An examination room monitoring system and method based on artificial intelligence

Through an examination room monitoring system based on artificial intelligence, the candidates' answering behavior in real time and the calculation of violation assessment scores is solved, and the problem of difficulty in identifying advanced cheating equipment in the existing technology is solved, achieving a more accurate examination room management and a fair and just examination environment.

CN118646850BActive Publication Date: 2024-11-08YIJIAO NETWORK TECH (FOSHAN) CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202411116797.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-15
Publication Date
2024-11-08
Estimated Expiration
2044-08-15

AI Technical Summary

Technical Problem

The existing examination room monitoring technology is difficult to identify advanced cheating equipment in a timely manner, resulting in successful cheating and affecting the fairness and justice of the exam.

Method used

The examination room monitoring system based on artificial intelligence is adopted to monitor candidates' answering behavior through full coverage monitoring, real-time video entry, examination content analysis and artificial intelligence evaluation, calculate the scores for illegal examinations, and dynamically adjust them based on the number of testers and the threshold for illegal assessment scores.

Benefits of technology

It improves the accuracy and fairness of examination room monitoring, effectively cracks down on candidates' use of cheating equipment, and ensures the fairness of the examination.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118646850B_ABST
    Figure CN118646850B_ABST
Patent Text Reader

Abstract

The present invention discloses an examination room monitoring system based on artificial intelligence, which belongs to the technical field of examination room monitoring. The system includes an examination room monitoring module, a real-time video input module, an examination content analysis module, an artificial intelligence evaluation module, and an early warning module. The output end of the examination room monitoring module is connected to the input end of the real-time video input module; the output end of the real-time video input module is connected to the input end of the examination content analysis module; the output end of the examination content analysis module is connected to the input end of the artificial intelligence evaluation module; the output end of the artificial intelligence evaluation module is connected to the input end of the early warning module; and an examination room monitoring method based on artificial intelligence is also provided to further ensure the fairness of the examination and severely crack down on the behavior of candidates using external devices to cheat.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of examination room monitoring, and in particular to an examination room monitoring system and method based on artificial intelligence. Background Art

[0002] In the field of examination room monitoring, judging from the current cheating situation in the examination room, carrying some cheating devices to cheat is one of the most common means, and people's prevention and control means are generally to shield the signal through a shielding device or to conduct entry detection with a metal detector. However, due to the increasing development of science, cheating devices are also constantly updated. When the level of the cheating device exceeds the signal range that can be shielded by the shielding device, people will not be able to detect the cheating means in time, resulting in the final successful cheating of the examinee, which seriously affects the fairness and justice of the examination. Therefore, the present invention is based on monitoring the behavior of the examinee answering questions in the examination room when the external instrument cannot identify the cheating device, so that the effect of examination room monitoring is more accurate, rather than just limited to the detection between devices. Summary of the invention

[0003] The purpose of the present invention is to provide an examination room monitoring system and method based on artificial intelligence to solve the problems raised in the above background technology.

[0004] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0005] An examination room monitoring system based on artificial intelligence, the system includes an examination room monitoring module, a real-time video input module, an examination content analysis module, an artificial intelligence evaluation module, and an early warning module;

[0006] The examination room monitoring module is used to monitor the examination room in full coverage to ensure video monitoring of each candidate; the real-time video recording module is used to record and store the monitored real-time video; the examination content analysis module is used to receive the examination paper content information for classification processing; the artificial intelligence evaluation module is used to perform comparative evaluation based on the preset examination behavior information and identify abnormal candidate behavior; the early warning module is used to issue an early warning, and record the candidate information and output it to the administrator port to remind the invigilator to pay special attention;

[0007] The output end of the examination room monitoring module is connected to the input end of the real-time video recording module; the output end of the real-time video recording module is connected to the input end of the examination content analysis module; the output end of the examination content analysis module is connected to the input end of the artificial intelligence evaluation module; the output end of the artificial intelligence evaluation module is connected to the input end of the early warning module.

[0008] According to the above technical solution, the examination room monitoring module includes a recording unit and a time recording unit;

[0009] The recording unit is used to record the process of the examinee answering questions in the examination room; the time recording unit is used to record the time of various actions of the examinee in the process of answering questions;

[0010] The output end of the recording unit is connected to the input end of the time recording unit; the output end of the time recording unit is connected to the input end of the real-time video recording module.

[0011] According to the above technical solution, the real-time video recording module includes a receiving unit and a storage unit;

[0012] The receiving unit is used to receive the real-time test video content recorded in the test room; the storage unit is used to store the video content for subsequent retrieval and analysis;

[0013] The output end of the receiving unit is connected to the input end of the storage unit; the output end of the storage unit is connected to the input end of the test content analysis module.

