. It is possible to stop counting the learning time by judging.

The embodiments described in this specification and the accompanying drawings are merely illustrative of some of the technical ideas included in the present invention. Therefore, since the embodiments disclosed in the present specification are for explanation rather than limitation of the technical spirit of the present invention, it is obvious that the scope of the technical spirit of the present invention is not limited by these embodiments. Modifications and specific embodiments that can be easily inferred by a person of ordinary skill in the art within the scope of the technical idea included in the specification and drawings of the present invention are included in the scope of the present invention. will have to be interpreted.

100: academic achievement calculation module 200: learning time measurement module
300: academic grade calculation module 400: test generation module
500: big data storage module 600: training data generation module
700: study time measuring device

Claims (6)

an academic achievement calculation module that calculates academic achievement based on the student's academic achievement test details;
a learning time measuring module for measuring the actual learning time of the student; and
an academic grade calculation module for calculating the student's academic grade based on first characteristic data including the student's academic achievement and learning time;
including,
The academic achievement calculation module calculates the academic achievement based on the number of times the student takes the academic achievement test and the academic achievement test score,
The academic grade calculation module determines the academic grade of the student using an algorithm set based on the correlation between the academic achievement and learning time and the academic grade,
A plurality of students' incorrect answer notes are stored, and a test generation module for generating a problem of the test based on the academic grade and the problem described in the incorrect note note; further comprising, the test generating module,
an incorrect note storage unit in which the incorrect answer notes written by a plurality of students are stored;
a problem extracting unit for extracting customized problems for the student from among the problems listed in the incorrect note storage unit in consideration of the academic grade;
a problem arranging unit for randomly arranging the problems extracted by the problem extracting unit to generate a test; and
a reward unit for giving a predetermined point to a student who provided a problem constituting the test; including,
The algorithm is implemented as a deep learning algorithm composed of a neural network, wherein the neural network is
an input layer receiving the first characteristic data;
at least one hidden layer configured to receive the first feature data from the input layer and include a plurality of nodes; and
an output layer for determining the student's academic grade based on the operation result of the hidden layer; including,
A function that is a correlation between an input factor transmitted to the node and an output factor output from the node is defined in the node, and the function includes a weight and a bias;
A big data storage module that stores big data including the number of times a plurality of students take the academic achievement test, the academic achievement test scores, learning time, and actual entrance exam scores; further comprising,
The weight and bias are determined through machine learning performed on a neural network using the big data stored in the big data storage module as training data,
Further comprising; a training data generation module for generating training data for determining the weight and bias;
The training data generation module,
a determination unit for receiving the first characteristic data and determining the student's academic grade of the student based on a previously learned determination criterion; and
and a generator configured to generate second feature data corresponding to the first feature data;
The discrimination unit generates the discrimination criterion through learning using a generative adversarial network (GAN), the discrimination unit is trained to distinguish the first characteristic data and the second characteristic data, and the generation unit is the A learning management system for generating the second characteristic data by learning so that the determining unit cannot distinguish the first characteristic data and the second characteristic data.
delete delete delete The method according to claim 1,
The learning time measuring module measures the actual study time of the student by the study time measuring device, the study time measuring device,
a transfer rail formed on one side of the desk used by the student;
a seating member formed to move in a left and right direction along the transfer rail, the seating member being rotatable up and down based on the transfer rail; and
a terminal formed to be seated on the seating member and having a seating sensor and a camera;
including,
The seating member includes a seating part for seating the terminal, a first driving unit for moving the seating member along the transfer rail, and a second driving unit for rotating the seating member up and down,
When the terminal is seated in the seating unit, the terminal starts the counting of the student's learning time, the learning management system.
6. The method of claim 5,
The seating unit drives the first driving unit and the second driving unit so that the camera can detect the pupil's eyes,
In spite of the driving of the first driving unit and the second driving unit, if the camera does not detect the eyes of the student, the terminal stops counting the learning time.

Images