Education data processing method and verification system based on privacy calculation

Abstract

The invention provides an educational data processing method and verification system based on privacy calculation, relates to the technical field of electrical digital data processing, can accurately capture the sight direction and fixed duration of a student by combining a common camera with an embedded eye movement tracking algorithm, overcomes the identification error problem of traditional expression identification in a complex environment, and improves the identification accuracy. Accurate evaluation of the real attention state of the student is realized; by constructing the comprehensive attention evaluation index, the attention level of the student is effectively quantified, more scientific and reasonable attention scores are provided based on the mathematical model of normalization processing and the deviation degree coefficient, and the teaching strategy can be fed back in real time and dynamically adjusted, and the teaching effect can be optimized.

Description

Technical Field

[0001] The present invention relates to the field of electronic digital data processing technology, and specifically to an educational data processing method and verification system based on privacy computing. Background Art

[0002] Video analysis technology, a key component of this approach, ingests classroom video data to monitor and analyze student behavior, attention distribution, and interactions in real time, aiming to improve teaching quality and learning outcomes. Furthermore, privacy-preserving computing, a key means of ensuring data security and privacy, has garnered widespread attention and application in data processing in recent years. Privacy-preserving computing utilizes technologies such as encryption algorithms and multi-party secure computation to ensure that sensitive information is not leaked during data analysis and processing, enabling the effective use of data while protecting privacy.

[0003] The existing technology, with publication number CN118036075A, is titled "A Comprehensive, Easy-to-Operate, Secure and Reliable University Personal Information Anonymization Protection System and Implementation Method." This system addresses key issues currently faced by universities in managing student personal information, such as information leakage and ambiguous authority divisions. Developed using the Python and PHP programming languages, it effectively protects student personal information and precisely controls authority. The system primarily includes multiple functional modules, including a student personal information portal, a university department work platform, a student personal information database, and a student alumni database. It also features an efficient anonymization module, capable of desensitizing sensitive information in accordance with the "University Personal Information System Department Authority Settings," maximizing privacy protection without hindering data analysis and use.

[0004] The shortcomings are:

[0005] Inadequate attention monitoring accuracy: Traditional expression recognition and speech analysis methods struggle to accurately reflect students' true attention status and are significantly affected by ambient light, camera angle, and noise, resulting in significant errors.

[0006] Lack of privacy protection: Existing data processing technologies do not fully consider the protection of student privacy in educational scenarios, and personal behavior data is easily leaked or abused;

[0007] Lack of real-time and scientific nature: There is a lack of effective real-time feedback mechanisms and scientific quantitative indicators in the attention assessment process, making it difficult to dynamically adjust teaching strategies and optimize teaching effects.

[0008] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention

[0009] The purpose of the present invention is to provide an educational data processing method and verification system based on privacy computing to solve the problems raised in the above background technology.

[0010] To achieve the above object, the present invention provides the following technical solutions:

[0011] The educational data processing method based on privacy computing includes the following specific steps:

[0012] Step S1: collecting video data of a target classroom, and dividing the area where students are located in the video data by a preset segmentation algorithm to generate a corresponding target student area sequence set;

[0013] Step S2: determining whether the pupil center position of each student in each target student area is within a preset range based on the preset gaze stability area of the teaching screen, thereby extracting the gaze fixation duration data of each student in the gaze stability area;

[0014] The gaze fixation duration data includes the gaze deviation coefficient and the total gaze fixation duration;

[0015] Step S3: Analyze the gaze fixation duration data of each target student area to obtain the gaze deviation frequency and gaze deviation degree distribution results respectively;

[0016] The distribution results of sight deviation degree are used to divide the target student area sequence set into multiple deviation area subsets with different deviation degrees;

[0017] Step S4: receiving the gaze deviation frequency and total gaze fixation duration of each deviation area subset and performing comprehensive analysis to construct a comprehensive attention evaluation index for each deviation area subset;

[0018] Step S5: Combine the teaching quality feedback data of each deviation area subset with the comprehensive attention evaluation index for comprehensive analysis to generate a verification and tuning model for optimizing the preset range of the stable sight area in the teaching screen and formulating a strategy for verifying the quality and applicability of relevant teaching content.

[0019] A privacy-preserving computing-based educational data processing and verification system, the system being used to execute the privacy-preserving computing-based educational data processing method, specifically comprising:

[0020] Region division module: used to collect video data of the target classroom, and divide the area where the students are located in the video data into regions using a preset segmentation algorithm to generate a corresponding target student area sequence set;

[0021] The judgment and extraction module is used to determine whether the pupil center position of each student in each target student area is within a preset range based on the preset stable sight area of the teaching screen, thereby extracting the fixed sight time data of each student in the stable sight area;

[0022] The gaze fixation duration data includes the gaze deviation coefficient and the total gaze fixation duration;

[0023] Deviation area division module: used to analyze the gaze fixation duration data of each target student area and obtain the gaze deviation frequency and gaze deviation degree distribution results respectively;

[0024] The distribution results of sight deviation degree are used to divide the target student area sequence set into multiple deviation area subsets with different deviation degrees;

[0025] Index construction module: used to receive the gaze deviation frequency and total gaze fixation duration of each deviation area subset for comprehensive analysis to construct a comprehensive attention evaluation index for each deviation area subset;

[0026] Strategy Verification Module: This module is used to conduct a comprehensive analysis based on the teaching quality feedback data of each deviation area subset and the comprehensive attention evaluation index, generate a verification and tuning model to optimize the preset range of the stable line of sight area in the teaching screen, and formulate a strategy to verify the quality and applicability of relevant teaching content.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] 1. Using a common camera combined with an embedded eye-tracking algorithm, it can accurately capture the direction and duration of a student's gaze, overcoming the recognition errors of traditional expression recognition in complex environments and achieving accurate assessment of a student's true attention state.

[0029] 2. By constructing a comprehensive attention evaluation index, we effectively quantify students' attention levels. Based on a mathematical model of normalization and deviation coefficients, we provide a more scientific and reasonable attention score, enabling real-time feedback and dynamic adjustment of teaching strategies to optimize teaching effectiveness.

[0030] 3. Utilizing the privacy computing technology of the Trusted Execution Environment (TEE), sensitive information is encrypted and protected during data collection and processing, ensuring the high security of student privacy and complying with the requirements of relevant laws and regulations. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Schematic diagram of the overall method flow of the present invention;

[0032] Figure 2 This is a block diagram of the overall system module of the present invention. DETAILED DESCRIPTION

[0033] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to specific embodiments.