[0014] According to the above technical solution, the test content analysis module includes a test content classification unit, a time standard interval calculation unit, and a historical database;

[0015] The test content classification unit is used to classify the test content, classify the content according to whether a draft paper is needed, and classify the question type according to whether the question type is a multiple-choice question; the time standard interval calculation unit is used to retrieve historical data to calculate the time standard interval of various actions when answering questions; the historical database is used to store historical data for the system to retrieve;

[0016] The output end of the historical database is connected to the input end of the time standard interval calculation unit; the output ends of the time standard interval calculation unit and the test content classification unit are connected to the input end of the artificial intelligence evaluation module.

[0017] According to the above technical solution, the artificial intelligence evaluation module includes a simulation scenario establishment unit and an artificial intelligence evaluation unit;

[0018] The simulation scenario establishment unit is used to establish a simulation scenario for this examination; the artificial intelligence evaluation unit is used to simulate the scenario using machine learning, thereby obtaining an evaluation score for the violation of the examination for each examinee in this examination;

[0019] The output end of the simulation scenario establishment unit is connected to the input end of the artificial intelligence evaluation unit; the output end of the artificial intelligence evaluation unit is connected to the input end of the early warning module.

[0020] According to the above technical solution, the early warning module includes a judgment unit and an early warning unit;

[0021] The judgment unit is used to calculate a new threshold value of the illegal examination assessment score according to the number of candidates and the average value of the illegal examination assessment score; the early warning unit is used to mark the candidates who exceed the threshold value of the illegal examination assessment score in this examination and submit it to the invigilator;

[0022] The output end of the judgment unit is connected to the input end of the early warning unit.

[0023] An examination room monitoring method based on artificial intelligence, the method comprising the following steps:

[0024] S1. Use the camera device that fully covers the examination room to obtain the current examination monitoring video data inside the examination room and create a single file for each candidate;

[0025] S2. Classify according to the test content, including content professional classification and content question type classification, and monitor the time of each candidate's actions when answering questions;

[0026] S3, obtaining a preset test content simulation scenario, performing comparative analysis on each candidate, and calculating the illegal test evaluation score;

[0027] S4. Obtain the assessment score of the illegal examination in step S3, establish a warning threshold based on the number of examinees, mark the examinees who have violated the rules and exceed the threshold, and remind the invigilator to pay special attention to them.

[0028] According to the above technical solution, in step S2, the specific steps are as follows:

[0029] S6-1. Classify the examination contents according to whether or not a draft paper is needed. The examination subjects that require a draft paper for calculation are recorded as the first examination subjects, such as mathematics and physics, and the examination subjects that do not require a draft paper for calculation are recorded as the second examination subjects, such as Chinese and English.

[0030] S6-2. Classify the test content question types, using multiple-choice questions and non-multiple-choice questions as the classification basis, obtain the test paper layout format, and mark the positions of multiple-choice questions and non-multiple-choice questions respectively;

[0031] S6-3. Obtain the time of each action of the examinee during the test, and record the time of answering the i-th question on the test paper as t i , then there is the formula:

[0032] t i =a i -a i-1;

[0033] Among them, a irepresents the time it takes the examinee to complete the i-th question on the test paper; a i-1 It represents the time taken by the candidate to complete question i-1 on the test paper;

[0034] That is, the time difference between every two questions is selected as the answering time;

[0035] S6-4. For the first test subject, obtain the draft paper calculation time for the i-th question, denoted as t ij ; For the first and second test subjects, obtain the interval length of filling in the multiple-choice questions, recorded as set B={b1, b2, b3, ..., b n-1}, where b1, b2, b3, ..., b n-1 They correspond to the length of time between filling in every two multiple-choice questions starting from the first question.

[0036] According to the above technical solution, in step S3, the specific steps are as follows:

[0037] S7-1, obtaining a preset test content simulation scenario, wherein the preset test content simulation scenario includes a time interval for answering each question of the examinee, a corresponding time interval for calculating the draft paper, and a time interval for filling in between every two multiple-choice questions;

[0038] S7-2, obtaining a sample of examinees, wherein the sample of examinees includes historical examination data and examination room data of examinees in a region;

[0039] S7-3. According to the ranking of historical test scores in the historical test data, the candidates are classified into three categories: excellent, good, and poor. That is, the candidates ranked at the top m1 are recorded as excellent; the candidates ranked between m1 and m2 are recorded as good; the candidates ranked after m2 are recorded as poor;

[0040] S7-4, randomly select L groups from the three categories of candidates, namely, excellent, good, and poor, for intelligent learning, and obtain the examination room data of the selected candidates, wherein the examination room data includes the answering time of each question and the corresponding draft paper calculation time;

[0041] S7-5, obtain the current test question content, compare the current test question content with the historical test data in the candidate sample in step S7-2 for similarity in question type difficulty, according to the formula:

