Classroom teaching quality evaluation method and system based on big data

Abstract

The invention relates to the technical field of education informatization, and discloses a classroom teaching quality evaluation method and system based on big data, and the method comprises the steps: obtaining the eye watching orientation information of a student, the face direction information of the student, the classroom problem information and the total classroom number; obtaining a concentration degree parameter according to the eye gazing azimuth information and the face direction information; according to the concentration degree parameter, a particle swarm optimization algorithm is adopted for optimization to obtain an optimal concentration degree coefficient; training by adopting a random forest algorithm according to the classroom question information to obtain a correct rate coefficient and a question proportion coefficient; according to the correct rate coefficient and the problem proportion coefficient, calculating to obtain a normalized correct rate coefficient and a normalized problem proportion coefficient; according to the total classroom number, the optimal concentration degree coefficient, the classroom problem information, the normalized correct rate coefficient and the problem proportion coefficient, classroom quality evaluation parameters are obtained through calculation; and finally, matching with a preset evaluation threshold parameter to obtain a classroom teaching quality evaluation result. According to the method, the classroom teaching quality can be correctly evaluated.

Description

Technical Field

[0001] The present invention relates to the technical field of educational informatization, and particularly to a method and system for evaluating the quality of classroom teaching based on big data. Background Art

[0002] With the application and research of big data and artificial intelligence technologies in the field of education, in the existing evaluation methods of classroom teaching, by detecting the effective target faces matched in the video images of learners and the video images of teaching, the learning time of each effective target face is obtained and recorded as the concentration time, and finally the teaching effect coefficient is obtained by calculating the ratio of the concentration time to the classroom teaching time, and the teaching quality is reflected by this teaching effect coefficient. In this teaching effect evaluation method, since it is impossible to determine whether the students are interested in the teaching pictures, or interested in other things, or only interested in a certain image and other factors, the error rate of this teaching effect evaluation and judgment is relatively high.

[0003] In summary, in the prior art, only measuring the teaching quality by concentration has a relatively high error rate of judgment and cannot correctly evaluate the quality of classroom teaching. Summary of the Invention

[0004] The present invention provides a method and system for evaluating the quality of classroom teaching based on big data to solve the problem that in the existing evaluation methods of classroom teaching, only measuring the teaching quality by concentration has a relatively high error rate of judgment and cannot correctly evaluate the quality of classroom teaching.

[0005] In a first aspect, to solve the above technical problem, the present invention provides a method for evaluating the quality of classroom teaching based on big data, including:

[0006] Obtaining the eye gaze direction information of students, the facial direction information of students, classroom question information, and the total number of classes;

[0007] Determining the concentration based on the eye gaze direction information and the facial direction information to obtain a concentration parameter;

[0008] Constructing a concentration data set according to the concentration parameter and performing optimization using a particle swarm optimization algorithm to obtain an optimal concentration coefficient;

[0009] Constructing a classroom information data set according to the classroom question information and performing training using a random forest algorithm to obtain a correct rate coefficient and a question proportion coefficient;

[0010] Performing normalization calculation according to the correct rate coefficient and the question proportion coefficient to obtain a normalized correct rate coefficient and a normalized question proportion coefficient;

[0011] Calculate the classroom quality parameter based on the total number of classes, the optimal concentration coefficient, the classroom question information, the normalized correct rate coefficient, and the normalized question proportion coefficient to obtain the classroom quality evaluation parameter;

[0012] Match the classroom quality evaluation parameter with the preset evaluation threshold parameter to obtain the classroom teaching quality evaluation result.

[0013] In an optional implementation manner, the eye gaze direction information includes gazing at the teacher's face and gazing elsewhere;

[0014] The face direction information includes the face facing forward and the face facing sideways;

[0015] The classroom question information includes the number of questions actively answered by students, the number of questions passively answered by students, and the number of questions correctly answered by students.

[0016] In an optional implementation manner, the determining the concentration parameter based on the eye gaze direction information and the face direction information includes:

[0017] When the face direction information is the face facing forward and the eye gaze direction information is gazing at the teacher's face, it is determined as the first concentration parameter;

[0018] When the face direction information is the face facing sideways and the eye gaze direction information is gazing elsewhere, it is determined as the second concentration parameter;

[0019] When the face direction information is the face facing forward and the eye gaze direction information is gazing elsewhere, it is determined as the third concentration parameter.

[0020] In an optional implementation manner, the constructing the concentration data set based on the concentration parameter and using the particle swarm optimization algorithm for optimization to obtain the optimal concentration coefficient includes:

[0021] Perform data cleaning and preprocessing on the concentration data set to obtain the concentration training set data;

[0022] Establish a velocity set and a position set based on the concentration training set data, and initialize the data of the velocity set and the position set to obtain the initial velocity data and the initial position data;

[0023] Perform parameter optimization using the particle swarm optimization algorithm based on the initial velocity data and the initial position data to obtain the optimal concentration coefficient;

[0024] Among them, the optimal concentration coefficient includes the first optimal parameter, the second optimal parameter, and the third optimal parameter.

[0025] In an alternative embodiment, constructing a classroom information data set according to the classroom question information and training using a random forest algorithm to obtain a correct rate parameter and a question proportion parameter includes:

[0026] Performing data cleaning and preprocessing on the classroom information data set to obtain classroom information training set data;

[0027] Constructing a decision tree based on the classroom information training set data to obtain a decision tree data set;

[0028] Performing model training using a random forest algorithm on the decision tree data set to obtain a correct rate coefficient and a question proportion coefficient.