[0034] It should be noted that, unless otherwise defined, the technical or scientific terms used in the present invention should have the usual meanings understood by people with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative position relationships. When the absolute position of the object being described changes, the relative position relationship may also change accordingly.

[0035] Example 1:

[0036] See also Figure 1 , the present invention provides a technical solution:

[0037] The educational data processing method based on privacy computing includes the following specific steps:

[0038] Step S1: The edge server collects video data of the target classroom, and divides the area where the students are located in the video data into regions using a preset segmentation algorithm to generate a corresponding target student area sequence set;

[0039] To further explain, the specific implementation steps are as follows:

[0040] 1.1) Install a high-definition camera in the target classroom. The camera must have a resolution of at least 1920×1080 pixels and a frame rate of at least 30 frames per second. Adjust the camera's installation position to ensure full coverage of the target student area.

[0041] 1.2) Using the image processing module, the collected video data is preprocessed. Specifically, a Gaussian mixture model based background segmentation algorithm is used to separate areas unrelated to the students from the video data. Based on the segmentation results, the seating distribution map of the target classroom is delineated.

[0042] 1.3) Based on the pre-defined seating distribution map, each seat is uniquely digitally identified based on the actual location of the student in the target classroom. This digital identification maps the seat's physical location to the target student area in the video data, thereby clearly defining the area corresponding to each student in the video data.

[0043] The target student region sequence set is recorded as {1, 2, ..., i, ..., n}, where i represents the index mark of the target student region and n is the total number of target student regions;

[0044] 1.4) Video data from the target student area is transmitted to the edge server through a real-time monitoring system. The edge server encrypts the video data using an encryption module and then transmits it to the Trusted Execution Environment (TEE) via a secure data channel for data privacy processing and analysis.

[0045] Data encryption transmission and processing within the trusted execution environment (TEE) ensure privacy protection and data security, complying with the requirements of desensitization technology based on privacy computing;

[0046] Step S2: The edge server determines whether the pupil center position of each student in each target student area is within a preset range based on the preset stable sight area of the teaching screen, thereby extracting the fixed sight duration data of each student in the stable sight area;

[0047] The gaze fixation duration data includes the gaze deviation coefficient and the total gaze fixation duration;

[0048] To further explain, the specific implementation steps are as follows:

[0049] The pupil center position uses the embedded eye tracking algorithm to detect the student's face and pupil center position in each target student area;

[0050] The embedded eye tracking algorithm includes a face detection module and a pupil detection algorithm;

[0051] 2.1) Obtain video data of the target pupil area isolated by the trusted execution environment (TEE), call the facial detection module within the trusted execution environment (TEE) to locate the face of the target pupil area i, and use the pupil detection algorithm to determine the pupil center position of the target pupil area i;

[0052] In this embodiment, the face detection module adopts the trained convolutional neural network model MTCNN; the pupil detection algorithm adopts the Haar cascade or deep learning eye movement detection model;

[0053] Description of the facial detection module: It is used to locate the student face in the target pupil area as the basis for further pupil detection; in this embodiment, a trained convolutional neural network model MTCNN (Multi-task Cascaded Convolutional Networks) is used, which is good at simultaneously detecting the coordinates of the face bounding box and key points (such as eyes, nose, mouth, etc.).

[0054] The MTCNN face detection model consists of three cascaded networks, each of which gradually narrows the search range and improves detection accuracy:

[0055] P-Net (Proposal Network): Performs a fast, coarse-grained scan of the input image to generate a set of candidate face regions.

[0056] The output candidate regions are where the faces may appear.

[0057] R-Net (Refinement Network): Further filters and accurately locates the candidate regions generated by P-Net, removes noise and adjusts the bounding box.

[0058] Output the preliminarily optimized face boundary.

[0059] O-Net (Output Network): performs the final refinement on the results generated by R-Net and predicts the key points of the face, such as the eyes;

[0060] Finally, the accurate face bounding box and key point coordinates are output.

[0061] The steps are as follows:

[0062] Input video data: The video frame data of the target student area is obtained from the trusted execution environment (TEE) and provided to the MTCNN model.

[0063] Model processing: Each input frame image will be processed step by step through the three networks of MTCNN to obtain the face bounding box and key point coordinates.

[0064] Output: The face bounding box and key points provide candidate locations for the next step of pupil detection.

[0065] The pupil detection algorithm uses a Haar cascade or deep learning eye movement detection model. The specific logic includes:

[0066] Haar cascade detection: Haar cascade is a method for detecting target areas by training cascade classifiers. It is suitable for lightweight devices and has high speed but limited accuracy.

[0067] According to the coordinates of the key points of the eyes output by the face detection module, the detection area is limited to the eyes;

[0068] Haar cascade detection: Use a pre-trained eye classifier to scan the local candidate area, detect the small area bounding box where the pupil is located, and calculate the center point coordinates as the pupil center;

[0069] 2.2) Based on optical detection using an infrared camera, a pupil tracking algorithm is used to extract the pupil center position of each frame in the video data corresponding to the target pupil area i;

[0070] The pupil center position of the target pupil area i is represented by the coordinate value (x p ,y p );

[0071] The coordinate information of the detected pupil center position is calculated and recorded in real time; this embodiment uses the pixel value in the teaching screen to represent the pupil center position;

[0072] It should be noted that the “pixel value in the screen” is to design a coordinate system to locate the teaching screen in pixels, divide the display area of the teaching screen into a two-dimensional pixel grid, and then map the coordinates of the pupil center position to the grid coordinate system of the teaching screen;

[0073] 2.3) Based on the position and functional settings of the teaching screen, the center area of the teaching screen is pre-set as the "stable sight area" of the students corresponding to the target student area;

[0074] The preset range of the vision stability area is defined as: a circular area with the center of the teaching screen as the center and a radius of d;

[0075] The radius of the stable sight area is defined as d≤50 pixels; and d is the maximum deviation radius allowed in the stable sight area;

[0076] 2.4) In the continuous frame sequence of the video data corresponding to the target pupil area i, the pupil center position of each frame is detected in real time, and it is determined whether the pupil center position falls within the stable line of sight area;

[0077] If the pupil center is within the stable line of sight area, the line of sight is considered fixed;

[0078] If the pupil center position is not within the stable sight area, it is considered sight deviation; and the sight deviation degree of the target pupil area i is analyzed;

