Multi-dimensional education comprehensive evaluation system based on intelligent education AI model

Abstract

The invention belongs to the technical field of education management, and particularly relates to a multi-dimensional education comprehensive evaluation system based on a wisdom education AI model, which comprises a wisdom education AI model unit, a multi-dimensional comprehensive evaluation unit and an education management end, according to the invention, all test score information, homework completion record information and classroom monitoring videos of corresponding students in each subject in a detection period are sent to the AI model analysis module through the test score record transmission module, the homework completion record transmission module and the classroom real-time monitoring module; the intelligent education AI model analyzes the performance of the corresponding students in all aspects one by one based on the score information, the homework completion record information and the classroom monitoring video, and the multi-dimensional comprehensive evaluation unit comprehensively evaluates the learning conditions of the corresponding students based on the analysis result of the intelligent education AI model. And targeted student management can be realized, the overall teaching performance of the class can be improved, the teaching management difficulty can be obviously reduced, and the intelligent degree is high.

Description

Technical Field

[0001] The present invention relates to the technical field of education management, and specifically to a multi-dimensional education comprehensive evaluation system based on an intelligent education AI model. Background Art

[0002] With the rapid development of information technology, intelligent education has become an important trend in the field of education. In a Chinese invention patent with the publication number CN117952315A, a college education teaching management information system is disclosed. By setting a learning effect evaluation unit to comprehensively consider the mid-term grades, final grades, and usual test grades of students, the evaluation of students' online learning effects becomes more comprehensive, the weight of phased key exams is reduced, the learning effect evaluation is more accurate, and effective supervision and early warning of students' learning are realized.

[0003] However, in the actual application process of the above invention technical solution, it mainly focuses on the single evaluation of exam scores, ignores the performance of students in other key aspects, cannot comprehensively evaluate the learning status of corresponding students from multiple dimensions, is not conducive to realizing targeted student management and improving the overall teaching performance of the class, has a large teaching management difficulty, and a low degree of intelligence.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the Invention

[0005] The purpose of the present invention is to provide a multi-dimensional education comprehensive evaluation system based on an intelligent education AI model, which solves the problems that the existing technology mainly focuses on the single evaluation of exam scores, ignores the performance of students in other key aspects, cannot comprehensively evaluate the learning status of corresponding students from multiple dimensions, is not conducive to realizing targeted student management and improving the overall teaching performance of the class, has a large teaching management difficulty, and a low degree of intelligence.

[0006] To achieve the above purpose, the present invention provides the following technical solution:

[0007] A multi-dimensional education comprehensive evaluation system based on an intelligent education AI model includes an intelligent education AI model unit, a multi-dimensional comprehensive evaluation unit, and an education management terminal. Among them, the intelligent education AI model unit marks the class to be supervised as the target class, marks the corresponding students in the target class as the supervised object i, and i is a natural number greater than 1; a detection period with a set number of days L1 is set. When the number of days reaches L1, the test feedback value, homework feedback value, and classroom feedback value of the supervised object i are obtained, and the test feedback value, homework feedback value, and classroom feedback value of the supervised object i are sent to the multi-dimensional comprehensive evaluation unit.

[0008] The multi-dimensional comprehensive evaluation unit performs numerical calculations on the test feedback value, homework feedback value, and classroom feedback value to obtain a multi-dimensional teaching evaluation coefficient. It numerically compares the multi-dimensional teaching evaluation coefficient with a preset multi-dimensional teaching evaluation coefficient threshold. If the multi-dimensional teaching evaluation coefficient exceeds the preset multi-dimensional teaching evaluation coefficient threshold, the regulatory object i is marked as a strongly managed object. If the multi-dimensional teaching evaluation coefficient does not exceed the preset multi-dimensional teaching evaluation coefficient threshold, the regulatory object i is marked as a weakly managed object, and the marking information of the regulatory object i is sent to the education management terminal.