[0042] ;

[0043] in, represents the similarity between two sets of questions i1 and i2; u represents any feature; k represents the total number of features; represents the feature u of question i1; represents the characteristic u of question i2; R represents the adjustment value, which is set according to the assessment of the question setter;

[0044] S7-6, according to the calculation of step S7-5, obtain historical questions whose similarity with the current test questions is higher than a threshold value N, where N represents the similarity threshold, and obtain the test site data of these historical questions;

[0045] S7-7. Obtain the video data of the real-time test input and calculate the illegal test evaluation score of each candidate as follows:

[0046] Calculate the standard interval for each question:

[0047] For excellent candidates, select the minimum value of the answering time for each question and the corresponding draft paper calculation time, recorded as c1 and d1;

[0048] For poor candidates, select the average of the time taken to answer each question and the corresponding time spent on the draft paper, and record them as c2 and d2;

[0049] Then the standard intervals are [c1,c2] or [c2,c1]; [d1,d2] or [d2,d1], and the standard intervals take valid values ​​and discard the intervals when there is no solution;

[0050] Considering that some students with poor grades do not understand the questions well, that is, they may write randomly when answering the questions, which makes their time very short. In this system, they are easily judged as cheating. Therefore, we set the good and bad grades to intercept the time interval, so as to reduce the system's misjudgment rate;

[0051] The illegal examination assessment score for each candidate is calculated as follows:

[0052] ;

[0053] Among them, P represents the candidate's illegal test assessment score; i represents the question; V represents the total number of questions; w1 represents the coefficient value of the answering time for each question; w2 represents the coefficient value of the corresponding draft paper calculation time; w3 represents the coefficient value of the interval time for filling in multiple-choice questions; Represents the difference between the answer time of question i and the standard interval, that is, the time portion below the standard interval, which is a positive value; if it meets or exceeds the standard interval, it is 0; The difference between the draft paper calculation time of question i and the standard interval, that is, the time portion below the standard interval, is a positive value; if it meets or exceeds the standard interval or there is no draft paper, it is 0; Represents the time portion below the standard selection time threshold, takes a positive value, and is 0 if it meets or exceeds the standard interval;

[0054] The standard selection time threshold is the standard value of the interval between filling in every two multiple-choice questions, and is obtained based on the average filling value of the three categories of candidates: excellent, good, and poor in historical data.

[0055] In the exam, each candidate generally has an answer sheet, a test paper and a draft paper;

[0056] Therefore, a normal examinee usually calculates and writes answers on the scratch paper and the test paper, and submits them to the answer sheet when handing in the test paper. However, if he wears a cheating device, the other side of the device will give the answer directly, and the time is tight. Because the cheater will try to save time as much as possible during the cheating process, he will directly give the answer to each question when answering, and lacks the time to think and calculate on the scratch paper or test paper. Therefore, the time required for the examinee wearing the cheating device must be less than that of the normal examinee, and the examinee's illegal test evaluation score is finally calculated.

[0057] According to the above technical solution, in step S4, the specific steps are as follows:

[0058] S8-1, obtaining the number of people who took the last exam G1 and the threshold value P1 of the illegal exam assessment score in the historical data;

[0059] The illegal examination assessment score threshold P1=P 均值 *y1;

[0060] P 均值 is the average of the evaluation scores of the illegal examinations in the previous examination; y1 is the adjustment coefficient value. According to the constant value set by the system, y1 is greater than 1;

[0061] S8-2. Get the total number of people G2 in this exam and the average value P of the evaluation score of this illegal exam 新 ; Then, according to the formula, calculate the threshold P2 of the illegal examination assessment score for this examination;

[0062] ;

[0063] S8-3, based on the assessment score threshold P2 of this illegal examination, all candidates whose scores are higher than the threshold P2 are marked, and the invigilator is reminded to pay special attention to them.

[0064] In the above steps, first, the threshold of the illegal examination evaluation score is determined based on the average score of the illegal examination evaluation score and the coefficient value, and the coefficient value is a constant value set according to the system situation and the local status of the school. Therefore, due to the different number of people taking the exam each time, the representative ability of the average value will change. When the number of people increases or decreases, the threshold value this time will be different from the previous time. Therefore, y2 is set for adjustment. Y2 is an influence coefficient, that is, the influence of the change in the number of people on the representative data of the average value. Based on this method, the selection of the threshold can be more accurate, which will not increase the burden on the system and ensure that cheating methods are detected.