[0079] Set a preset gaze fixation time window for the students corresponding to target student area i, and record the total number of gaze fixations of target student area i within the preset gaze fixation time window as J;

[0080] When the pupil center position exceeds the stable visual range, a deviation event is recorded, and the total number of deviation events is counted as N1, N1≤J;

[0081] In this embodiment, the sight deviation degree of the target student area i is represented as the sight deviation degree coefficient α i ;

[0082] The sight deviation coefficient is used to quantitatively describe the degree of deviation of the pupil center position relative to the sight stability area;

[0083] Determine (x p ,y p ) is the coordinate of the pupil center; (x c ,y c ) is the center coordinate of the stable sight area;

[0084] The calculation formula of the line of sight deviation coefficient is defined as follows:

[0085]

[0086] Among them, α i is the sight deviation coefficient of target student area i;

[0087] Molecular part (x p,i -x c,i ) 2 +(y p -y c ) 2 is the Euclidean distance, which is used to calculate the actual deviation distance between the pupil center position and the center of the stable sight area;

[0088] The denominator represents the maximum allowable deviation radius of the line of sight stable area. By dividing it by d, the actual deviation distance can be standardized so that different deviation degrees can be directly compared.

[0089] α i =1 indicates that the pupil center position of the target pupil region i is on the boundary of the deviation radius;

[0090] α i <1 means that the pupil center position of the target pupil region i is within the visual stability region or near the boundary;

[0091] α i >1 means that the pupil center position of the target pupil region i is beyond the stable sight area, and the degree of deviation increases with the increase of the value;

[0092] When the pupil center position of the student corresponding to the target student area i is detected to enter the stable sight area, the timing starts immediately. When the pupil center position leaves the stable sight area, the timing stops. The time interval from start to stop is recorded, which is called the fixed sight time T of the student corresponding to the target student area i. fix,i ; This embodiment uses seconds as the unit for cumulative timing;

[0093] The above timing process is calculated with the help of frame timestamp to ensure accurate accumulation of time;

[0094] 2.5) The fixed sight duration T obtained by all timings within the preset fixed sight time window fix,i Summarize to calculate the total fixation time T of the students corresponding to the target student area i fix,i,raw ;T fix,i,raw The specific calculation formula is:

[0095]

[0096] Where T fix,i,j represents the gaze fixation duration of the student corresponding to target student area i in the jth gaze fixation, where J is the total number of gaze fixations within a preset gaze fixation time window; in this embodiment, the preset gaze fixation time window is 2 minutes;

[0097] The total fixation time T of the target student area i fix,i,raw and the line of sight deviation coefficient α i Represented as gaze fixation duration data.

[0098] It should be noted that: d≤50 pixels and T fix,i,j The timing unit parameter design needs to be adjusted by an expert group based on actual conditions such as classroom screen size, resolution, and student distance range to ensure the rationality of the judgment criteria.

[0099] Step S3: The edge server analyzes the gaze fixation duration data of each target student area to obtain the gaze deviation frequency and gaze deviation degree distribution results respectively;

[0100] The distribution results of sight deviation degree are used to divide the target student area sequence set into multiple deviation area subsets with different deviation degrees;

[0101] Further explanation: The target student areas within the deviation area subset are set to be clustered;

[0102] 3.1) The calculation formula for defining the line of sight deviation frequency is:

[0103] Among them, Fp iN1 is the frequency of sight deviation of the students corresponding to the target student area i in the preset sight fixed time window; i is the total number of deviation events in target student region i; J i is the total number of gaze fixations in target student region i;

[0104] 3.2) Based on the target student area sequence set {1, 2, …, i, …, n}, the edge server divides the target classroom seat distribution map into independent parts of the left area set, the middle area set, and the right area set, and analyzes the sight deviation coefficient α of the target student area in each area set. i Value distribution; specific operation steps:

[0105] Obtain the seating distribution map of the target classroom, including the spatial coordinates of each target student area i;

[0106] According to the spatial layout of the classroom seat distribution diagram, the classroom is divided into a left area set, a middle area set, and a right area set; the dividing line is determined according to the geometric center of the classroom or a preset seat grouping standard. In this embodiment:

[0107] x <x center -Δ

[0108] x center -Δ≤x≤x center +Δ

[0109] x>x center +Δ

[0110] Among them, x center is the center x-coordinate of the classroom seating distribution map, and Δ is the offset of the area division;

[0111] For each target student region i in the target student region sequence set {1,2,…,i,…,n}, read its spatial coordinates (x i ,y i ).

[0112] If x i <x center -Δ, then the target student region i is assigned to the left region set;

[0113] If x center -Δ≤x i ≤x center +Δ, then the target student region i is classified into the middle region set;

[0114] If x i >x center +Δ, then the target student region i is classified into the right region set;

[0115] Add each target student area i to the corresponding area set according to the above determination results;

[0116] In the database of the edge server, independent data tables or sub-databases are established for the left region set, the middle region set, and the right region set;

[0117] The divided target student area i and its related data α i The values are stored in the corresponding area sets for subsequent independent analysis;

[0118] 3.3) Calculate the sight deviation coefficient α corresponding to the left area set, the middle area set, and the right area set i Statistical characteristics of , including mean and standard deviation;

[0119] Specific steps:

[0120] Extract all sight deviation coefficients α from the left area set, the middle area set, and the right area set i value;

[0121] The following statistical indicators are calculated for each region set:

[0122] Mean:

[0123]

[0124]

[0125] Standard Deviation:

[0126]

[0127] Among them, S L ∈{1, 2, ..., N L}, S M ∈{1, 2, ..., N M}, S R ∈{1, 2, ..., N R};N L , N M and N R are the total number of target student areas in the left area set, the middle area set, and the right area set;

[0128] According to the calculated mean and standard deviation, the line of sight deviation coefficient α in each area set is determined. i the concentration of values;

[0129] Generate the sight deviation coefficient α for the left, middle and right area sets respectively i Distribution statistics report of values, including mean and standard deviation;

[0130] The statistical results are stored in the distributed analysis database of the edge server for subsequent call and reference.