[0029] In an alternative embodiment, performing model training using a random forest algorithm on the decision tree data set to obtain a correct rate coefficient and a question proportion coefficient includes:

[0030] Dividing the decision tree data set into decision tree training set data and decision tree test set data;

[0031] Performing random sampling on the decision tree training set data to obtain decision tree sampling set data;

[0032] Calculating the root mean square error of the decision tree based on the decision tree sampling set data and the decision tree test set data to obtain the root mean square error of the decision tree;

[0033] When the root mean square error of the decision tree is less than a preset error threshold, performing a majority voting operation on the decision tree sampling set data to obtain a correct rate coefficient and a question proportion coefficient.

[0034] In an alternative embodiment, performing normalization calculation based on the correct rate coefficient and the question proportion coefficient to obtain a normalized correct rate coefficient and a normalized question proportion coefficient includes:

[0035] The normalized correct rate coefficient is calculated by the following formula:

[0036]

[0037] In the formula, θ′ is the normalized correct rate coefficient, θ is the correct rate coefficient, μ is the question proportion coefficient, and λ is a preset focus parameter;

[0038] The normalized question proportion coefficient is calculated by the following formula:

[0039]

[0040] In the formula, μ′ is the normalized question proportion coefficient.

[0041] In an alternative embodiment, calculating a classroom quality parameter based on the total number of classes, the optimal concentration coefficient, the classroom question information, the normalized correct rate coefficient, and the normalized question proportion coefficient to obtain a classroom quality evaluation parameter, including:

[0042] The classroom quality evaluation parameter is calculated by the following formula:

[0043]

[0044] In the formula, Q is the classroom quality evaluation parameter, θ′ is the normalized correct rate coefficient, I is the number of students who actively answer questions, P is the number of students who passively answer questions, R is the number of students who answer questions correctly, μ′ is the normalized question proportion coefficient, M is the total number of classes, α is the first optimal parameter, β is the second optimal parameter, γ is the third optimal parameter, A is the first concentration parameter, B is the second concentration parameter, C is the third concentration parameter, and λ is the preset concentration parameter.

[0045] In an alternative embodiment, matching the classroom quality evaluation parameter with a preset evaluation threshold parameter to obtain a classroom teaching quality evaluation result, including:

[0046] When Q > Q1, the classroom teaching quality evaluation result is excellent;

[0047] When Q2 < Q ≤ Q1, the classroom teaching quality evaluation result is good;

[0048] When Q3 < Q ≤ Q2, the classroom teaching quality evaluation result is passing;

[0049] When Q ≤ Q3, the classroom teaching quality evaluation result is failing;

[0050] Wherein, Q is the classroom quality evaluation parameter, and Q1, Q2, and Q3 are respectively the preset first evaluation threshold, second evaluation threshold, and third evaluation threshold.

[0051] In a second aspect, the present invention provides a classroom teaching quality evaluation system based on big data, including:

[0052] A data acquisition module for acquiring students' eye gaze orientation information, students' facial direction information, classroom question information, and the total number of classes;

[0053] A concentration determination module for determining the concentration based on the eye gaze orientation information and the facial direction information to obtain a concentration parameter;

[0054] A concentration optimization module for constructing a concentration data set based on the concentration parameter and performing optimization using a particle swarm optimization algorithm to obtain an optimal concentration coefficient;

[0055] A classroom information training module, configured to construct a classroom information dataset according to the classroom question information, and perform training using a random forest algorithm to obtain a correct rate coefficient and a question proportion coefficient;

[0056] A normalization calculation module, configured to perform normalization calculation according to the correct rate coefficient and the question proportion coefficient to obtain a normalized correct rate coefficient and a normalized question proportion coefficient;

[0057] An evaluation parameter calculation module, configured to calculate classroom quality parameters according to the total number of classes, the optimal concentration coefficient, the classroom question information, the normalized correct rate coefficient, and the normalized question proportion coefficient to obtain classroom quality evaluation parameters;

[0058] An evaluation result matching module, configured to match the classroom quality evaluation parameters with preset evaluation threshold parameters to obtain a classroom teaching quality evaluation result.

[0059] In a third aspect, the present invention further provides an electronic device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a method for evaluating classroom teaching quality based on big data as described in any one of the above.

[0060] In a fourth aspect, the present invention further provides a computer-readable storage medium, which includes a stored computer program. When the computer program runs, it controls the device where the computer-readable storage medium is located to execute a method for evaluating classroom teaching quality based on big data as described in any one of the above.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] (1) The method for evaluating classroom teaching quality of the present invention collects students' eye gaze direction information, students' facial direction information, classroom question information, and total number of classroom information. By combining eye gaze direction with facial direction to judge students' concentration, it can more objectively and comprehensively judge whether students are interested in the teaching picture; at the same time, the correctness of students' answers and the activity of active answers are placed under the same standard, and these two aspects are considered comprehensively to obtain a more accurate evaluation of classroom performance; compared with the prior art, it can complete the evaluation of teaching according to multiple factors and correctly evaluate the quality of classroom teaching.