[0065] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0066] The present invention can provide an examination room monitoring system and method based on artificial intelligence, which monitors various actions of candidates in the process of answering questions in the examination room, obtains video and time, creates a simulation scene, establishes a standard time interval, and uses artificial intelligence to monitor and evaluate the status of candidates, obtains the illegal examination evaluation score of each candidate, and changes the illegal examination evaluation score threshold according to the change of the number of examinees and the change of the average score of the illegal examination, so as to further improve the accuracy of the model; it can effectively combat candidates from using external devices to cheat in the examination, and provide more accurate monitoring services when the shielding equipment in the examination room has problems or cannot be shielded, so as to ensure the fairness and justice of the examination. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0068] Figure 1 It is a flow chart of an examination room monitoring system and method based on artificial intelligence of the present invention;

[0069] Figure 2 It is a schematic diagram of the steps of an examination room monitoring method based on artificial intelligence of the present invention;

[0070] Figure 3 It is a schematic diagram of the steps of calculating the evaluation score of the illegal examination in the examination room monitoring method based on artificial intelligence of the present invention. DETAILED DESCRIPTION

[0071] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0072] See also Figure 1-Figure 3 , the present invention provides a technical solution:

[0073] An examination room monitoring system based on artificial intelligence, the system includes an examination room monitoring module, a real-time video input module, an examination content analysis module, an artificial intelligence evaluation module, and an early warning module;

[0074] The examination room monitoring module is used to monitor the examination room in full coverage to ensure video monitoring of each candidate; the real-time video recording module is used to record and store the monitored real-time video; the examination content analysis module is used to receive the examination paper content information for classification processing; the artificial intelligence evaluation module is used to perform comparative evaluation based on the preset examination behavior information and identify abnormal candidate behavior; the early warning module is used to issue an early warning, and record the candidate information and output it to the administrator port to remind the invigilator to pay special attention;

[0075] The output end of the examination room monitoring module is connected to the input end of the real-time video recording module; the output end of the real-time video recording module is connected to the input end of the examination content analysis module; the output end of the examination content analysis module is connected to the input end of the artificial intelligence evaluation module; the output end of the artificial intelligence evaluation module is connected to the input end of the early warning module.

[0076] The examination room monitoring module includes a recording unit and a time recording unit;

[0077] The recording unit is used to record the process of the examinee answering questions in the examination room; the time recording unit is used to record the time of various actions of the examinee in the process of answering questions;

[0078] The output end of the recording unit is connected to the input end of the time recording unit; the output end of the time recording unit is connected to the input end of the real-time video recording module.

[0079] The real-time video recording module includes a receiving unit and a storage unit;

[0080] The receiving unit is used to receive the real-time test video content recorded in the test room; the storage unit is used to store the video content for subsequent retrieval and analysis;

[0081] The output end of the receiving unit is connected to the input end of the storage unit; the output end of the storage unit is connected to the input end of the test content analysis module.

[0082] The test content analysis module includes a test content classification unit, a time standard interval calculation unit, and a history database;

[0083] The test content classification unit is used to classify the test content, classify the content according to whether a draft paper is needed, and classify the question type according to whether the question type is a multiple-choice question; the time standard interval calculation unit is used to retrieve historical data to calculate the time standard interval of various actions when answering questions; the historical database is used to store historical data for the system to retrieve;

[0084] The output end of the historical database is connected to the input end of the time standard interval calculation unit; the output ends of the time standard interval calculation unit and the test content classification unit are connected to the input end of the artificial intelligence evaluation module.

[0085] The artificial intelligence evaluation module includes a simulation scenario establishment unit and an artificial intelligence evaluation unit;

[0086] The simulation scenario establishment unit is used to establish a simulation scenario for this examination; the artificial intelligence evaluation unit is used to simulate the scenario using machine learning, thereby obtaining an evaluation score for the violation of the examination for each examinee in this examination;

[0087] The output end of the simulation scenario establishment unit is connected to the input end of the artificial intelligence evaluation unit; the output end of the artificial intelligence evaluation unit is connected to the input end of the early warning module.

[0088] The early warning module includes a judgment unit and an early warning unit;

[0089] The judgment unit is used to calculate a new threshold value of the illegal examination assessment score according to the number of candidates and the average value of the illegal examination assessment score; the early warning unit is used to mark the candidates who exceed the threshold value of the illegal examination assessment score in this examination and submit it to the invigilator;

[0090] The output end of the judgment unit is connected to the input end of the early warning unit.

[0091] An examination room monitoring method based on artificial intelligence, the method comprising the following steps:

[0092] S1. Use the camera device that fully covers the examination room to obtain the current examination monitoring video data inside the examination room and create a single file for each candidate;

[0093] S2. Classify according to the test content, including content professional classification and content question type classification, and monitor the time of each candidate's actions when answering questions;

[0094] S3, obtaining a preset test content simulation scenario, performing comparative analysis on each candidate, and calculating the illegal test evaluation score;

[0095] S4. Obtain the assessment score of the illegal examination in step S3, establish a warning threshold based on the number of examinees, mark the examinees who have violated the rules and exceed the threshold, and remind the invigilator to pay special attention to them.