[0009] Furthermore, the intelligent education AI model unit includes a test score record transmission module, a homework completion record transmission module, a classroom real-time monitoring module, and an AI model analysis module;

[0010] The test score record transmission module obtains all the test score information of the regulatory object i in each subject during the detection period and sends all the test score information of each subject to the AI model analysis module; the homework completion record transmission module obtains the homework completion record information of the regulatory object i in each subject during the detection period and sends the homework completion record information to the AI model analysis module;

[0011] The classroom real-time monitoring module conducts real-time monitoring of the classroom and sends all the classroom monitoring videos during the detection period to the AI model analysis module; the AI model analysis module constructs an intelligent education AI model based on deep learning and machine learning technologies. The intelligent education AI model analyzes all aspects of the regulatory object i based on the score information, homework completion record information, and classroom monitoring videos, and obtains the test feedback value, homework feedback value, and classroom feedback value of the regulatory object i through analysis.

[0012] Furthermore, the method for analyzing and obtaining the test feedback value is as follows:

[0013] The test scores of the regulatory object i for the corresponding test are collected, and the test scores of the regulatory object i for the corresponding test are numerically compared with the average score of the corresponding class. If the test score does not exceed the average score of the corresponding class, the corresponding test is marked as a non-compliant test for the regulatory object i;

[0014] By analyzing to determine the non-dominant subjects of the regulatory object i during the detection period, obtaining the number of non-dominant subjects corresponding to the regulatory object i during the detection period and marking it as the non-dominant detection value, calculating the average value of the subject measurement values of all subjects corresponding to the regulatory object i to obtain the learning performance value, and marking the subject measurement value with the largest value as the subject inferior value; the test feedback value of the regulatory object i is obtained by numerically calculating the non-dominant detection value, learning performance value, and subject inferior value.

[0015] Furthermore, the method for analyzing and determining non-dominant subjects is specifically as follows:

[0016] Obtain the number of non-compliant tests for the corresponding subject of the supervised object i during the detection period, calculate the ratio with the total number of tests for the corresponding subject in the target class during the detection period to obtain the subject non-compliance matching value, mark the difference value between the test scores of the corresponding non-compliant tests and the average score of the corresponding class as the test difference detection value, and calculate the average value of all test difference detection values for the corresponding subject of the supervised object i during the detection period to obtain the subject disadvantage value;

[0017] Calculate the subject measurement table value by performing numerical calculations on the subject non-compliance matching value and the subject disadvantage value, and perform a numerical comparison between the subject measurement table value and the preset subject measurement table threshold. If the subject measurement table value exceeds the preset subject measurement table threshold, mark the corresponding subject as a non-dominant subject of the supervised object i.

[0018] Furthermore, the method for analyzing and obtaining the homework feedback value is as follows:

[0019] Based on the homework completion record information, obtain the number of times the supervised object i fails to complete the homework on time during the detection period and mark it as the homework unfinished value. And when the supervised object i fails to complete the homework on time, start tracking and stop tracking when the supervised object i completes the homework, so as to obtain the homework make-up value;

[0020] Obtain all the homework make-up values of the supervised object i during the detection period and calculate their average value to obtain the make-up performance value. And perform a numerical comparison between the homework make-up value and the preset homework make-up threshold. If the homework make-up value exceeds the preset homework make-up threshold, mark the corresponding homework make-up value as the homework abnormal make-up value, and mark the number of homework abnormal make-up values corresponding to the supervised object i during the detection period as the abnormal make-up detection value; Calculate the homework feedback value by performing numerical calculations on the homework unfinished value, the make-up performance value, and the abnormal make-up detection value.

[0021] Furthermore, the method for analyzing and obtaining the classroom feedback value is as follows:

[0022] Obtain all the classroom monitoring videos of the target class during the detection period and focus on the supervised object i in the classroom monitoring videos; Based on the classroom monitoring videos, collect the number of times the supervised object i is late and leaves early during the detection period and mark it as the late and early leave frequency, and mark the total duration of the supervised object i being late and leaving early during the detection period as the late and early leave duration;

[0023] Based on the classroom monitoring videos, collect the number of times the supervised object i raises their hand during the classroom Q&A session during the detection period and mark it as the Q&A interaction value, and obtain the classroom behavior anomaly value through classroom posture monitoring and identification analysis. Calculate the classroom feedback value of the supervised object i by performing numerical calculations on the Q&A interaction value, the late and early leave frequency, the late and early leave duration, and the classroom behavior anomaly value.