[0063] (2) The classroom teaching quality evaluation method of the present invention is based on big data, uses the particle swarm optimization algorithm for optimization, and adopts the random forest algorithm to establish decision trees and predict results, achieving a more accurate intelligent evaluation of teaching quality. Accordingly, a classroom teaching quality evaluation system based on big data is provided, which provides great assistance to the modern teaching evaluation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 is a schematic flow chart of a classroom teaching quality evaluation method based on big data provided by the first embodiment of the present invention;

[0065] Figure 2 is a schematic structural diagram of a classroom teaching quality evaluation system based on big data provided by the second embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0066] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0067] Refer to Figure 1 , the first embodiment of the present invention provides a classroom teaching quality evaluation method based on big data, including the following steps:

[0068] S11, obtaining the eye gaze direction information of students, the facial direction information of students, classroom question information, and the total number of classes;

[0069] S12, determining the concentration degree according to the eye gaze direction information and the facial direction information to obtain a concentration parameter;

[0070] S13, constructing a concentration data set according to the concentration parameter, and using the particle swarm optimization algorithm for optimization to obtain an optimal concentration coefficient;

[0071] S14, constructing a classroom information data set according to the classroom question information, and using the random forest algorithm for training to obtain a correct rate coefficient and a question proportion coefficient;

[0072] S15, performing normalization calculation according to the correct rate coefficient and the question proportion coefficient to obtain a normalized correct rate coefficient and a normalized question proportion coefficient;

[0073] S16. Calculate the classroom quality parameter based on the total number of classes, the optimal concentration coefficient, the classroom question information, the normalized correct rate coefficient, and the normalized question proportion coefficient to obtain the classroom quality evaluation parameter.

[0074] S17. Match the classroom quality evaluation parameter with the preset evaluation threshold parameter to obtain the classroom teaching quality evaluation result.

[0075] In step S11, obtain the student's eye gaze direction information, the student's face direction information, the classroom question information, and the total number of classes.

[0076] The eye gaze direction information includes gazing at the teacher's face and gazing elsewhere.

[0077] The face direction information includes facing forward and facing sideways.

[0078] The classroom question information includes the number of questions the student actively answers, the number of questions the student is passively asked to answer, and the number of questions the student answers correctly.

[0079] It should be noted that for obtaining the student's eye gaze direction information, it can be obtained through eye tracking technology. For obtaining the student's face direction information, it can be obtained through face recognition and pose estimation technology. For obtaining information such as the number of questions the student actively answers, the number of questions the student is passively asked to answer, and the number of questions the student answers correctly, it can be obtained through speech recognition and natural language processing technology. For obtaining the total number of class information, it can be directly obtained through the school's educational administration system or course schedule.

[0080] In one implementation, obtain the eye gaze direction information through eye tracking technology. First, install eye tracking devices near the podium, blackboard, or ceiling in the classroom. These devices are cameras integrated with eye tracking functions and can accurately capture the student's eye fixation points and line-of-sight directions, thereby collecting the student's line-of-sight data in real time. For example, the device will record the pixel coordinates and duration of the student's eyes gazing at the teacher's face, as well as the pixel coordinates and duration of gazing elsewhere (such as textbooks, blackboards, classmates, etc.). Analyze the collected line-of-sight data to calculate the proportion of the time the student gazes at the teacher's face and the time the student gazes elsewhere. For example, by analyzing student Xiaoming in a 45-minute math class, the time he gazes at the teacher's face is 36 minutes, and the time he gazes elsewhere is 9 minutes. It can be concluded that the proportion of Xiaoming gazing at the teacher's face is 80%, and the proportion of gazing elsewhere is 20%.

[0081] In one implementation, the facial orientation information of students is obtained through a facial recognition algorithm. First, high-definition cameras are installed at multiple angles in the classroom to ensure that all student positions can be covered. The resolution and frame rate of the cameras should be high enough to clearly capture the facial details and dynamic changes of students. Then, using a facial recognition algorithm, such as a convolutional neural network (CNN) based on deep learning, the faces of students are recognized from the videos captured by the cameras. Through a pose estimation model, such as a model based on key point detection, the facial orientation of students is analyzed, the key points of the face (such as eyes, nose, mouth, etc.) are recognized, and based on the positional relationship of these key points, it is determined whether the face of the student is facing the teacher directly or facing other classmates sideways. Finally, the time proportion of different facial orientations of students in the classroom is counted. For example, in a 40-minute English class, the facial recognition system recognizes that the time when student Xiaohong's face is facing the teacher directly is 28 minutes, and the time when it is facing other classmates sideways is 12 minutes. It can be concluded that the proportion of Xiaohong's face facing the teacher directly is 70%, and the proportion of facing other classmates sideways is 30%.

[0082] In one implementation, the classroom question information is obtained through speech recognition and natural language processing technologies. First, microphone arrays are installed in all corners of the classroom to ensure that the voices of teachers and students can be clearly captured. The microphone arrays can suppress noise and improve the quality of speech signals. The collected speech signals are converted into text data through speech recognition technology. The speech recognition model can recognize the words and sentences in the speech and generate the corresponding text. Using natural language processing technologies, such as text classification, keyword extraction, etc., relevant information about students' answers to questions is extracted from the text data. For example, by analyzing the keywords and sentence structures in the text, it is recognized which texts are those when students actively answer questions, such as the speeches when students raise their hands to answer questions actively; which texts are those when students passively answer questions, such as the speeches when the teacher calls on students to answer questions; and which answers are correct, such as the positive feedback given by the teacher after the answer. Finally, the number of times students actively answer questions, the number of times they passively answer questions, and the number of correct answers in the classroom are counted. For example, in a physics class, by analyzing the results of speech recognition and natural language processing, it is counted that student Xiaogang actively answers questions 3 times and passively answers questions 2 times, and the number of correct answers is 4 times.