[0096] In step S2, the specific steps are as follows:

[0097] S6-1. Classify the examination contents according to whether or not a draft paper is needed. The examination subjects that require a draft paper for calculation are recorded as the first examination subjects, such as mathematics and physics, and the examination subjects that do not require a draft paper for calculation are recorded as the second examination subjects, such as Chinese and English.

[0098] S6-2. Classify the test content question types, using multiple-choice questions and non-multiple-choice questions as the classification basis, obtain the test paper layout format, and mark the positions of multiple-choice questions and non-multiple-choice questions respectively;

[0099] S6-3. Obtain the time of each action of the examinee during the test, and record the time of answering the i-th question on the test paper as t i , then there is the formula:

[0100] t i =a i -a i-1;

[0101] Among them, a i represents the time it takes the examinee to complete the i-th question on the test paper; a i-1 It represents the time taken by the candidate to complete question i-1 on the test paper;

[0102] S6-4. For the first test subject, obtain the draft paper calculation time for the i-th question, denoted as t ij ; For the first and second test subjects, obtain the interval length of filling in the multiple-choice questions, recorded as set B={b1, b2, b3, ..., b n-1}, where b1, b2, b3, ..., b n-1 They correspond to the length of time between filling in every two multiple-choice questions starting from the first question.

[0103] In step S3, the specific steps are as follows:

[0104] S7-1, obtaining a preset test content simulation scenario, wherein the preset test content simulation scenario includes a time interval for answering each question of the examinee, a corresponding time interval for calculating the draft paper, and a time interval for filling in between every two multiple-choice questions;

[0105] S7-2, obtaining a sample of examinees, wherein the sample of examinees includes historical examination data and examination room data of examinees in a region;

[0106] S7-3. According to the ranking of historical test scores in the historical test data, the candidates are classified into three categories: excellent, good, and poor. That is, the candidates ranked at the top m1 are recorded as excellent; the candidates ranked between m1 and m2 are recorded as good; the candidates ranked after m2 are recorded as poor;

[0107] S7-4, randomly select L groups from the three categories of candidates, namely, excellent, good, and poor, for intelligent learning, and obtain the examination room data of the selected candidates, wherein the examination room data includes the answering time of each question and the corresponding draft paper calculation time;

[0108] S7-5, obtain the current test question content, compare the current test question content with the historical test data in the candidate sample in step S7-2 for similarity in question type difficulty, according to the formula:

[0109] ;

[0110] in, represents the similarity between two sets of questions i1 and i2; u represents any feature; k represents the total number of features; represents the feature u of question i1; represents the characteristic u of question i2; R represents the adjustment value, which is set according to the assessment of the question setter;

[0111] S7-6, according to the calculation of step S7-5, obtain historical questions whose similarity with the current test questions is higher than a threshold value N, where N represents the similarity threshold, and obtain the test site data of these historical questions;

[0112] S7-7. Obtain the video data of the real-time test input and calculate the illegal test evaluation score of each candidate as follows:

[0113] Calculate the standard interval for each question:

[0114] For excellent candidates, select the minimum value of the answering time for each question and the corresponding draft paper calculation time, recorded as c1 and d1;

[0115] For poor candidates, select the average of the time taken to answer each question and the corresponding time spent on the draft paper, and record them as c2 and d2;

[0116] Then the standard intervals are [c1,c2] or [c2,c1]; [d1,d2] or [d2,d1], and the standard intervals take valid values ​​and discard the intervals when there is no solution;

[0117] The illegal examination assessment score for each candidate is calculated as follows:

[0118] ;

[0119] Among them, P represents the candidate's illegal test assessment score; i represents the question; V represents the total number of questions; w1 represents the coefficient value of the answering time for each question; w2 represents the coefficient value of the corresponding draft paper calculation time; w3 represents the coefficient value of the interval time for filling in multiple-choice questions; Represents the difference between the answer time of question i and the standard interval, that is, the time portion below the standard interval, which is a positive value; if it meets or exceeds the standard interval, it is 0; The difference between the draft paper calculation time of question i and the standard interval, that is, the time portion below the standard interval, is a positive value; if it meets or exceeds the standard interval or there is no draft paper, it is 0; Represents the time portion below the standard selection time threshold, takes a positive value, and is 0 if it meets or exceeds the standard interval;

[0120] The standard selection time threshold is the standard value of the interval between filling in every two multiple-choice questions, and is obtained based on the average filling value of the three categories of candidates: excellent, good, and poor in historical data.

