An online reading quality evaluation method and system applied to students

Abstract

This invention relates to a method and system for evaluating the online reading quality of students. The method includes: Q1. During online reading, collecting data on the user's viewing distance, blink frequency, and heart rate at the same moment, and performing normalization processing to obtain normalized data on the user's viewing distance, blink frequency, and heart rate at the same moment. Then, using an ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm, the method predicts the user's viewing distance, blink frequency, and heart rate at the next moment, obtaining data on the user's viewing distance, blink frequency, and heart rate at the next moment. This invention not only evaluates the user's reading quality from multiple aspects, ensuring the accuracy of the reading quality assessment, but also assesses and reminds users of their reading quality in advance, thereby improving the user's enthusiasm and stability in reading.

Description

Technical Field

[0001] This invention relates to the field of online reading quality assessment technology, and in particular to a method and system for assessing the online reading quality of students. Background Technology

[0002] With the continuous development of technology, online reading has become the preferred reading choice for many users, which in turn has led to the rise of online reading among students. However, online reading lacks supervision, and students often engage in fidgeting, which is detrimental to the development of their concentration. Furthermore, feedback to parents cannot be promptly relayed, making it difficult to make adjustments and help children cultivate good reading habits. Therefore, how to effectively evaluate the quality of students' online reading has become an urgent problem to be solved.

[0003] In the prior art, Chinese patent application (application number: 201510671738.5, publication number: CN105159465B) discloses a method and system for monitoring user reading comprehension, including: acquiring the user's blink frequency, brainwave attention value, and reading speed at the same moment; the user's brainwave attention value is a parameter for measuring whether attention is focused, evaluated based on brainwave monitoring data; calculating the user's reading comprehension at that moment based on the user's blink frequency, brainwave attention value, and reading speed; wherein, the user's reading comprehension is positively correlated with the user's blink frequency and brainwave attention value, and negatively correlated with the user's reading speed. However, this solution cannot provide early warning of the reading quality at the next moment, thus leading to a decline in students' reading quality. At the same time, it does not consider the health, comfort, and stability of reading, lacking a comprehensive approach.

[0004] In the prior art, Chinese patent application (application number: 202310651691.0, publication number: CN116385229A) discloses an online reading monitoring method and system for students, which includes the steps of: when a reading task setting request is received from a mobile terminal, sending a reading task screening table to the mobile terminal; when the book information selected by the mobile terminal based on the reading task screening table and the end time of the reading task are received, identifying the chapter information to be read by the mobile terminal for this reading task based on the book information, and filtering monitoring test data associated with the chapter information to be read from a knowledge point database; the student reads offline based on the chapter information to be read, and when a reading end instruction is received from the student via the mobile terminal, sending the monitoring test data... The data is sent to the mobile device; when the answer data is received from the mobile device based on the monitoring test data, completion analysis data is output based on the answer data. This solution does not start from the user's own state, but only considers the correctness of the answer data, thus ignoring the fact that students' reading attitudes change over time, and the conclusions drawn are one-sided.

[0005] In the prior art, Chinese patent application (application number: 202010963185.1, publication number: CN112084978A) discloses a reading monitoring method, a reading robot, and a computing device. The method includes: the reading robot collecting a first vector from a preset point to a user's target location, and a second vector from the preset point to a picture book; calculating the line-of-sight distance from the user's target location to the picture book based on the first and second vectors; determining whether the line-of-sight distance is within a preset distance range; and if so, determining that the reading method is standard. This solution only considers the user's viewing distance to estimate the user's reading quality, resulting in a simplistic evaluation and inaccurate judgment of the user's reading habits. Summary of the Invention

[0006] In view of the shortcomings of the prior art, the present invention provides a method and system for evaluating the online reading quality of students. It not only evaluates the user's reading quality from multiple aspects to ensure the accuracy of the reading quality assessment, but also evaluates and reminds the user of the reading quality in advance, thereby improving the user's enthusiasm and stability in reading.

[0007] To achieve the above and other related objectives, the present invention provides the following technical solution: A method for assessing the quality of students' online reading, the method comprising: Q1. During the online reading process, data on the user's viewing distance, blink frequency, and heart rate at the same time are collected and normalized to obtain the normalized data on the user's viewing distance, blink frequency, and heart rate at the same time. Q2. Based on the normalized data of the user's viewing distance, blink frequency, and heart rate at the same moment, the ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm is used to predict the user's viewing distance, blink frequency, and heart rate at the next moment, so as to obtain the data of the user's viewing distance, blink frequency, and heart rate at the next moment. Q3. Based on the user's viewing distance, blink frequency, and heart rate data at the next moment, construct the user's reading expectation function G. expect It represents the user's reading expectation value at the next moment and obtains the data information of the user's reading expectation value at the next moment; Q4. Based on the user's reading expectation value at the next moment, and combined with the user's reading speed, construct the user's reading quality evaluation function H. evaluate The system assesses the user's reading quality and obtains data on the assessment value of the user's reading quality.