[0083] In step S12, according to the eye gaze orientation information and the facial orientation information, the concentration is determined to obtain a concentration parameter.

[0084] When the facial orientation information is that the face is facing forward and the eye gaze orientation information is that the eyes are gazing at the teacher's face, it is determined as the first concentration parameter.

[0085] When the facial direction information is that the face is in a side view and the eye gaze direction information is that the eyes are looking elsewhere, it is determined as the second concentration parameter;

[0086] When the facial direction information is that the face is in a front view and the eye gaze direction information is that the eyes are looking elsewhere, it is determined as the third concentration parameter.

[0087] It should be noted that the concentration determination needs to be carried out in real time to timely understand the student's concentration state in class. The device and algorithm quickly process the collected data and output the concentration parameter in real time. For example, in a 45-minute class, the system should be able to update the concentration parameter every minute or every few minutes to reflect the student's concentration at different time periods, and finally comprehensively judge which specific concentration parameter it is.

[0088] Exemplarily, by judging the concentration parameter through segmented time, in a 45-minute math class, the concentration determination process of student Xiaohua is as follows:

[0089] Minutes 1 - 10:

[0090] Eye gaze direction information: The proportion of looking at the teacher's face is 80%, and the proportion of looking elsewhere is 20%; Facial direction information: The proportion of the face in a front view is 90%, and the proportion of the face in a side view is 10%. Concentration determination: Most of the time (80%) the face is in a front view and looking at the teacher's face, so it is determined as the first concentration parameter.

[0091] Minutes 11 - 20:

[0092] Eye gaze direction information: The proportion of looking at the teacher's face is 30%, and the proportion of looking elsewhere is 70%; Facial direction information: The proportion of the face in a front view is 50%, and the proportion of the face in a side view is 50%. Concentration determination: Half of the time (50%) the face is in a side view and looking elsewhere, so it is determined as the second concentration parameter.

[0093] Minutes 21 - 30:

[0094] Eye gaze direction information: The proportion of looking at the teacher's face is 60%, and the proportion of looking elsewhere is 40%; Facial direction information: The proportion of the face in a front view is 90%, and the proportion of the face in a side view is 10%. Concentration determination: Most of the time (90%) the face is in a front view, but for part of the time (40%) looking elsewhere, so it is determined as the third concentration parameter.

[0095] Minutes 31 - 40:

[0096] Eye gaze direction information: The proportion of looking at the teacher's face is 90%, and the proportion of looking elsewhere is 10%; Facial direction information: The proportion of the face in a front view is 100%, and the proportion of the face in a side view is 0%. Concentration determination: The whole process shows the face in a front view and looking at the teacher's face, so it is determined as the first concentration parameter.

[0097] Minute 41 - 45:

[0098] Eye gaze direction information: 50% of the time, the eyes are fixed on the teacher's face, and 50% of the time, they are fixed on other places. Facial direction information: 70% of the time, the face is facing forward, and 30% of the time, the face is turned sideways. For the determination of attentiveness, when the face is turned sideways and the eyes are fixed on other places for a certain period (30% of the time), it is determined as the second attentiveness parameter.

[0099] Comprehensive judgment:

[0100] The time percentage of the first attentiveness parameter appearing is: (10 minutes + 10 minutes) / 45 minutes ≈ 44.4%

[0101] The time percentage of the second attentiveness parameter appearing is: (10 minutes + 3 minutes) / 45 minutes ≈ 28.9%

[0102] The time percentage of the third attentiveness parameter appearing is: (10 minutes + 7 minutes) / 45 minutes ≈ 26.7%. Final comprehensive judgment: Xiaohua's attentiveness parameter in this class is the first attentiveness parameter.

[0103] In step S13, an attentiveness dataset is constructed based on the attentiveness parameter, and a particle swarm optimization algorithm is used for optimization to obtain the optimal attentiveness coefficient.

[0104] Based on the attentiveness dataset, data cleaning and preprocessing are performed to obtain the attentiveness training set data;

[0105] A velocity set and a position set are established based on the attentiveness training set data, and the data of the velocity set and the position set are initialized to obtain the initial velocity data and the initial position data;

[0106] Based on the initial velocity data and the initial position data, a particle swarm optimization algorithm is used for parameter optimization to obtain the optimal attentiveness coefficient;

[0107] Among them, the optimal attentiveness coefficient includes the first optimal parameter, the second optimal parameter, and the third optimal parameter.

[0108] It should be noted that the concentration dataset is cleaned by removing outliers, missing values, and duplicate values. For example, if the concentration parameter for a certain time period is missing, it is filled in by interpolation or using the values of adjacent time periods. The data is standardized or normalized so that data with different dimensions is under the same standard, and the values of all concentration parameters are converted to the range of 0 - 1. In the particle swarm optimization algorithm, each particle represents a set of concentration coefficients. The position set of the particle represents the values of this set of coefficients, and the velocity set represents the moving speed of the particle in the solution space. By initializing the velocity set and position set, that is, randomly generating a set of initial velocity data and position data. Through the iterative update of the positions and velocities of the first concentration parameter, the second concentration parameter, and the third concentration parameter by the particle swarm optimization algorithm, and performing multiple iterations until the preset accuracy range is met, the first optimal parameter, the second optimal parameter, and the third optimal parameter can be obtained.