[0024] Furthermore, the specific analysis process of classroom posture monitoring and identification analysis is as follows:

[0025] Monitor the classroom postures of the supervised object i to identify classroom misbehaviors, obtain all classroom misbehaviors of the supervised object i during the detection period, classify all classroom misbehaviors, count the occurrence times of classroom misbehaviors of corresponding types and mark them as behavior occurrence frequency values, and obtain the single-duration of classroom misbehaviors of corresponding types. Sum up all single-durations of classroom misbehaviors of corresponding types during the detection period to calculate the behavior occurrence duration value;

[0026] Perform numerical calculations on the behavior occurrence frequency value and the behavior occurrence duration value to obtain the behavior impact value. Preset a set of preset impact weight values for each type of classroom misbehavior in advance, multiply the behavior impact value of the corresponding type of classroom misbehavior by the corresponding preset impact weight value to obtain the behavior detection value; obtain the behavior detection values of all types of classroom misbehaviors corresponding to the supervised object i during the detection period, and mark the sum value of all behavior detection values of the supervised object i as the classroom behavior anomaly value.

[0027] Further, the education management terminal is communicatively connected to the teacher management and evaluation unit. The teacher management and evaluation unit analyzes the teaching management performance of the corresponding teacher during the detection period, marks the corresponding teacher as an excellent-performance teacher or a poor-performance teacher through the analysis, and sends the marking information of the corresponding teacher to the education management terminal.

[0028] Further, the specific analysis process of the teacher management and evaluation unit is as follows:

[0029] Obtain the teaching subjects and the classes taught by the corresponding teacher, collect the class average score of the corresponding class in the corresponding test of this subject, numerically compare the class average score with the corresponding preset class average score threshold. If the class average score does not exceed the preset class average score threshold, assign the test non-excellent symbol ZP-1 to the corresponding test;

[0030] Obtain the ratio of the number of tests with the test non-excellent symbol ZP-1 corresponding to the corresponding teacher during the detection period and mark it as the test non-excellent value. Numerically compare the test non-excellent value with the preset test non-excellent threshold. If the test non-excellent value exceeds the preset test non-excellent threshold, mark the corresponding teacher as a poor-performance teacher;

[0031] If the non-optimal value of the test does not exceed the preset non-optimal test threshold, the total class hours of the corresponding teacher during the detection period are obtained, the ratio of the number of people with classroom misbehaviors and the ratio of the average number of people raising their hands in the classroom Q&A session during the corresponding class hours of teaching are collected, and the ratio of the average number of people raising their hands and the ratio of the number of people with classroom misbehaviors are respectively compared numerically with the preset ratio threshold of the average number of people raising their hands and the preset ratio threshold of the number of people with classroom misbehaviors. If the ratio of the average number of people raising their hands does not exceed the preset ratio threshold of the average number of people raising their hands or the ratio of the number of people with classroom misbehaviors exceeds the preset ratio threshold of the number of people with classroom misbehaviors, the corresponding class hour teaching process is given the teaching management anomaly symbol ZX-1;

[0032] Obtain the number of class hours with the teaching management anomaly symbol ZX-1 corresponding to the corresponding teacher during the detection period and calculate the ratio with the total class hours to obtain the classroom management anomaly value; compare the classroom management anomaly value numerically with the preset classroom management anomaly threshold. If the classroom management anomaly value exceeds the preset classroom management anomaly threshold, mark the corresponding teacher as a teacher with poor performance; if the classroom management anomaly value does not exceed the preset classroom management anomaly threshold, mark the corresponding teacher as a teacher with excellent performance.