[0121] In step S4, the specific steps are as follows:

[0122] S8-1, obtaining the number of people who took the last exam G1 and the threshold value P1 of the illegal exam assessment score in the historical data;

[0123] The illegal examination assessment score threshold P1=P 均值 *y1;

[0124] P 均值 is the average of the evaluation scores of the illegal examinations in the previous examination; y1 is the adjustment coefficient value. According to the constant value set by the system, y1 is greater than 1;

[0125] S8-2. Get the total number of people G2 in this exam and the average value P of the evaluation score of this illegal exam 新 ; Then, according to the formula, calculate the threshold P2 of the illegal examination assessment score for this examination;

[0126] ;

[0127] S8-3, based on the assessment score threshold P2 of this illegal examination, all candidates whose scores are higher than the threshold P2 are marked, and the invigilator is reminded to pay special attention to them.

[0128] In this embodiment:

[0129] Set the current exam to a math exam; get the exam content to be a test paper without multiple-choice questions;

[0130] There are 5 questions in total on the test paper;

[0131] There is a candidate X;

[0132] Monitor candidate X and obtain the time of each action of candidate X during the test;

[0133] The time to answer the five questions on the test paper is recorded as t1-t5;

[0134] The calculation time for the draft paper of the 5 questions on the test paper is recorded as t ij1 -t ij5 ;

[0135] Obtaining a preset test content simulation scenario, wherein the preset test content simulation scenario includes a time interval for answering each question of the examinee, a corresponding time interval for calculating the draft paper, and a time interval for filling in between every two multiple-choice questions;

[0136] Obtaining a sample of examinees, wherein the sample of examinees includes historical examination data and examination room data of examinees in a region;

[0137] According to the ranking of historical test scores in the historical test data, the candidates are classified into three categories: excellent, good, and poor. That is, the candidates ranked at the top m1 are recorded as excellent; the candidates ranked between m1 and m2 are recorded as good; the candidates ranked after m2 are recorded as poor;

[0138] Randomly select L groups from the three categories of candidates: excellent, good, and poor, and conduct intelligent learning to obtain the examination room data of the selected candidates, wherein the examination room data includes the answering time of each question and the corresponding calculation time on the draft paper;

[0139] Get the current exam question content, and compare the current exam question content with the historical exam data in the above candidate sample for question type difficulty similarity, according to the formula:

[0140] ;

[0141] in, represents the similarity between two sets of questions i1 and i2; u represents any feature; k represents the total number of features; represents the feature u of question i1; represents the characteristic u of question i2; R represents the adjustment value, which is set according to the assessment of the question setter;

[0142] Obtain historical questions whose similarity to the current exam questions is higher than a threshold value N, where N represents the similarity threshold, and obtain the exam room data of these historical questions;

[0143] Calculate the standard interval for each question:

[0144] For excellent candidates, the minimum time for answering each question and the corresponding time for calculating on the draft paper are 8 minutes and 6 minutes respectively;

[0145] For poor candidates, the average of the time to answer each question and the corresponding time to calculate on the draft paper is selected, which is recorded as 12min and 9min;

[0146] Then the standard intervals are [8min, 12min] and [6min, 9min] respectively;

[0147] Calculate the irregular exam assessment score of candidate X using the following formula:

[0148] ;

[0149] Among them, P represents the candidate's illegal test assessment score; i represents the question; V represents the total number of questions; w1 represents the coefficient value of the answering time for each question; w2 represents the coefficient value of the corresponding draft paper calculation time; w3 represents the coefficient value of the interval time for filling in multiple-choice questions; Represents the difference between the answer time of question i and the standard interval, that is, the time portion below the standard interval, which is a positive value; if it meets or exceeds the standard interval, it is 0; The difference between the draft paper calculation time of question i and the standard interval, that is, the time portion below the standard interval, is a positive value; if it meets or exceeds the standard interval or there is no draft paper, it is 0; Represents the time portion below the standard selection time threshold, takes a positive value, and is 0 if it meets or exceeds the standard interval;

[0150] Among them, w1=0.6; w2=0.8; w3=2.5;

[0151] Five questions , , They are as follows:

[0152] : 2, 2.5, 1.5, 3, 1;

[0153] : 5, 6, 5, 5, 6;

[0154] : 0, 0, 0, 0, 0;

[0155] Finally, the evaluation score of candidate X for the illegal examination is

[0156] P=0.6*2+0.8*5+0.6*2.5+0.8*6+0.6*1.5+0.8*5+0.6*3+0.8+5+0.6*1+0.8*6=27.6;

[0157] Obtain the number of people who took the last exam G1 and the threshold value of the illegal exam assessment score P1 from the historical data;

[0158] The illegal examination assessment score threshold P1=P 均值 *y1;

[0159] P 均值 is the average of the evaluation scores of the illegal examinations in the previous examination; y1 is the adjustment coefficient value. According to the constant value set by the system, y1 is greater than 1;

[0160] Get the total number of people G2 in this exam and the average value P of the evaluation score of this illegal exam 新 ; Then, according to the formula, calculate the threshold P2 of the illegal examination assessment score for this examination;

[0161] ;

[0162] After calculation, the threshold value P2 is 12.5;

[0163] Since the score of candidate X is higher than the threshold value P2, it is marked and the invigilator is reminded to pay special attention to it.