[0109] Exemplarily, the initial position sets of the first optimal parameter, the second optimal parameter, and the third optimal parameter are [0.5, 0.3, 0.2], [0.6, 0.2, 0.2], and [0.4, 0.4, 0.2] respectively, and the initial velocity sets are [0.1, 0.1, 0.1], [0.1, 0.1, 0.1], and [0.1, 0.1, 0.1] respectively. Set the inertia weight ω = 7, the learning factor c1 = 2, c2 = 2, and the maximum number of iterations n = 100. Repeat the iterative process. After 100 iterations, the global optimal solution is [0.75, 0.15, 0.1]. Therefore, the first optimal parameter, the second optimal parameter, and the third optimal parameter are 0.75, 0.15, and 0.1 respectively.

[0110] In step S14, a classroom information dataset is constructed based on the classroom question information and trained using the random forest algorithm to obtain the correct rate coefficient and the question proportion coefficient.

[0111] Based on the classroom information dataset, data cleaning and preprocessing are performed to obtain classroom information training set data;

[0112] A decision tree is constructed based on the classroom information training set data to obtain a decision tree dataset;

[0113] Based on the decision tree dataset, model training is performed using the random forest algorithm to obtain the correct rate coefficient and the question proportion coefficient.

[0114] The decision tree dataset is divided into decision tree training set data and decision tree test set data;

[0115] Random sampling is performed based on the decision tree training set data to obtain decision tree sampling set data;

[0116] Calculate the root mean square error based on the decision tree sampling set data and the decision tree test set data to obtain the decision tree root mean square error;

[0117] When the decision tree root mean square error is less than the preset error threshold, perform a majority voting operation on the decision tree sampling set data to obtain the correct rate coefficient and the problem proportion coefficient.

[0118] It should be noted that data cleaning and preprocessing require checking for missing values, outliers, and duplicate values in the dataset and dealing with them. For example, if the number of passive question answers for a certain student is negative, it is regarded as an outlier and corrected or deleted, and the data is standardized or normalized for subsequent model training. For example, convert the number of question answers into a relative proportion so that its value is within the range of 0-1. Select features from the classroom information dataset to construct a decision tree and determine the target variable of the decision tree. For example, select the number of active question answers, the number of passive question answers, and the number of correct question answers as features, and use the number of correct question answers as the target variable to predict the correct rate of students' answers; for another example, select the number of active question answers, the number of passive question answers, and the total number of classroom questions as features, and use the number of active question answers as the target variable to predict the problem proportion rate of students. Use the decision tree algorithm to construct a decision tree model based on the features and the target variable. The decision tree can be split according to different values of the features, forming multiple nodes and branches, and finally obtaining a decision tree structure. Divide the constructed decision tree dataset into decision tree training set data and decision tree test set data. The decision tree training set is used to train the random forest model, and the decision tree test set is used to evaluate the performance of the model. Randomly sample from the decision tree training set data to obtain the decision tree sampling set data. Random sampling is a sampling method with replacement, that is, the same sample will be sampled multiple times, which can provide different training data for each decision tree in the random forest and increase the diversity between the trees. When the decision tree root mean square error is less than the preset error threshold, it indicates that the model performance meets the requirements. At this time, perform a majority voting operation on the decision tree sample set data. For example, for each sample, count the prediction results of multiple decision trees and take the prediction value that appears most frequently as the final prediction result. This final prediction result is the correct rate coefficient and the problem proportion coefficient.

[0119] Among them, the formula for the root mean square error is:

[0120]

[0121] In the formula, e RMSE is the root mean square error, n is the number of samples, is the i-th predicted value data, y i is the i-th actual value data.

[0122] Exemplarily, 70% of the decision tree dataset is divided into decision tree training set data, and 30% of the decision tree dataset is divided into decision tree test set data. For example, if there are 40 students in a class, then 28 students are divided into decision tree training set data and 12 students are divided into decision tree test set data. Random sampling is performed on the decision tree training set data to obtain decision tree sampling set data. The random forest algorithm is used to train the decision tree sampling set data, and then multiple decision trees are constructed and the prediction results are integrated. For example, the integrated decision tree sampling set data results are [0.8, 0.9, 0.8, 0.78, 0.17, 0.8, 0.8]. Through the majority voting result, the predicted correct rate coefficient result is 0.8. Calculated by the root mean square error formula, e RMSE = 0.256. If the preset error threshold is 0.3 and the convergence condition is satisfied at this time, then 0.8 is the final result of the correct rate coefficient.

[0123] In step S15, normalization calculation is performed according to the correct rate coefficient and the problem proportion coefficient to obtain a normalized correct rate coefficient and a normalized problem proportion coefficient.

[0124] The normalized correct rate coefficient is calculated by the following formula:

[0125]

[0126] In the formula, θ′ is the normalized correct rate coefficient, θ is the correct rate coefficient, μ is the problem proportion coefficient, and λ is the preset focus parameter;

[0127] The normalized problem proportion coefficient is calculated by the following formula:

[0128]

[0129] In the formula, μ′ is the normalized problem proportion coefficient.