[0131] 3.4) Apply the preset clustering rules to filter and delete the target student areas that do not meet the clustering rules; to ensure that the target student areas in each area set present the preset clustering form; specific operation steps:

[0132] Based on the statistical feature results, the preset clustering rules are defined for the left region set, the middle region set, and the right region set as follows:

[0133] respectively setting a first preset threshold value and a second preset threshold value for the mean and standard deviation of the statistical features;

[0134] The fuzzy analytic hierarchy process (FAHP) is used to deal with the fuzziness and multi-criteria weight relationship in the decision-making process. The first preset threshold of the mean of each area set and the second preset threshold of the standard deviation of each area set are calculated and evaluated by combining the mean and standard deviation of the data.

[0135] If the mean of each region set is less than the corresponding first preset threshold and the standard deviation is lower than the corresponding second preset threshold, it is considered that there is clustering in the region set;

[0136] It should be noted that the representation values of the first preset threshold and the second preset threshold of each region set are set differently;

[0137] If the sight deviation coefficient α of the target student area i is i If the value is higher than the preset aggregation threshold, the target student area i is deemed not to meet the aggregation rule, and the target student area i that does not meet the aggregation rule is deleted from the corresponding area set;

[0138] The preset aggregation threshold, the specific judgment conditions are expressed as:

[0139] α i >μ X +k·σ X , then delete the target student area i; μ X +k·σ X It is represented as the preset aggregation threshold;

[0140] Among them, X represents the left area set, the middle area set, or the right area set, μ X represents the mean of the corresponding area set, σ X represents the standard deviation of the corresponding region set, and k is a constant that controls the threshold;

[0141] For target student area i that does not meet the aggregation rules, do the following:

[0142] From the region set S L ∈{1, 2, ..., N L}, S M ∈{1, 2, ..., N M}, S R ∈{1, 2, ..., N R}Remove the target student area i;

[0143] Update the target student area sequence set in each area set to ensure data continuity and accuracy.

[0144] 3.6) The set of regions S after the deletion operation L ′∈{1,2,…,N′ L}, S M′ ∈{1,2,…,N′ M}, S R ′∈{1,2,…,N′ R}Update to the edge server's database; S L ', S M ' and S R ′ respectively represent the region set representations of the left region set, the middle region set, and the right region set after deletion operations; and N′ L , N' M and N' R Represent the total number of target student areas after deletion in the left area set, the middle area set, and the right area set respectively;

[0145] This technical solution divides the classroom seating map into three independent zones: left, center, and right. It analyzes the distribution of the target student's gaze deviation coefficient within each zone and applies preset clustering rules to filter and delete them. This not only improves the detail and accuracy of data analysis, but also enhances the system's ability to understand and manage student gaze behavior in different zones. The reasons for choosing this technical solution include:

[0146] Regional analysis: By dividing the classroom into three independent areas: left, center, and right, we can analyze students' gaze deviation behavior in each area in more detail and adapt to possible differences in gaze behavior in different areas.

[0147] Application of clustering rules: Apply preset clustering rules based on the distribution of the line of sight deviation coefficient to ensure that the target student area within the deviation area subset presents a reasonable clustering pattern, thereby improving the reliability of data analysis.

[0148] Quantitative clustering rules are used to achieve accurate screening and deletion of target student areas, avoid errors in subjective judgment, and ensure the objectivity and consistency of the screening process.

[0149] 3.7) Calculate SL ', S M ' and S R 'The sight deviation coefficient α corresponding to the target student area i The mean of L ', μ M ' and μ R ';

[0150] Different deviation degree classification standards are set to divide the left area set, the middle area set, and the right area set into low deviation level, medium deviation level, and high deviation level. The specific logic includes:

[0151] According to the actual teaching environment and needs, set S L ′,S M ′ and S R ′ represents the left region set, the middle region set and the right region set, the mean μ L or μ M ′ or μ R The classification thresholds of ′ are q1 and q2;

[0152] μ L or μ M ′ or μ R When the value of ′ is in the interval (0, q1], the degree of sight deviation is low deviation level;

[0153] μ L or μ M ′ or μ R When the value of ′ is in the interval (q1, q2], the degree of sight deviation is medium deviation level;

[0154] μ L or μ M ' or μ R When the value of ′ is greater than q2, the degree of sight deviation is high deviation level;

[0155] The set of regions that meet the low deviation level is recorded as the low deviation region subset, the set of regions that meet the medium deviation level is recorded as the medium deviation region subset, and the set of regions that meet the high deviation level is recorded as the high deviation region subset.

[0156] Step S4: The edge server receives the gaze deviation frequency and total gaze fixation duration of each deviation area subset and performs comprehensive analysis to construct a comprehensive attention evaluation index for each deviation area subset;

[0157] Further explanation: The calculation formula for the comprehensive attention evaluation index is defined as follows:

[0158]

[0159] Among them, A ris the comprehensive attention evaluation index of the region set r, w1 and w2 are weight coefficients, representing the importance of the frequency of gaze deviation and the total gaze fixation duration in the comprehensive evaluation, respectively, and satisfying w1+w2=1; the values of w1 and w2 are both in the interval (0,1); k1 is the exponential decay coefficient, which is used to adjust the impact of the frequency of gaze deviation on the comprehensive evaluation index; e is the base of the natural logarithm; ln represents the natural logarithm function; it is used to mitigate the contribution of the gaze fixation duration to the comprehensive evaluation index and avoid excessive evaluation index caused by linear growth;

[0160] is the average frequency of line of sight deviation in the region set r; is the average total fixation duration of the region set r; R r is the total number of target student regions in the region set r; i1 represents the index of the target student region;

[0161] Gaze deviation frequency: reflects the frequency of students' gaze deviation in the target student area. The higher the frequency, the easier it is for students to be distracted.

[0162] Total gaze fixation duration: indicates the total duration of students’ gaze fixation in the target student area. The longer the total gaze fixation duration, the higher the degree of student concentration.

[0163] Set the comprehensive attention evaluation index A of the region set r r The value range of is in the interval (0,1);

[0164] When A r The closer the value is to 0, the higher the average frequency of students' gaze deviation in region r, and the shorter the average total gaze fixation duration. This indicates that the students' attention in region r is more distracted. The system can trigger corresponding teaching intervention measures, such as adjusting classroom activity content or teacher guidance methods, to improve students' attention concentration in this region.

[0165] When A r The closer the value is to 1, the lower the average frequency of students' gaze deviation in region set r, and the longer the average total gaze fixation duration; this indicates that the students' attention in region set r is more highly focused. The system can record this region as a high-attention region, which serves as an important reference for future teaching optimization and resource allocation.