[0008] Furthermore, the method also includes: Q5. Based on the user's reading quality assessment data, a preset threshold is set. If the user's reading quality assessment value is less than the preset threshold, the user's reading quality is considered poor, and a rest reminder is issued. If the user's reading quality assessment value is greater than the preset threshold, the user's reading quality is considered high, and a refueling reminder is issued.

[0009] Furthermore, in step Q2, the prediction of the user's viewing distance, blink rate, and heart rate at the next moment using the ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm includes: Q21. Input the normalized data information of the user's viewing distance, blink frequency and heart rate at the same time into the ELMAN neural network regression prediction model, initialize the weights and biases of the model, and obtain the data information of the weights and biases of the initialized ELMAN neural network regression prediction model. Q22. Based on the weight and bias parameter data of the initialized ELMAN neural network regression prediction model, the sardine population is initialized, the individual parameters and the maximum number of iterations K are determined, and the data information of the initialized sardine population is obtained. Q23. Based on the data of the initialized sardine population, establish a location update function P for the sardine population. , in, Let represent the position of the i-th sardine individual in the k+1th iteration. Let represent the position of the i-th sardine individual in the population at time k. ω is the weight coefficient of the i-th sardine individual in the k+1th generation. min ω is the lower bound of the weighting coefficients. max Δk is the upper limit of the weight coefficient, a is the decay coefficient, K is the maximum number of iterations, and Δk is the search step size. For the parameters of the i-th sardine population in k iterations, iterative optimization is performed to optimize the weights and bias parameters of the initialized ELMAN neural network regression prediction model, resulting in the optimized weights and bias parameters of the ELMAN neural network regression prediction model. Q24. Based on the weights and bias parameters of the optimized ELMAN neural network regression prediction model, obtain the optimized ELMAN neural network regression prediction model, and input the normalized data information of the user's viewing distance, blink frequency, and heart rate at the same time moment to predict the user's viewing distance, blink frequency, and heart rate at the next time moment, and obtain the data information of the user's viewing distance, blink frequency, and heart rate at the next time moment.

[0010] Furthermore, the ELMAN neural network regression prediction model is a local feedback recurrent neural network, including an input layer, a hidden layer, a continuation layer, and an output layer, net=newelm((input), [11, 1], {"tansig", "purelin"}, "traingdx"), where input is the training set data, tansig is the activation function of the hidden layer, purelin is the activation function of the output layer, and traingdx is the gradient descent function.

[0011] Furthermore, the user's reading expectation function G expect for, , Among them, g t h represents the user's line-of-sight distance at the next moment. t r represents the user's blink frequency at the next moment. t Let λ be the user's heart rate at the next moment, and let ρ and η be any constant parameters between 0 and 1.

[0012] Furthermore, the constraints on the constant parameters λ, ρ, and η are as follows: .

[0013] Furthermore, the user's reading quality evaluation function H evaluate for, , δ health +δ comfortable +δ stable =1, Among them, v look For users' reading speed, δ health δ is a factor representing the proportion of users' reading health. comfortable δ represents the proportion of user reading comfort factors. stable δ is the user's reading stability factor. health δ comfortable and δ stable These are constant parameters ranging from 0 to 1.

[0014] Furthermore, the user's reading speed and the user's reading expectation at the next moment are incorporated into the user's reading quality evaluation function H. evaluate Standardization processes were performed previously.

[0015] To achieve the above and other related objectives, the present invention also provides an evaluation system for the online reading quality of students, for implementing the aforementioned evaluation method for the online reading quality of students, the system comprising: The data acquisition and preprocessing module is used to collect data on the user's visual distance, blink frequency, and heart rate at the same time, and perform normalization processing to obtain the normalized data on the user's visual distance, blink frequency, and heart rate at the same time. The user's viewing distance, blink frequency, and heart rate prediction module is connected to the data acquisition and preprocessing module. It is used to predict the user's viewing distance, blink frequency, and heart rate at the next moment using the ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm, so as to obtain the data information of the user's viewing distance, blink frequency, and heart rate at the next moment. The user's reading expectation calculation module is connected to the user's viewing distance, blink frequency, and heart rate prediction module, and is used to construct the user's reading expectation function G. expect It represents the user's reading expectation value at the next moment and obtains the data information of the user's reading expectation value at the next moment; The user's reading quality assessment module is connected to the user's reading expectation calculation module, and is used to construct the user's reading quality evaluation function H by combining the user's reading speed. evaluate The system assesses the user's reading quality and obtains data on the assessment value of the user's reading quality.

[0016] Furthermore, the system also includes a user reading quality discrimination module, which is connected to the user reading quality evaluation module. This module is used to set a preset threshold. If the user's reading quality evaluation value is less than the preset threshold, the user's reading quality is considered poor, and a rest reminder is issued. If the user's reading quality evaluation value is greater than the preset threshold, the user's reading quality is considered high, and a refueling reminder is issued.