[0130] It should be noted that through normalization processing, the correct rate coefficient and the problem proportion coefficient are converted into a value between 0 and 1, and the correctness of the students' answers and the active answering activity are placed under the same standard. Considering these two factors comprehensively, a more accurate classroom performance evaluation is obtained.

[0131] Exemplarily, the correct rate coefficient of student A is 0.8, the problem proportion coefficient is 0.5, and the preset focus parameter is 0.7. Then

[0132] In step S16, classroom quality parameter calculation is performed according to the total number of classes, the optimal focus coefficient, the classroom question information, the normalized correct rate coefficient, and the normalized problem proportion coefficient to obtain a classroom quality evaluation parameter.

[0133] The classroom quality evaluation parameter is calculated by the following formula:

[0134]

[0135] In the formula, Q is the classroom quality evaluation parameter, θ′ is the normalized correct rate coefficient, I is the number of times students actively answer questions, P is the number of times students passively answer questions, R is the number of correct answers by students, μ′ is the normalized question proportion coefficient, M is the total number of classes, α is the first optimal parameter, β is the second optimal parameter, γ is the third optimal parameter, A is the first focus parameter, B is the second focus parameter, C is the third focus parameter, and λ is the preset focus parameter.

[0136] It should be noted that the first term in the formula combines the normalized correct rate coefficient with the correct rate of students' answers, reflecting the proportion of the correctness of students' answers in the classroom quality. The second term in the formula combines the normalized question proportion coefficient with the activity of students' answering questions, reflecting the proportion of the activity of students' answering questions in the classroom quality. The third term λ(αA + βB + γC) in the formula combines the preset focus parameter with the optimal focus coefficient, reflecting the proportion of students' focus in the classroom quality.

[0137] Exemplarily, assuming that the normalized correct rate coefficient θ′ = 0.4, the number of correct answers by students R = 8, the number of times students passively answer questions P = 5, the total number of classes M = 20, the normalized question proportion coefficient μ′ = 0.25, the preset focus parameter λ = 0.35, the first optimal parameter α = 0.7, the second optimal parameter β = 0.15, the third optimal parameter γ = 0.1, the first focus parameter A = 0.8, the second focus parameter B = 0.1, and the third focus parameter C = 0.1, then the classroom quality evaluation parameter Q = 0.619 can be calculated.

[0138] In step S17, the classroom teaching quality evaluation result is obtained by matching the classroom quality evaluation parameter with the preset evaluation threshold parameter.

[0139] When Q > Q1, the classroom teaching quality evaluation result is excellent;

[0140] When Q2 < Q ≤ Q1, the classroom teaching quality evaluation result is good;

[0141] When Q3 < Q ≤ Q2, the classroom teaching quality evaluation result is passing;

[0142] When Q ≤ Q3, the classroom teaching quality evaluation result is failing;

[0143] Wherein, Q is the classroom quality evaluation parameter, and Q1, Q2, and Q3 are respectively the preset first evaluation threshold, second evaluation threshold, and third evaluation threshold.

[0144] The preset of the evaluation threshold is set according to teaching requirements. Exemplarily, the preset first evaluation threshold, second evaluation threshold, and third evaluation threshold are Q1 = 0.87, Q2 = 0.72, and Q3 = 0.6 respectively. According to the above comprehensive calculation, the classroom quality evaluation parameter Q = 0.619. Then, 0.6 < Q < 0.72, so the evaluation result of the classroom teaching quality is passing.

[0145] In summary, the present invention discloses a method for evaluating classroom teaching quality based on big data, including obtaining the eye gaze orientation information of students, the facial direction information of students, classroom question information, and the total number of classes; determining the concentration degree according to the eye gaze orientation information and the facial direction information to obtain a concentration parameter; constructing a concentration data set according to the concentration parameter and using a particle swarm optimization algorithm for optimization to obtain an optimal concentration coefficient; constructing a classroom information data set according to the classroom question information and using a random forest algorithm for training to obtain a correct rate coefficient and a question proportion coefficient; performing normalization calculation according to the correct rate coefficient and the question proportion coefficient to obtain a normalized correct rate coefficient and a normalized question proportion coefficient; calculating a classroom quality parameter according to the total number of classes, the optimal concentration coefficient, the classroom question information, the normalized correct rate coefficient, and the normalized question proportion coefficient to obtain a classroom quality evaluation parameter; and matching the classroom quality evaluation parameter with a preset evaluation threshold parameter to obtain an evaluation result of the classroom teaching quality. The method collects the eye gaze orientation information, facial direction information, classroom question information, and the total number of classes of students, judges the concentration degree of students according to the eye gaze and facial direction information to obtain a concentration parameter, then uses the concentration parameter in the particle swarm optimization algorithm to obtain an optimal concentration coefficient, uses the classroom question information, trains through the random forest algorithm to obtain a correct rate coefficient and a question proportion coefficient, performs normalization processing on the correct rate coefficient and the question proportion coefficient to obtain a normalized coefficient, and finally combines the total number of classes, the optimal concentration coefficient, the classroom question information, and the normalized coefficient to calculate a classroom quality evaluation parameter, thereby obtaining an evaluation result of the classroom teaching quality. The method can complete the evaluation of teaching according to multiple factors and correctly evaluate the classroom teaching quality.