[0166] This embodiment initially sets A r The low, middle and high intervals of are (0,0.3), [0.3,0.65), [0.65,1) respectively;

[0167] When A rWhen the value is in the low range (0, 0.3), it means that the degree of distraction of students in the region set r reaches more than 70%;

[0168] When A r When the value is in the middle interval [0.3, 0.65), it means that the degree of distraction of students in the region set r is between 30% and 70%;

[0169] When A r When the value is in the high interval [0.65, 1), it means that the degree of distraction of students in the region set r is below 30%;

[0170] Added: exponential decay term Decreases, resulting in A r Value decreases;

[0171] This indicates that the students’ attention distraction in the region set r has intensified, and the system needs to take measures to improve their concentration.

[0172] Increase: corresponding to A r The value increases, indicating that the students' concentration in the region set r has improved. The system can regard the teaching effect in this region as good and maintain the current teaching strategy.

[0173] Weight coefficient w1 increases: improve The weight in the comprehensive evaluation is enhanced To A r the impact of;

[0174] Exponential decay coefficient k1 increases: it will speed up To A r negative impact, making When A is higher r It is necessary to strengthen the penalty effect for high-frequency gaze deviations, so that the system pays more attention to the areas of students who are frequently distracted.

[0175] A r The formula achieves nonlinear data fusion by combining exponential functions and logarithmic functions. It can effectively suppress the extreme negative evaluation caused by high-frequency deviation events, and moderately amplify the positive impact of fixed gaze duration, thus achieving a balance in comprehensive evaluation.

[0176] In this embodiment, the frequency of gaze deviation and the total duration of gaze fixation correspond to the student's inattention and focus in class, respectively. The formula design reflects the interaction between the two in the actual teaching environment. By quantifying the performance through mathematical means, the system can accurately capture and evaluate the student's attention state.

[0177] A r The exponential decay function in the formula Similar to the attenuation process in physics, it means that as the frequency of sight deviation increases, the attention evaluation index decreases at an exponential rate. This reflects the trend of the attention evaluation index slowly rising with the increase in the duration of gaze fixation, which enables the comprehensive index to respond smoothly to changes in different parameters.

[0178] Step S5: The edge server performs a comprehensive analysis based on the teaching quality feedback data of each deviation area subset and the comprehensive attention evaluation index to generate a verification and tuning model for optimizing the preset range of the stable line of sight area in the teaching screen and formulating a strategy for verifying the quality and applicability of relevant teaching content.

[0179] Further explanation: 4.1) The edge server collects the teaching quality feedback data Tq related to each deviation area subset r , teaching quality feedback data Tq r It is composed of scoring indicators of students’ understanding of teaching content, participation in teaching activities and teaching effectiveness;

[0180]

[0181] Among them, Q i1 Indicates the specific feedback index of the teaching quality by students in the target student area i1; in this embodiment, Q i1 It is represented by the degree of students’ understanding of the teaching content;

[0182] It should be noted that: the definition of Q i1 The calculation formula is as follows:

[0183] Q i1 =w u1 ·U i1 +w p1 ·P i1 +w s1 ·S i1 ;

[0184] Q i1 It represents the specific feedback indicator of teaching quality from students in the target student area i1, and is composed of a weighted combination of the following three factors:

[0185] Students' understanding of teaching content i1 ;

[0186] Students' participation in teaching activities P i1 ;

[0187] Students' rating of teaching effectiveness S i1 ;

[0188] U i1It is calculated through comprehensive data such as classroom test scores and completion of after-class exercises;

[0189]

[0190] Test i1 is the classroom test score of students in the target student area i1, with a value range of (0,1);

[0191] At i1 is the homework score of students in the target student area i1, and its value range is (0,1);

[0192] If there is no homework data, only the class test score U is used. i1 =Test i1 ;

[0193] P i1 The results are obtained by comprehensively counting the number of times students answer questions, participate in discussions, and interactive behavior indicators;

[0194]

[0195] IC i1 is the actual number of interactions of students in the target student area i1 during the course; TI is the total number of possible interactions preset for the teaching activity;

[0196] S i1 Derived from questionnaires, rating scales (e.g. 1 to 5) or course satisfaction feedback data;

[0197]

[0198] RS i1 is the actual rating of the teaching effect by students in the target student area i1;

[0199] Se min and Se max Indicates the minimum and maximum values of the score; if the score range is 1 to 5, then Se min and Se max are 1 and 5 respectively; after standardization, S i1 The value range of is normalized to the interval (0,1);

[0200] Weight coefficient w u1 、w p1 、w s1 is the weight of each feedback sub-indicator, which is used to adjust the impact of each indicator on Q i1 The degree of influence of satisfies the following constraints: w u1 +w p1 +w s1 =1;

[0201] The weight setting is adjusted according to the actual application and is determined by the expert group using the fuzzy analytic hierarchy process;

[0202] In this embodiment, if more emphasis is placed on students’ understanding of the teaching content, w u1 、w p1 、w s1 They are 0.5, 0.3, and 0.2 respectively;

[0203] When Q i1 As it approaches 1, U i1 or P i1 or S i1 The higher; U i1 or P i1 or S i1 They are represented as follows:

[0204] The higher the level of understanding of teaching content, the higher the students' enthusiasm for classroom participation, and the higher the students' subjective satisfaction with teaching results.

[0205] Edge server pair Tq r The min-max normalization method is used to normalize Tq r The value range is limited to the same interval (0,1);

[0206] This embodiment initially sets the teaching quality feedback data The low, middle and high intervals of are (0,0.3), [0.3,0.65), [0.65,1) respectively;

[0207] when When the value is in the low interval (0, 0.3), it means that the students' ratings of teaching quality in the region set r are within 30% of the standard value;

[0208] when When the value is in the middle interval [0.3, 0.65), it means that the teaching quality rating of students in the regional set r is between 30% and 70% of the standard value;

[0209] when When the value is in the high interval [0.65, 1), it means that the teaching quality rating of students in the regional set r is above 70% of the standard value;

[0210] Further explanation: The edge server uses a weighted linear regression model to convert the standardized teaching quality feedback data As the dependent variable, the comprehensive attention evaluation index A r As independent variables, the following verification and tuning model is established:

[0211]

[0212] Among them, β0 is the intercept term of the model; β1 is A r The regression coefficient of A r The degree of impact on teaching quality; η r is the error term of the regional set r, reflecting the deviation between the model prediction and the actual feedback data;

[0213] The edge server assigns different weights w to each region set r according to its importance or data reliability. r , specifically η r ~N(0,σ 2 / w r );

[0214] w r is the weight coefficient of the regional set r. The higher the weight, the more reliable the data of the regional set is and the greater the impact on the model. 2 is the variance of the error term;

[0215] The edge server uses the least squares method to estimate the regression coefficients β0 and β1 to ensure the goodness of fit of the model on the training data;

[0216] Further explanation: Based on the output of the validation and tuning model, the edge server optimizes the preset radius d of the stable gaze area on the teaching screen. Specifically, the radius d of the circular area with a radius of d centered on the teaching screen is adjusted to improve student concentration.