[0033] Compared with the prior art, the beneficial effects of the present invention are:

[0034] 1. In the present invention, through the test score record transmission module, the homework completion record transmission module and the classroom real-time monitoring module, all the test score information, homework completion record information and classroom monitoring videos of the corresponding students in various subjects during the detection period are sent to the AI model analysis module. The AI model analysis module analyzes the various aspects of the corresponding students one by one based on the intelligent education AI model, and the multi-dimensional comprehensive evaluation unit comprehensively evaluates the learning status of the corresponding students based on the analysis results of the intelligent education AI model, which is conducive to realizing targeted student management and improving the overall teaching performance of the class, significantly reducing the teaching management difficulty and having a high degree of intelligence;

[0035] 2. In the present invention, through the teacher management evaluation unit, the teaching information corresponding to each teacher is obtained from the intelligent education AI model unit, and the teaching management performance of each teacher during the detection period is analyzed to determine excellent performance teachers and poor performance teachers, which is convenient for formulating corresponding management plans for different teachers in the future, improving the teacher quality and teaching management performance of teachers, significantly reducing the teacher supervision difficulty, having a high level of intelligence, and being conducive to ensuring the teaching reputation of the school. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] For the convenience of those skilled in the art to understand, the present invention will be further described below with reference to the accompanying drawings;

[0037] Figure 1 It is the system block diagram of Embodiment 1 in the present invention;

[0038] Figure 2It is the system block diagram of Embodiment 2 in the present invention. Detailed implementation manners

[0039] 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.

[0040] Embodiment 1: As Figure 1 shown, the multi-dimensional education comprehensive evaluation system based on the intelligent education AI model proposed by the present invention includes an intelligent education AI model unit, a multi-dimensional comprehensive evaluation unit, and an education management terminal. Among them, the intelligent education AI model unit marks the class to be supervised as the target class, marks the corresponding students in the target class as the supervised object i, and i is a natural number greater than 1;

[0041] Set a detection period of L1 days. Preferably, L1 = 60; when the number of days reaches L1, obtain the test feedback value YPi, homework feedback value WLi, and classroom feedback value QWi of the supervised object i, and send the test feedback value YPi, homework feedback value WLi, and classroom feedback value QWi of the supervised object i to the multi-dimensional comprehensive evaluation unit;

[0042] It should be noted that the intelligent education AI model unit includes a test score record transmission module, a homework completion record transmission module, a classroom real-time monitoring module, and an AI model analysis module. The test score record transmission module, the homework completion record transmission module, and the classroom real-time monitoring module are respectively communicatively connected to the AI model analysis module to provide information support for the analysis process of the AI model analysis module;

[0043] The test score record transmission module obtains all the test score information of the supervised object i in each subject during the detection period, and sends all the test score information of each subject to the AI model analysis module; the homework completion record transmission module obtains the homework completion record information of the supervised object i in each subject during the detection period, and sends the homework completion record information to the AI model analysis module; the classroom real-time monitoring module conducts real-time monitoring of the classroom and sends all the classroom monitoring videos during the detection period to the AI model analysis module;

[0044] The AI model analysis module constructs an intelligent education AI model based on deep learning and machine learning technologies. The intelligent education AI model analyzes all aspects of the supervised object i one by one based on the score information, homework completion record information, and classroom monitoring videos, and obtains the test feedback value YPi, homework feedback value WLi, and classroom feedback value QWi of the supervised object i through the analysis.

[0045] Among them, the method for analyzing and obtaining the test feedback value YPi is as follows: collect the test scores of the supervised object i for the corresponding test, compare the test scores of the supervised object i for the corresponding test with the average score of the corresponding class. If the test score does not exceed the average score of the corresponding class, it indicates that the corresponding test result of the supervised object i is poor, and then mark the corresponding test as a non-compliant test for the supervised object i;

[0046] Obtain the number of non-compliant tests of the corresponding subject for the supervised object i during the detection period and calculate the ratio with the total number of tests of the corresponding subject in the target class during the detection period to obtain the subject non-compliance matching value. Mark the difference value between the test scores of the corresponding non-compliant tests and the average score of the corresponding class as the test difference detection value, and calculate the average value of all test difference detection values of the corresponding subject for the supervised object i during the detection period to obtain the subject disadvantage value;

[0047] Perform numerical calculation on the subject non-compliance matching value XWi and the subject disadvantage value YSi through the formula TPi = a1×XWi + a2×YSi, where a1 and a2 are preset weight coefficients, and a1 > a2 > 0; moreover, the larger the numerical value of the subject test table value TPi, the worse the test performance of the supervised object i in the corresponding subject during the detection period;