[0146] Referring to Figure 2 , the second embodiment of the present invention provides a classroom teaching quality evaluation system based on big data, including:

[0147] A data acquisition module for obtaining the eye gaze orientation information of students, the facial direction information of students, classroom question information, and the total number of classes;

[0148] The concentration determination module is used to determine the concentration based on the eye gaze orientation information and the face direction information, and obtain the concentration parameter;

[0149] The concentration optimization module is used to construct a concentration data set according to the concentration parameter, and use the particle swarm optimization algorithm for optimization to obtain the optimal concentration coefficient;

[0150] The classroom information training module is used to construct a classroom information data set according to the classroom question information, and use the random forest algorithm for training to obtain the correct rate coefficient and the question proportion coefficient;

[0151] The normalization calculation module is used to perform normalization calculation according to the correct rate coefficient and the question proportion coefficient to obtain the normalized correct rate coefficient and the normalized question proportion coefficient;

[0152] The evaluation parameter calculation module is used to calculate the classroom quality parameters according to the total number of classes, the optimal concentration coefficient, the classroom question information, the normalized correct rate coefficient and the normalized question proportion coefficient, and obtain the classroom quality evaluation parameter;

[0153] The evaluation result matching module is used to match the classroom quality evaluation parameter with the preset evaluation threshold parameter to obtain the classroom teaching quality evaluation result.

[0154] It should be noted that a classroom teaching quality evaluation system based on big data provided by an embodiment of the present invention is used to execute all the process steps of a method for a classroom teaching quality evaluation system based on big data in the above embodiment. The working principles and beneficial effects of the two correspond one by one, so they will not be elaborated here.

[0155] An embodiment of the present invention also provides an electronic device. The electronic device includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a classroom teaching quality evaluation program based on big data. When the processor executes the computer program, the steps in the above-mentioned embodiments of various classroom teaching quality evaluation methods are implemented, such as Figure 1 The step S11 shown. Alternatively, when the processor executes the computer program, the functions of each module / unit in the above-mentioned device embodiments are implemented, such as the data acquisition module.

[0156] Exemplarily, the computer program can be divided into one or more modules / units. The one or more modules / units are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program in the electronic device.

[0157] The electronic device can be a computing device such as a desktop computer, notebook, handheld computer, and smart tablet. The electronic device may include, but is not limited to, a processor and a memory. Those skilled in the art can understand that the above components are only examples of the electronic device and do not constitute a limitation on the electronic device. It may include more or fewer components than the above, or combine certain components, or different components. For example, the electronic device may also include input / output devices, network access devices, a bus, etc.

[0158] The so-called processor may be a central processing unit (CPU), or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The processor is the control center of the electronic device and connects various parts of the entire electronic device through various interfaces and lines.

[0159] The memory can be used to store the computer program and / or module. The processor realizes various functions of the electronic device by running or executing the computer program and / or module stored in the memory, and by calling the data stored in the memory. The memory mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system, applications required for at least one function (such as a sound playback function, an image playback function, etc.); the data storage area can store data created according to the use of the mobile phone (such as audio data, phone book, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as a hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one magnetic disk storage device, flash device, or other volatile solid-state storage devices.

[0160] Among them, if the modules / units integrated in the electronic device are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above-described embodiment methods of the present invention, it can also be completed by a computer program instructing relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-described various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0161] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the attached drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that they have a communication connection, which can be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0162] The above-described specific embodiments have further elaborated on the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only the specific embodiments of the present invention and is not used to limit the protection scope of the present invention. It is particularly pointed out that for those skilled in the art, any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for evaluating the quality of classroom teaching based on big data, characterized in that, including: obtaining the eye gaze direction information of students, the facial direction information of students, the classroom question information, and the total number of classes; judging the concentration according to the eye gaze direction information and the facial direction information to obtain a concentration parameter; constructing a concentration data set according to the concentration parameter and performing optimization using a particle swarm optimization algorithm to obtain an optimal concentration coefficient; constructing a classroom information data set according to the classroom question information and performing training using a random forest algorithm to obtain a correct rate coefficient and a question proportion coefficient; performing normalization calculation according to the correct rate coefficient and the question proportion coefficient to obtain a normalized correct rate coefficient and a normalized question proportion coefficient; calculating a classroom quality parameter according to the total number of classes, the optimal concentration coefficient, the classroom question information, the normalized correct rate coefficient, and the normalized question proportion coefficient to obtain a classroom quality evaluation parameter; matching the classroom quality evaluation parameter with a preset evaluation threshold parameter to obtain a classroom teaching quality evaluation result.

2. The method for evaluating the quality of classroom teaching based on big data according to claim 1, wherein The eye gaze direction information includes gazing at the teacher's face and gazing at other places; The facial direction information includes the front face and the side face; The classroom question information includes the number of questions actively answered by students, the number of questions passively answered by students, and the number of questions correctly answered by students.

3. The method for evaluating the quality of classroom teaching based on big data according to claim 2, wherein The judging the concentration according to the eye gaze direction information and the facial direction information to obtain a concentration parameter includes: when the facial direction information is the front face and the eye gaze direction information is gazing at the teacher's face, it is determined as the first concentration parameter; when the facial direction information is the side face and the eye gaze direction information is gazing at other places, it is determined as the second concentration parameter; when the facial direction information is the front face and the eye gaze direction information is gazing at other places, it is determined as the third concentration parameter.