[0217] Select A from the region set r r The set of regions with the smallest values is optimized within a preset range. The specific optimization logic is as follows:

[0218] The edge server is based on the regression coefficient β1 and The radius d is dynamically adjusted as follows;

[0219]

[0220] Among them, d new is the radius of the optimized sightline stabilization area; α is the adjustment coefficient, which controls the amplitude of radius adjustment; To verify the teaching quality feedback data value predicted by the tuning model;

[0221] Setting of adjustment coefficient α:

[0222] Set α to a positive number less than 1 to ensure the gradualness of radius adjustment and the stability of the system; the adjustment is determined by the expert group system using the fuzzy hierarchical analysis method;

[0223] To avoid the radius of the sight stabilization area being too small or too large, set the radius d new The value range is d min ≤d new ≤d max ;

[0224] d min is the minimum radius of the sight line stable area. In this embodiment, d min 0.5 m;

[0225] d max is the maximum radius of the sight line stable area. In this embodiment, d max 2 meters;

[0226] After each adjustment, check d new Is it in d min and d max If it exceeds the range, it will be limited to the corresponding boundary value;

[0227] Further explanation: 4.2) Based on the output of the validation and tuning model, select A from the region set r r The strategy for verifying the quality and applicability of relevant teaching content is formulated based on the set of regions with the smallest values. The specific logic includes:

[0228] The edge server obtains the content difficulty, interactivity and information volume and performs weighted calculation to obtain the teaching content quality index Cq; and the standardized value of the teaching content quality index Cq is recorded as and The value range is the interval (0,1);

[0229] The calculation formula for defining the teaching content quality index Cq is as follows:

[0230] Cq=w d D1+w l I1+w f Formula One

[0231] Among them, w d is the weight of content difficulty D1; w l is the weight of interactivity I1; w f is the weight of the information F1; w d +w l +w f =1;

[0232] The formula for D1 is

[0233] Among them, C2 is the complexity score (Complexity), with a value range of (0,1);

[0234] B2 is the knowledge point breadth score (Breadth), with a value range of (0,1);

[0235] P2 is the depth score of the knowledge point (Depth), and its value range is (0,1);

[0236] The formula for I1 is Among them, A2 is the actual number of activities (Activity Count), which indicates the frequency of interaction between students and teaching content during the teaching process.

[0237] A max It is the maximum number of activities, which represents the total number of all possible interactive events in the teaching content design.

[0238] The formula for F1 is Among them, V2 is the actual information volume (Content Volume), such as the number of words and knowledge points; V max It is the maximum information capacity of teaching content.

[0239] This embodiment initially sets the teaching content quality index The low, middle and high intervals of are (0,0.3), [0.3,0.65), [0.65,1) respectively;

[0240] when When the value is in the low range (0, 0.3), it means that the teaching content quality score is within 30% of the standard value;

[0241] when When the value is in the middle interval [0.3, 0.65), it means that the teaching content quality score is between 30% and 70% of the standard value;

[0242] when When the value is in the high interval [0.65, 1), it means that the teaching content quality score is above 70% of the standard value;

[0243] Furthermore, combined with the comprehensive attention evaluation index A r and teaching quality feedback data Conduct comprehensive assessments;

[0244] The edge server adopts a multivariate weighted decision model, taking into account A r 、 and Cq three indicators to formulate a teaching content adjustment strategy; the strategy formulation formula in this embodiment is as follows:

[0245]

[0246] Among them, S qScoring the teaching content quality and applicability strategy, γ1, γ2, γ3 are the weight coefficients of each indicator, satisfying γ1+γ2+γ3=1; is the standardized value of the teaching content quality indicator;

[0247] Setting S q The value range of is the interval (0,1), and different strategy classification standards are set:

[0248]

[0249] In this embodiment, the value intervals of h1 and h2 are set to be within the range of (0.2, 0.85);

[0250] When A r When the importance is high, this embodiment sets γ1 = 0.4, γ2 = 0.3, and γ3 = 0.3;

[0251] when When the importance is high, this embodiment sets γ1 = 0.25, γ2 = 0.5, and γ3 = 0.25;

[0252] when When the importance is high, this embodiment sets γ1 = 0.2, γ2 = 0.35, and γ3 = 0.45;

[0253] It should be noted that the above settings of γ1, γ2, and γ3 are initial settings, which will be adjusted by the expert group using the fuzzy analytic hierarchy process according to the application situation;

[0254] When 0<S q When r or or In the low range, the first-level teaching content adjustment strategies include increasing interactive sessions, reducing content difficulty, and adding visual aids;

[0255] When h1≤S q When <h2, implement the secondary teaching content adjustment strategy; A in this interval r or or In the middle section, the secondary teaching content adjustment strategies include adjusting the teaching rhythm, adding case analysis, and introducing more practical applications;

[0256] When h2≤S q When <1, implement the teaching content maintenance strategy. r or or ​In the high range, the three-level teaching content adjustment strategy includes recording the teaching content as high-quality content, continuously monitoring its effectiveness, maintaining the current teaching strategy, and considering reuse in similar teaching situations.

[0257] Among them, the adjustment content complexity of the first-level teaching content adjustment strategy is greater than that of the second-level teaching content adjustment strategy;

[0258] The strategy application is as follows: the edge server sends the formulated teaching content quality and applicability strategy to the teaching management system or the teacher side to guide the specific adjustment and optimization implementation of the teaching content.

[0259] After the teaching content is adjusted, the system continues to monitor A r 、 and Cq indicators to form new teaching quality feedback data, enter the next round of verification and tuning model generation, and achieve closed-loop feedback optimization.

[0260] The weighted regression model links the attention evaluation index with the teaching quality feedback data through a linear relationship, enabling the system to rationally adjust the preset range of the stable vision area through mathematical methods to ensure the applicability and effectiveness of the teaching content.