[0048] Compare the subject test table value TPi with the preset subject test table threshold. If the subject test table value TPi exceeds the preset subject test table threshold, it indicates that the test performance of the supervised object i in the corresponding subject during the detection period is poor, and the knowledge mastery performance of the supervised object i in the corresponding subject is poor, then mark the corresponding subject as a non-dominant subject for the supervised object i;

[0049] Obtain the number of non-dominant subjects corresponding to the supervised object i during the detection period and mark it as the non-dominant detection value, calculate the average value of the subject test table values of all subjects corresponding to the supervised object i to obtain the learning performance value, and mark the largest subject test table value as the subject inferior amplitude value;

[0050] Perform numerical calculation on the non-dominant detection value FYi, the learning performance value XMi, and the subject inferior amplitude value HFi through the formula YPi = mg×FYi + (rg×XMi + nq×HFi) / 2 to obtain the test feedback value YPi of the supervised object i; where mg, rg, and nq are preset weight coefficients, and mg > rg > nq > 0; moreover, the larger the numerical value of the test feedback value YPi, the worse the comprehensive learning effect performance of the supervised object i during the detection period.

[0051] Moreover, the method for analyzing and obtaining the homework feedback value WLi is as follows: Based on the homework completion record information, obtain the number of times that the supervised object i fails to complete the homework on time during the detection period and mark it as the homework outstanding value. And when the supervised object i fails to complete the homework on time, start tracking, and stop tracking when the supervised object i completes and submits the homework. Accordingly, obtain the homework make-up value, that is, the time interval between the time when the homework should be submitted and the time when the homework is finally completed and submitted;

[0052] Obtain all the homework make-up values of the supervised object i during the detection period and calculate their average value to obtain the make-up time performance value. And compare the homework make-up value with the preset homework make-up time threshold. If the homework make-up value exceeds the preset homework make-up time threshold, mark the corresponding homework make-up value as the abnormal make-up value of the homework, and mark the number of abnormal make-up values of the homework corresponding to the supervised object i during the detection period as the abnormal make-up detection value;

[0053] Perform numerical calculation on the homework outstanding value FNi, the make-up time performance value QLi, and the abnormal make-up detection value PXi through the formula WLi = (hu × FNi + sq × PXi) / 2 + rw × QLi to obtain the homework feedback value WLi; where hu, sq, and rw are preset weight coefficients greater than zero. And the larger the value of the homework feedback value WLi, the worse the comprehensive homework completion performance of the supervised object i during the detection period.

[0054] Furthermore, the method for analyzing and obtaining the classroom feedback value QWi is as follows: Obtain all the classroom monitoring videos of the target class during the detection period, and focus on the supervised object i in the classroom monitoring videos; Based on the classroom monitoring videos, collect the number of times that the supervised object i is late and leaves early during the detection period and mark it as the late and early leave frequency, and mark the total duration of the supervised object i being late and leaving early during the detection period as the late and early leave duration;

[0055] Based on the classroom monitoring videos, collect the number of times that the supervised object i raises his / her hand during the classroom Q&A session and mark it as the Q&A interaction value; Monitor the classroom posture of the supervised object i to identify classroom bad behaviors, obtain all the classroom bad behaviors of the supervised object i during the detection period, and classify all the classroom bad behaviors (including dozing off, playing with mobile phones, fighting, etc.);

[0056] Mark the number of occurrences of the corresponding type of classroom bad behavior and mark it as the behavior occurrence frequency value, and obtain the single duration of the corresponding type of classroom bad behavior. Sum up all the single durations of the corresponding type of classroom bad behavior during the detection period to obtain the behavior occurrence time value;

[0057] The behavior generation frequency value SWi and the behavior generation time value NXi are numerically calculated through the formula LPi = b1×SWi + b2×NXi to obtain the behavior influence value LPi; where b1 and b2 are preset weight coefficients, b1 > b2 > 0.25; and the larger the value of the behavior influence value LPi, the greater the adverse impact on classroom teaching during the detection period due to the corresponding type of classroom misbehavior by the supervision object i.

[0058] A set of preset influence weight values greater than zero are set in advance for each type of classroom misbehavior. Among them, the greater the adverse impact of the corresponding type of classroom misbehavior on classroom teaching, the greater the value of the preset influence weight value matched with it.