4. The method for evaluating the quality of classroom teaching based on big data according to claim 1, wherein, The constructing a concentration data set according to the concentration parameter and performing optimization using a particle swarm optimization algorithm to obtain an optimal concentration coefficient includes: performing data cleaning and preprocessing on the concentration data set to obtain concentration training set data; establishing a velocity set and a position set according to the concentration training set data and initializing the data of the velocity set and the position set to obtain initial velocity data and initial position data; performing parameter optimization using a particle swarm optimization algorithm according to the initial velocity data and the initial position data to obtain an optimal concentration coefficient; wherein, the optimal concentration coefficient includes a first optimal parameter, a second optimal parameter, and a third optimal parameter.

5. The method for evaluating the quality of classroom teaching based on big data according to claim 1, wherein The constructing a classroom information data set according to the classroom question information and performing training using a random forest algorithm to obtain a correct rate parameter and a question proportion parameter includes: performing data cleaning and preprocessing on the classroom information data set to obtain classroom information training set data; constructing a decision tree according to the classroom information training set data to obtain a decision tree data set; performing model training using a random forest algorithm according to the decision tree data set to obtain a correct rate coefficient and a question proportion coefficient.

6. The method for evaluating the quality of classroom teaching based on big data according to claim 5, characterized in that, Based on the decision tree dataset, the random forest algorithm is used for model training to obtain the correct rate coefficient and the problem proportion coefficient, including: Dividing the decision tree dataset into decision tree training set data and decision tree test set data; Performing random sampling according to the decision tree training set data to obtain decision tree sampling set data; Calculating the root mean square error according to the decision tree sampling set data and the decision tree test set data to obtain the decision tree root mean square error; When the decision tree root mean square error is less than the preset error threshold, performing a majority voting operation on the decision tree sampling set data to obtain the correct rate coefficient and the problem proportion coefficient.

7. The method for evaluating the quality of classroom teaching based on big data according to claim 1, wherein The normalization calculation based on the correct rate coefficient and the problem proportion coefficient to obtain the normalized correct rate coefficient and the normalized problem proportion coefficient includes: The normalized correct rate coefficient is calculated by the following formula: In the formula, θ′ is the normalized correct rate coefficient, θ is the correct rate coefficient, μ is the problem proportion coefficient, and λ is the preset focus parameter; The normalized problem proportion coefficient is calculated by the following formula: In the formula, μ′ is the normalized problem proportion coefficient.

8. The method for evaluating the quality of classroom teaching based on big data according to claim 1, characterized in that The calculation of the classroom quality parameter based on the total number of classes, the optimal focus coefficient, the classroom problem information, the normalized correct rate coefficient, and the normalized problem proportion coefficient to obtain the classroom quality evaluation parameter includes: The classroom quality evaluation parameter is calculated by the following formula: In the formula, Q is the classroom quality evaluation parameter, θ′ is the normalized correct rate coefficient, I is the number of students who actively answer questions, P is the number of students who passively answer questions, R is the number of students who answer questions correctly, μ′ is the normalized problem proportion coefficient, M is the total number of classes, α is the first optimal parameter, β is the second optimal parameter, γ is the third optimal parameter, A is the first focus parameter, B is the second focus parameter, C is the third focus parameter, and λ is the preset focus parameter.

9. The method for evaluating the quality of classroom teaching based on big data according to claim 1, wherein The matching of the classroom quality evaluation parameter and the preset evaluation threshold parameter to obtain the classroom teaching quality evaluation result includes: When Q > Q1, the classroom teaching quality evaluation result is excellent; When Q2 < Q ≤ Q1, the classroom teaching quality evaluation result is good; When Q3 < Q ≤ Q2, the classroom teaching quality evaluation result is passing; When Q ≤ Q3, the classroom teaching quality evaluation result is failing; Among them, Q is the classroom quality evaluation parameter, and Q1, Q2, and Q3 are the preset first evaluation threshold, second evaluation threshold, and third evaluation threshold respectively.

10. A classroom teaching quality evaluation system based on big data, characterized in that, Including: A data acquisition module for acquiring the eye gaze orientation information of students, the facial direction information of students, the classroom problem information, and the total number of classes; A focus determination module for determining the focus according to the eye gaze orientation information and the facial direction information to obtain a focus parameter; A focus optimization module for constructing a focus dataset according to the focus parameter and performing optimization using the particle swarm optimization algorithm to obtain the optimal focus coefficient; A classroom information training module for constructing a classroom information dataset according to the classroom problem information and performing training using the random forest algorithm to obtain the correct rate coefficient and the problem proportion coefficient; A normalization calculation module, configured to perform normalization calculation according to the correct rate coefficient and the problem proportion coefficient, so as to obtain a normalized correct rate coefficient and a normalized problem proportion coefficient; An evaluation parameter calculation module, configured to calculate classroom quality parameters according to the total number of classes, the optimal concentration coefficient, the classroom problem information, the normalized correct rate coefficient, and the normalized problem proportion coefficient, so as to obtain classroom quality evaluation parameters; An evaluation result matching module, configured to match the classroom quality evaluation parameters with preset evaluation threshold parameters to obtain a classroom teaching quality evaluation result.

Images