[0261] The radius d of the stable visual area on the teaching screen directly affects the student's visual focus range, which in turn affects the distribution of attention. Guided by mathematical models, optimizing radius d ensures that teaching content is presented within the student's core visual range, improving visual concentration.

[0262] Example 2:

[0263] Please parameter Figure 2 : A privacy-preserving computing-based educational data processing and verification system, the system being used to execute the privacy-preserving computing-based educational data processing method, specifically comprising:

[0264] Region division module: used to collect video data of the target classroom, and divide the area where the students are located in the video data into regions using a preset segmentation algorithm to generate a corresponding target student area sequence set;

[0265] The judgment and extraction module is used to determine whether the pupil center position of each student in each target student area is within a preset range based on the preset stable sight area of the teaching screen, thereby extracting the fixed sight time data of each student in the stable sight area;

[0266] The gaze fixation duration data includes the gaze deviation coefficient and the total gaze fixation duration;

[0267] Deviation area division module: used to analyze the gaze fixation duration data of each target student area and obtain the gaze deviation frequency and gaze deviation degree distribution results respectively;

[0268] The distribution results of sight deviation degree are used to divide the target student area sequence set into multiple deviation area subsets with different deviation degrees;

[0269] Index construction module: used to receive the gaze deviation frequency and total gaze fixation duration of each deviation area subset for comprehensive analysis to construct a comprehensive attention evaluation index for each deviation area subset;

[0270] Strategy Verification Module: This module is used to conduct a comprehensive analysis based on the teaching quality feedback data of each deviation area subset and the comprehensive attention evaluation index, generate a verification and tuning model to optimize the preset range of the stable line of sight area in the teaching screen, and formulate a strategy to verify the quality and applicability of relevant teaching content.

[0271] It should be noted that: All calculation formulas in this application document use regression analysis including but not limited to machine learning algorithms to deeply analyze the relevant parameters collected and identify their natural trends and relationships. Use professional software, such as Python's Scikit-learn library or R language, to automatically generate mathematical models that match the data. Then, objectively evaluate the performance of the model through methods such as cross-validation, and combine continuous feedback and optimization to ensure that the created formula truly reflects the inherent laws of the data, thereby ensuring its effectiveness and accuracy. In all calculation formulas in this application, the parameters in each formula are dimensionally non-dimensionalized within a consistent range to ensure that different physical quantities are compared on the same scale; dimensionless technical means include but are not limited to Min-Max Normalization and Z-Score standardization;

[0272] The technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0273] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0274] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0275] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An educational data processing method based on privacy computing, characterized in that: The specific steps include: Step S1: collecting video data of a target classroom, and dividing the area where students are located in the video data by a preset segmentation algorithm to generate a corresponding target student area sequence set; Step S2: determining whether the pupil center position of each student in each target student area is within a preset range based on the preset gaze stability area of the teaching screen, thereby extracting the gaze fixation duration data of each student in the gaze stability area; The gaze fixation duration data includes the gaze deviation coefficient and the total gaze fixation duration; Step S3: Analyze the gaze fixation duration data of each target student area to obtain the gaze deviation frequency and gaze deviation degree distribution results respectively; The distribution results of sight deviation degree are used to divide the target student area sequence set into multiple deviation area subsets with different deviation degrees; Step S4: receiving the gaze deviation frequency and total gaze fixation duration of each deviation area subset and performing comprehensive analysis to construct a comprehensive attention evaluation index for each deviation area subset; Step S5: Combine the teaching quality feedback data of each deviation area subset with the comprehensive attention evaluation index for comprehensive analysis to generate a verification and tuning model for optimizing the preset range of the stable sight area in the teaching screen and formulating a strategy for verifying the quality and applicability of relevant teaching content.

2. The method for processing educational data based on privacy computing according to claim 1, characterized in that: Based on the pre-defined seating distribution map, each seat is uniquely digitally identified according to the actual location of the students in the target classroom, and the physical location of the seat is mapped to the target student area in the video data through the digital identification; The target student region sequence set is recorded as {1, 2, ..., i, ..., n}, where i represents the index mark of the target student region and n is the total number of target student regions; The video data in the target student area is transmitted to the edge server through the real-time monitoring system. After the video data is encrypted by the encryption module in the edge server, it is transmitted to the trusted execution environment TEE through a secure data channel for data privacy processing and analysis.

3. The method for processing educational data based on privacy computing according to claim 2, characterized in that: The pupil center position uses the embedded eye tracking algorithm to detect the student's face and pupil center position in each target student area; The embedded eye tracking algorithm includes a face detection module and a pupil detection algorithm; Call the facial detection module in the trusted execution environment (TEE) to locate the face of the target pupil area i, and use the pupil detection algorithm to determine the pupil center position of the target pupil area i; Use the pixel value in the teaching screen to represent the pupil center position; Pre-set the center area of the teaching screen as the "stable sight area" for students corresponding to the target student area; The preset range of the vision stability area is defined as: a circular area with the center of the teaching screen as the center and a radius of d; In the continuous frame sequence of the video data corresponding to the target pupil area i, the pupil center position of each frame is detected in real time, and it is determined whether the pupil center position falls within the stable line of sight area; If the pupil center is within the stable line of sight area, the line of sight is considered fixed; If the pupil center is not within the stable vision area, it is considered to be vision deviation; And analyze the results of the degree of sight deviation of the target student area i; Set a preset gaze fixation time window for the students corresponding to target student area i, and record the total number of gaze fixations of target student area i within the preset gaze fixation time window as J; When the pupil center position exceeds the stable visual range, a deviation event is recorded, and the total number of deviation events is counted as N1, N1≤J; The sight deviation degree of the target student area i is represented as the sight deviation degree coefficient α i ; The fixed sight duration T obtained by all timings within the preset fixed sight time window fix,i Summarize to calculate the total fixation time T of the students corresponding to the target student area i fix,i,raw ; The total fixation time T of the target student area i fix,i,raw and the line of sight deviation coefficient α i Represented as gaze fixation duration data.