[0059] Multiply the behavior influence value of the corresponding type of classroom misbehavior by the corresponding preset influence weight value to obtain the behavior detection value; obtain the behavior detection values of all types of classroom misbehavior corresponding to the supervision object i during the detection period, and mark the sum value of all behavior detection values of the supervision object i as the classroom behavior anomaly value.

[0060] The question-and-answer interaction value HXi, the late arrival and early departure frequency ZFi, the late arrival and early departure duration HPi, and the classroom behavior anomaly value SYi are numerically calculated through the formula QWi = (tu×ZFi + uy×HPi + kp×SYi) / (wq×HXi + 1) to obtain the classroom feedback value QWi of the supervision object i; where tu, uy, kp, and wq are preset proportionality coefficients greater than zero, and the larger the value of the classroom feedback value QWi, the worse the overall classroom performance of the supervision object i during the detection period.

[0061] The multi-dimensional comprehensive evaluation unit obtains the test feedback value, the homework feedback value, and the classroom feedback value of the supervision object i, and numerically calculates the test feedback value YPi, the homework feedback value WLi, and the classroom feedback value QWi through the formula DXi = (eu1*YPi + eu2*WLi + eu3*QWi) / 3 to obtain the multi-dimensional teaching evaluation coefficient DXi; where eu1, eu2, and eu3 are preset weight coefficients greater than zero, and the larger the value of the multi-dimensional teaching evaluation coefficient DXi, the worse the learning status of the supervision object i during the detection period overall.

[0062] Numerically compare the multi-dimensional teaching evaluation coefficient DXi with the preset multi-dimensional teaching evaluation coefficient threshold. If the multi-dimensional teaching evaluation coefficient DXi exceeds the preset multi-dimensional teaching evaluation coefficient threshold, it indicates that overall, the learning status of the supervised object i during the detection period is poor, and it is necessary to strengthen the learning supervision of the supervised object i in the follow-up. Then, mark the supervised object i as a strongly managed object. If the multi-dimensional teaching evaluation coefficient DXi does not exceed the preset multi-dimensional teaching evaluation coefficient threshold, mark the supervised object i as a weakly managed object, indicating that overall, the learning status of the supervised object i during the detection period is good, and send the marking information of the supervised object i to the education management terminal to strengthen the learning supervision of the weakly managed object in the follow-up, realizing targeted student management, which is beneficial to improving the overall teaching performance of the class, significantly reducing the teaching difficulty, and having a high degree of intelligence.

[0063] Embodiment 2: As Figure 2 shown, the difference between this embodiment and Embodiment 1 is that the education management terminal is communicatively connected to the teacher management evaluation unit. The teacher management evaluation unit obtains the teaching information corresponding to the corresponding teacher from the intelligent education AI model unit, analyzes the teaching management performance of the corresponding teacher during the detection period, and marks the corresponding teacher as a teacher with excellent performance or poor performance through the analysis;

[0064] and send the marking information of the corresponding teacher to the education management terminal to facilitate the formulation of corresponding management plans for different teachers in the follow-up. For example, strengthen the teacher management and teaching training guidance for teachers with poor performance, improve the teacher quality and teaching management performance of the corresponding teachers, significantly reduce the teacher supervision difficulty, have a high level of intelligence, and are beneficial to ensuring the teaching reputation of the school; The specific analysis process of the teacher management evaluation unit is as follows:

[0065] Obtain the teaching subjects corresponding to the corresponding teacher and the classes taught, collect the class average score of the corresponding class in the corresponding test for the subject, numerically compare the class average score with the corresponding preset class average score threshold. If the class average score does not exceed the preset class average score threshold, it indicates that the test result status of the corresponding test is poor, and then assign a non-excellent test symbol ZP-1 to the corresponding test;

[0066] Obtain the ratio of the number of tests with the non-excellent test symbol ZP-1 corresponding to the corresponding teacher during the detection period and mark it as the non-excellent test value. Numerically compare the non-excellent test value with the preset non-excellent test threshold. If the non-excellent test value exceeds the preset non-excellent test threshold, it indicates that the teaching quality status of the corresponding teacher during the detection period is poor, and then mark the corresponding teacher as a teacher with poor performance;