[0164] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0165] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An examination room monitoring method based on artificial intelligence, characterized in that: The method comprises the following steps: S1. Use the camera device that fully covers the examination room to obtain the current examination monitoring video data inside the examination room and create a single file for each candidate; S2. Classify according to the test content, including content professional classification and content question type classification, and monitor the time of each candidate's actions when answering questions; S3, obtaining a preset test content simulation scenario, performing comparative analysis on each candidate, and calculating the illegal test evaluation score; S4, obtaining the assessment score of the illegal examination in step S3, establishing a warning threshold based on the number of examinees, marking the examinees who have violated the rules and reminding the invigilators to pay special attention to them; In step S2, the specific steps are as follows: S6-1. Classify the exam content by subject, and use whether or not a draft paper is needed as the basis for classification. Exam subjects that require a draft paper for calculation are recorded as the first exam subject; exam subjects that do not require a draft paper for calculation are recorded as the second exam subject; S6-2. Classify the test content question types, using multiple-choice questions and non-multiple-choice questions as the classification basis, obtain the test paper layout format, and mark the positions of multiple-choice questions and non-multiple-choice questions respectively; S6-3. Obtain the time of each action of the examinee during the test, and record the time of answering the i-th question on the test paper as t i , then there is the formula: t i =a i -a i-1; Among them, a i represents the time it takes the examinee to complete the i-th question on the test paper; a i-1 It represents the time taken by the candidate to complete question i-1 on the test paper; S6-4. For the first test subject, obtain the draft paper calculation time for the i-th question, denoted as t ij ; For the first and second test subjects, obtain the interval length of filling in the multiple-choice questions, recorded as set B={b1, b2, b3, ..., b n-1 }, where b1, b2, b3, ..., b n-1 They correspond to the interval between filling in every two multiple-choice questions starting from the first question; In step S3, the specific steps are as follows: S7-1, obtaining a preset test content simulation scenario, wherein the preset test content simulation scenario includes a time interval for answering each question of the examinee, a corresponding time interval for calculating the draft paper, and a time interval for filling in between every two multiple-choice questions; S7-2, obtaining a sample of examinees, wherein the sample of examinees includes historical examination data and examination room data of examinees in a region; S7-3. According to the ranking of historical test scores in the historical test data, the candidates are classified into three categories: excellent, good, and poor. That is, the candidates ranked at the top m1 are recorded as excellent; the candidates ranked between m1 and m2 are recorded as good; the candidates ranked after m2 are recorded as poor; S7-4, randomly select L groups from the three categories of candidates, namely, excellent, good, and poor, for intelligent learning, and obtain the examination room data of the selected candidates, wherein the examination room data includes the answering time of each question and the corresponding draft paper calculation time; S7-5, obtain the current test question content, and compare the current test question content with the historical test data in the candidate sample in step S7-2 for similarity in question type difficulty, according to the formula: ; in, represents the similarity between two sets of questions i1 and i2; u represents any feature; k represents the total number of features; represents the feature u of question i1; represents the characteristic u of question i2; R represents the adjustment value, which is set according to the assessment of the question setter; S7-6, according to the calculation of step S7-5, obtain historical questions whose similarity with the current test questions is higher than a threshold value N, where N represents the similarity threshold, and obtain the test site data of these historical questions; S7-7. Obtain the video data of the real-time test input and calculate the illegal test evaluation score of each candidate as follows: Calculate the standard interval for each question: For excellent candidates, select the minimum value of the answering time for each question and the corresponding draft paper calculation time, recorded as c1 and d1; For poor candidates, select the average of the time taken to answer each question and the corresponding time spent on the draft paper, and record them as c2 and d2; Then the standard intervals are [c1,c2] or [c2,c1]; [d1,d2] or [d2,d1], and the standard intervals take valid values ​​and discard the intervals when there is no solution; The illegal examination assessment score for each candidate is calculated as follows: ; Among them, P represents the candidate's illegal test assessment score; i represents the question; V represents the total number of questions; w1 represents the coefficient value of the answering time for each question; w2 represents the coefficient value of the corresponding draft paper calculation time; w3 represents the coefficient value of the interval time for filling in multiple-choice questions; Represents the difference between the answer time of question i and the standard interval, that is, the time portion below the standard interval, which is a positive value; if it meets or exceeds the standard interval, it is 0; The difference between the calculation time of the draft paper of question i and the standard interval, that is, the time portion below the standard interval, is a positive value; if it meets or exceeds the standard interval or there is no draft paper, it is 0; Represents the time portion below the standard selection time threshold, takes a positive value, and is 0 if it meets or exceeds the standard interval; The standard selection time threshold is the standard value of the interval between filling in every two multiple-choice questions, and is obtained based on the average filling value of excellent, good, and poor candidates in historical data.