4. The method for processing educational data based on privacy computing according to claim 3, characterized in that: The target student areas within the deviation area subset are set to be clustered; The calculation formula for defining the line of sight deviation frequency is: Among them, Fp i N1 is the frequency of sight deviation of the students corresponding to the target student area i in the preset sight fixed time window; i is the total number of deviation events in target student region i; J i is the total number of gaze fixations in target student region i; Based on the target student area sequence set {1,2,…,i,…,n}, the edge server divides the target classroom seat distribution map into independent parts of the left area set, the middle area set, and the right area set, and analyzes the sight deviation coefficient α of the target student area in each area set respectively i Value distribution; Calculate the sight deviation coefficient α corresponding to the left area set, the middle area set, and the right area set i Statistical characteristics of , including mean and standard deviation; Apply the preset aggregation rules to filter and delete the target student areas that do not meet the aggregation rules; the preset aggregation rules are as follows: respectively setting a first preset threshold value and a second preset threshold value for the mean and standard deviation of the statistical features; If the mean of each region set is less than the corresponding first preset threshold, and the standard deviation is lower than the corresponding second preset threshold, it is considered that there is clustering in the region set; If the sight deviation coefficient α of the target student area i is i If the value is higher than the preset aggregation threshold, the target student area i is deemed not to meet the aggregation rules, and the target student area i that does not meet the aggregation rules is deleted from the corresponding area set.

5. The method for processing educational data based on privacy computing according to claim 4, characterized in that: The set of regions S after the deletion operation L ′∈{1,2,…,N′ L }, S M ′∈{1,2,…,N′ M }, S R ′∈{1,2,…,N′ R }Update to the edge server's database; S L ′,S M ′ and S R ′ respectively represent the region set representations of the left region set, the middle region set, and the right region set after deletion operations; and N′ L , N′ M and N′ R Represent the total number of target student areas after deletion in the left area set, the middle area set, and the right area set respectively; Calculate S L ′,S M ′ and S R The sight deviation coefficient α corresponding to the target student area in ′ i The mean of L ′,μ M ′ and μ R '; Based on μ L ′,μ M ′ and μ R ′, set different deviation degree classification standards to divide the left area set, the middle area set and the right area set into low deviation level, medium deviation level and high deviation level; The set of regions that meet the low deviation level is recorded as the low deviation region subset, the set of regions that meet the medium deviation level is recorded as the medium deviation region subset, and the set of regions that meet the high deviation level is recorded as the high deviation region subset.

6. The method for processing educational data based on privacy computing according to claim 5, characterized in that: The calculation formula for the comprehensive attention evaluation index is defined as follows: Among them, A r is the comprehensive attention evaluation index of the region set r, w1 and w2 are weight coefficients, and satisfy w1+w2=1; w1 and w2 values are both in the interval (0,1); k1 is the exponential decay coefficient; e is the base of the natural logarithm; ln represents the natural logarithm function; is the average frequency of line of sight deviation in the region set r; is the average total fixation duration of the region set r; R r is the total number of target student regions in the region set r; i1 represents the index of the target student region; Set the comprehensive attention evaluation index A of the region set r r The value range of is in the interval (0,1); When A r The closer the value is to 0, the higher the average frequency of students' gaze deviation in the region set r and / or the shorter the average total gaze fixation time; this means that the students' attention in the region set r is more distracted; When A r The closer the value is to 1, the lower the average frequency of gaze deviation of students in the area set r and / or the longer the average total gaze fixation time; it indicates that the attention of students in the area set r is more highly concentrated.

7. The method for processing educational data based on privacy computing according to claim 6, characterized in that: Collect teaching quality feedback data Tq related to each deviation area subset r , teaching quality feedback data Tq r It is composed of scoring indicators of students’ understanding of teaching content, participation in teaching activities and teaching effectiveness; The edge server uses a weighted linear regression model to convert the standardized teaching quality feedback data As the dependent variable, the comprehensive attention evaluation index A r As independent variables, the following verification and tuning model is established: Among them, β0 is the intercept term of the model; β1 is A r The regression coefficient of A r The degree of impact on teaching quality; η r is the error term of the region set r; Select A from the region set r r The set of regions with the smallest values is optimized within a preset range. The specific optimization logic is as follows: According to the regression coefficients β1 and The radius d is dynamically adjusted as follows; Among them, d new is the radius of the optimized sightline stability area; α is the adjustment coefficient; To verify the teaching quality feedback data value predicted by the tuning model.

8. The method for processing educational data based on privacy computing according to claim 7, characterized in that: Based on the output of the validation and tuning model, select A from the region set r r The strategy for verifying the quality and applicability of relevant teaching content is formulated based on the set of regions with the smallest values. The specific logic includes: The teaching content quality index Cq is obtained by weighted calculation of content difficulty, interactivity and information volume, and the standardized value of the teaching content quality index Cq is recorded as A multivariate weighted decision model is used to comprehensively consider A r 、 and Three indicators to formulate teaching content adjustment strategies; Among them, S q Scoring of teaching content quality and applicability strategy, γ1, γ2, γ3 are weight coefficients of each indicator, satisfying γ1+γ2+γ3=1; Setting S q The value range of is the interval (0,1), and different strategy classification standards are set: When 0<S q When < h1, implement the first-level teaching content adjustment strategy; When h1≤S q When < h2, implement the secondary teaching content adjustment strategy; When h2≤S q When <1, implement the teaching content maintenance strategy.

9. An educational data processing and verification system based on privacy computing, characterized by: The system is used to execute the educational data processing method based on privacy computing according to any one of claims 1 to 8, specifically comprising: Region division module: used to collect video data of the target classroom, and divide the area where the students are located in the video data into regions using a preset segmentation algorithm to generate a corresponding target student area sequence set; The judgment and extraction module is used to determine whether the pupil center position of each student in each target student area is within a preset range based on the preset stable sight area of the teaching screen, thereby extracting the fixed sight time data of each student in the stable sight area; The gaze fixation duration data includes the gaze deviation coefficient and the total gaze fixation duration; Deviation area division module: used to analyze the gaze fixation duration data of each target student area and obtain the gaze deviation frequency and gaze deviation degree distribution results respectively; The distribution results of sight deviation degree are used to divide the target student area sequence set into multiple deviation area subsets with different deviation degrees; Index construction module: used to receive the gaze deviation frequency and total gaze fixation duration of each deviation area subset for comprehensive analysis to construct a comprehensive attention evaluation index for each deviation area subset; Strategy Verification Module: This module is used to conduct a comprehensive analysis based on the teaching quality feedback data of each deviation area subset and the comprehensive attention evaluation index, generate a verification and tuning model to optimize the preset range of the stable line of sight area in the teaching screen, and formulate a strategy to verify the quality and applicability of relevant teaching content.

Images