[0067] If the non-optimal value of the test does not exceed the preset non-optimal test threshold, the total class hours of the corresponding teacher during the detection period are obtained, and the ratio of the number of students with classroom misbehaviors during the corresponding class hours (i.e., the ratio of the number of students with classroom misbehaviors in the corresponding class to the total number of students in the class) and the ratio of the average number of students raising their hands during the classroom Q&A session (i.e., the ratio of the average number of students raising their hands during the classroom Q&A session in the corresponding class hours to the total number of students in the class) are collected;

[0068] The ratio of the average number of students raising their hands and the ratio of the number of students with classroom misbehaviors are respectively compared numerically with the preset ratio threshold of the average number of students raising their hands and the preset ratio threshold of the number of students with classroom misbehaviors. If the ratio of the average number of students raising their hands does not exceed the preset ratio threshold of the average number of students raising their hands or the ratio of the number of students with classroom misbehaviors exceeds the preset ratio threshold of the number of students with classroom misbehaviors, indicating that the teaching management during the corresponding class hours is poor, then the corresponding class hours are given the teaching management exception symbol ZX-1;

[0069] The number of class hours with the teaching management exception symbol ZX-1 corresponding to the corresponding teacher during the detection period is obtained and the ratio is calculated with the total class hours to obtain the classroom management exception value; the classroom management exception value is compared numerically with the preset classroom management exception threshold. If the classroom management exception value exceeds the preset classroom management exception threshold, indicating that the teaching management performance of the corresponding teacher during the detection period is poor, then the corresponding teacher is marked as a teacher with poor performance; if the classroom management exception value does not exceed the preset classroom management exception threshold, indicating that the overall teaching performance of the corresponding teacher during the detection period is good, then the corresponding teacher is marked as a teacher with excellent performance.

[0070] The working principle of the present invention: When in use, all test score information, homework completion record information, and classroom monitoring videos of the corresponding students in various subjects during the detection period are sent to the AI model analysis module through the test score record transmission module, homework completion record transmission module, and classroom real-time monitoring module. The intelligent education AI model analyzes the various aspects of the corresponding students one by one based on the score information, homework completion record information, and classroom monitoring videos. The multi-dimensional comprehensive evaluation unit comprehensively evaluates the learning status of the corresponding students based on the analysis results of the intelligent education AI model and marks the corresponding students as strong management objects or weak management objects, and strengthens the learning supervision of the weak management objects in the follow-up, realizing targeted student management, which is beneficial to improving the overall teaching performance of the class, significantly reducing the teaching difficulty, and having a high degree of intelligence.

[0071] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation. The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only the specific embodiments. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification in order to better explain the principles and practical applications of the present invention, so that those skilled in the relevant technical field can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A multi-dimensional education comprehensive evaluation system based on the smart education AI model, characterized by: It includes a smart education AI model unit, a multi-dimensional comprehensive evaluation unit and an education management end, wherein the smart education AI model unit marks the class that needs to be supervised as the target class, and marks the corresponding students in the target class as the supervision object i, where i is a natural number greater than 1; obtains the test feedback value, homework feedback value and classroom feedback value of the supervision object i and sends them to the multi-dimensional comprehensive evaluation unit; The multidimensional comprehensive evaluation unit numerically calculates the test feedback value, homework feedback value and classroom feedback value to obtain the multidimensional teaching evaluation coefficient. If the multidimensional teaching evaluation coefficient exceeds the preset multidimensional teaching evaluation coefficient threshold, the supervision object i is marked as a strong management object; otherwise, the supervision object i is marked as a weak management object, and the marking information of the supervision object i is sent to the education management end.