2. The method for monitoring an examination room based on artificial intelligence according to claim 1, characterized in that: In step S4, the specific steps are as follows: S8-1, obtaining the number of people who took the last exam G1 and the threshold value P1 of the illegal exam assessment score in the historical data; The illegal examination assessment score threshold P1=P 均值 *y1; P 均值 is the average of the evaluation scores of the illegal examinations in the previous examination; y1 is the adjustment coefficient value. According to the constant value set by the system, y1 is greater than 1; S8-2. Get the total number of people G2 in this exam and the average value P of the evaluation score of this illegal exam 新 ; Then, according to the formula, calculate the threshold P2 of the illegal examination assessment score for this examination; ; S8-3, based on the assessment score threshold P2 of this illegal examination, all candidates whose scores are higher than the threshold P2 are marked, and the invigilator is reminded to pay special attention to them.

3. An examination room monitoring system based on artificial intelligence, using an examination room monitoring method based on artificial intelligence as claimed in claim 1, characterized in that: The system includes an examination room monitoring module, a real-time video input module, an examination content analysis module, an artificial intelligence evaluation module, and an early warning module; The examination room monitoring module is used to monitor the entire examination room and ensure video monitoring of each examinee; The real-time video recording module is used to record and store the monitored real-time video; The test content analysis module is used to receive test paper content information and perform classification processing; The artificial intelligence evaluation module is used to perform comparative evaluation based on preset test behavior information and identify abnormal test taker behavior; the early warning module is used to issue early warnings and input test taker information to the administrator port to remind the invigilator to pay special attention; The output end of the examination room monitoring module is connected to the input end of the real-time video recording module; the output end of the real-time video recording module is connected to the input end of the examination content analysis module; the output end of the examination content analysis module is connected to the input end of the artificial intelligence evaluation module; the output end of the artificial intelligence evaluation module is connected to the input end of the early warning module; The examination room monitoring module includes a recording unit and a time recording unit; The recording unit is used to record the process of the examinee answering questions in the examination room; the time recording unit is used to record the time of various actions of the examinee in the process of answering questions; The output end of the recording unit is connected to the input end of the time recording unit; the output end of the time recording unit is connected to the input end of the real-time video recording module; The real-time video recording module includes a receiving unit and a storage unit; The receiving unit is used to receive real-time test video content recorded in the test room; The storage unit is used to store video content for subsequent retrieval and analysis; The output end of the receiving unit is connected to the input end of the storage unit; the output end of the storage unit is connected to the input end of the test content analysis module.

4. The artificial intelligence-based examination room monitoring system according to claim 3, characterized in that: The test content analysis module includes a test content classification unit, a time standard interval calculation unit, and a history database; The test content classification unit is used to classify the test content, classify the content according to whether a draft paper is needed, and classify the question type according to whether the question type is a multiple-choice question; the time standard interval calculation unit is used to retrieve historical data to calculate the time standard interval of various actions when answering questions; the historical database is used to store historical data for the system to retrieve; The output end of the historical database is connected to the input end of the time standard interval calculation unit; the output ends of the time standard interval calculation unit and the test content classification unit are connected to the input end of the artificial intelligence evaluation module.

5. The artificial intelligence-based examination room monitoring system according to claim 3 is characterized in that: The artificial intelligence evaluation module includes a simulation scenario establishment unit and an artificial intelligence evaluation unit; The simulation scenario establishment unit is used to establish a simulation scenario for this examination; the artificial intelligence evaluation unit is used to simulate the scenario using machine learning, thereby obtaining an evaluation score for the violation of the examination for each examinee in this examination; The output end of the simulation scenario establishment unit is connected to the input end of the artificial intelligence evaluation unit; the output end of the artificial intelligence evaluation unit is connected to the input end of the early warning module.

6. The artificial intelligence-based examination room monitoring system according to claim 5, characterized in that: The early warning module includes a judgment unit and an early warning unit; The judgment unit is used to calculate a new threshold value of the illegal examination assessment score according to the number of candidates and the average value of the illegal examination assessment score; the early warning unit is used to mark the candidates who exceed the threshold value of the illegal examination assessment score in this examination and submit it to the invigilator; The output end of the judgment unit is connected to the input end of the early warning unit.

Citation Information

Patent Citations

  • Intermodulation signal frequency searching apparatus and positioning method

    CN101039289A

  • Face recognition system based on big data analysis

    CN114360035A

  • Intelligent pen test monitoring system and method based on visual inspection

    CN118470645A