2. According to claim 1, the multi-dimensional education comprehensive evaluation system based on the smart education AI model is characterized in that: The smart education AI model unit includes a test score record transmission module, a homework completion record transmission module, a classroom real-time monitoring module, and an AI model analysis module; The test score record transmission module obtains all the test score information of the supervised object i in each subject during the detection period, and sends all the test score information of each subject to the AI ​​model analysis module; The homework completion record transmission module obtains the homework completion record information of the supervised object i in each subject during the detection period, and sends the homework completion record information to the AI ​​model analysis module; The real-time classroom monitoring module monitors the classroom in real time and sends all classroom monitoring videos within the detection period to the AI ​​model analysis module; the AI ​​model analysis module uses the smart education AI model to analyze the performance of the supervised object i in all aspects based on the grade information, homework completion record information and classroom monitoring videos, and obtains the test feedback value, homework feedback value and classroom feedback value of the supervised object i through analysis.

3. The multi-dimensional education comprehensive evaluation system based on the smart education AI model according to claim 2 is characterized in that: The analysis and acquisition method of the test feedback value is as follows: The non-advantageous subjects of the supervision object i during the test period are determined through analysis, the number of non-advantageous subjects corresponding to the supervision object i during the test period is obtained and marked as non-advantageous test values, and the subject test table values ​​of all subjects corresponding to the supervision object i are averaged to obtain the learning performance value, and the subject test table value with the largest value is marked as the subject inferior value; the test feedback value of the supervision object i is obtained by numerically calculating the non-advantageous test value, the learning performance value and the subject inferior value.

4. The multi-dimensional education comprehensive evaluation system based on the smart education AI model according to claim 3 is characterized in that: The specific method for analyzing and determining non-advantageous subjects is as follows: the subject measurement value is obtained by numerically calculating the subject non-standard matching value and the subject disadvantage value. If the subject measurement value exceeds the preset subject measurement threshold, the corresponding subject will be marked as a non-advantageous subject of the supervision object i.

5. The multi-dimensional education comprehensive evaluation system based on the smart education AI model according to claim 2 is characterized in that: The analysis and acquisition method of the operation feedback value is as follows: Based on the job completion record information, the number of times the supervised object i failed to complete the job on time during the detection period is obtained and marked as the job unfinished value. The job feedback value is obtained by numerically calculating the job unfinished value, the make-up time performance value and the abnormal make-up detection value.

6. The multi-dimensional education comprehensive evaluation system based on the smart education AI model according to claim 2 is characterized in that: The analysis and acquisition method of classroom feedback value is as follows: obtain all classroom surveillance videos of the target class during the detection period, and focus on the supervised object i in the classroom surveillance video; obtain the classroom behavior abnormality value through classroom posture monitoring and recognition analysis, and numerically calculate the question-and-answer interaction value, the frequency of lateness and early departure, the duration of lateness and early departure, and the classroom behavior abnormality value to obtain the classroom feedback value of supervised object i.

7. The multi-dimensional education comprehensive evaluation system based on the smart education AI model according to claim 6 is characterized in that: The specific analysis process of classroom posture monitoring and recognition analysis is as follows: Monitor the classroom posture of the supervised object i to identify bad classroom behaviors, obtain all bad classroom behaviors of the supervised object i during the detection period, classify all bad classroom behaviors, obtain the behavior detection values ​​of all types of bad classroom behaviors corresponding to the supervised object i during the detection period, and mark the sum of all behavior detection values ​​of the supervised object i as the classroom behavior abnormality value.

8. The multi-dimensional education comprehensive evaluation system based on the smart education AI model according to claim 1 is characterized in that: The education management end is communicated with the teacher management evaluation unit, which analyzes the teaching management performance of the corresponding teacher during the detection period, marks the corresponding teacher as a good-performing teacher or a poor-performing teacher through analysis, and sends the marking information of the corresponding teacher to the education management end.

9. The multi-dimensional education comprehensive evaluation system based on the smart education AI model according to claim 8 is characterized in that: The specific analysis process of the teacher management evaluation unit is as follows: The test non-excellence value of the corresponding teacher is obtained. If the test non-excellence value exceeds the preset test non-excellence threshold, the corresponding teacher is marked as a poor-performing teacher; if the test non-excellence value does not exceed the preset test non-excellence threshold, the classroom management anomaly value is numerically compared with the preset classroom management anomaly threshold. If the classroom management anomaly value exceeds the preset classroom management anomaly threshold, the corresponding teacher is marked as a poor-performing teacher; otherwise, the corresponding teacher is marked as an excellent-performing teacher.

Images