[0017] The present invention has the following positive effects: 1. This invention uses an ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm to predict the user's viewing distance, blink frequency, and heart rate at the next moment, and combines this with the construction of the user's reading expectation function G. expect By representing the user's reading expectations at the next moment and obtaining data on these expectations, we can not only assess and remind the user's reading quality in advance, thereby improving the user's reading enthusiasm and stability, but also promptly reflect this information to parents, enabling them to make adjustments and help children read and develop good reading habits.

[0018] 2. This invention constructs a user reading quality evaluation function H by combining the user's reading speed. evaluateThe system assesses users' reading quality and uses preset thresholds. If a user's reading quality assessment score is below the threshold, their reading quality is considered poor, and a break reminder is issued. Conversely, if the score is above the threshold, their reading quality is considered high, and a "keep going" reminder is issued. This approach not only evaluates reading quality from multiple perspectives to ensure accuracy but also comprehensively considers health, comfort, and stability to further improve reading efficiency. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 This is a flowchart illustrating the ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm of the present invention. Figure 3 This is a schematic diagram of the system framework of the present invention. Detailed Implementation

[0020] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0021] Example 1: As Figure 1 As shown, a method for assessing the quality of students' online reading is provided, the method comprising: Q1. During the online reading process, data on the user's viewing distance, blink frequency, and heart rate at the same time are collected and normalized to obtain the normalized data on the user's viewing distance, blink frequency, and heart rate at the same time. Q2. Based on the normalized data of the user's viewing distance, blink frequency, and heart rate at the same moment, the ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm is used to predict the user's viewing distance, blink frequency, and heart rate at the next moment, so as to obtain the data of the user's viewing distance, blink frequency, and heart rate at the next moment. Q3. Based on the user's viewing distance, blink frequency, and heart rate data at the next moment, construct the user's reading expectation function G. expect It represents the user's reading expectation value at the next moment and obtains the data information of the user's reading expectation value at the next moment; Q4. Based on the user's reading expectation value at the next moment, and combined with the user's reading speed, construct the user's reading quality evaluation function H. evaluate The system assesses the user's reading quality and obtains data on the assessment value of the user's reading quality.

[0022] In this embodiment, the method further includes: Q5. Based on the user's reading quality assessment data, a preset threshold is set. If the user's reading quality assessment value is less than the preset threshold, the user's reading quality is considered poor, and a rest reminder is issued. If the user's reading quality assessment value is greater than the preset threshold, the user's reading quality is considered high, and a refueling reminder is issued.

[0023] In this embodiment, as Figure 2 As shown, in step Q2, the prediction of the user's visual distance, blink frequency, and heart rate at the next moment using the ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm includes: Q21. Input the normalized data information of the user's viewing distance, blink frequency and heart rate at the same time into the ELMAN neural network regression prediction model, initialize the weights and biases of the model, and obtain the data information of the weights and biases of the initialized ELMAN neural network regression prediction model. Q22. Based on the weight and bias parameter data of the initialized ELMAN neural network regression prediction model, the sardine population is initialized, the individual parameters and the maximum number of iterations K are determined, and the data information of the initialized sardine population is obtained. Q23. Based on the data of the initialized sardine population, establish a location update function P for the sardine population. , in, Let represent the position of the i-th sardine individual in the k+1th iteration. Let represent the position of the i-th sardine individual in the population at time k. ω is the weight coefficient of the i-th sardine individual in the k+1th generation. min ω is the lower bound of the weighting coefficients. max Δk is the upper limit of the weight coefficient, a is the decay coefficient, K is the maximum number of iterations, and Δk is the search step size. For the parameters of the i-th sardine population in k iterations, iterative optimization is performed to optimize the weights and bias parameters of the initialized ELMAN neural network regression prediction model, resulting in the optimized weights and bias parameters of the ELMAN neural network regression prediction model. Q24. Based on the weights and bias parameters of the optimized ELMAN neural network regression prediction model, obtain the optimized ELMAN neural network regression prediction model, and input the normalized data information of the user's viewing distance, blink frequency, and heart rate at the same time moment to predict the user's viewing distance, blink frequency, and heart rate at the next time moment, and obtain the data information of the user's viewing distance, blink frequency, and heart rate at the next time moment.

[0024] In this embodiment, the ELMAN neural network regression prediction model is a local feedback recurrent neural network, including an input layer, a hidden layer, a continuation layer, and an output layer, net=newelm((input), [11, 1], {"tansig", "purelin"}, "traingdx"), where input is the training set data, tansig is the activation function of the hidden layer, purelin is the activation function of the output layer, and traingdx is the gradient descent function.

[0025] In this embodiment, the user's reading expectation function G expect for, , Among them, g t h represents the user's line-of-sight distance at the next moment. t r represents the user's blink frequency at the next moment. t Let λ be the user's heart rate at the next moment, and let ρ and η be any constant parameters between 0 and 1.

[0026] In this embodiment, the constraints on the constant parameters λ, ρ, and η are as follows: .

[0027] In this embodiment, the user's reading quality evaluation function H evaluate for, , δ health +δ comfortable +δ stable =1, Among them, v look For users' reading speed, δ health δ is a factor representing the proportion of users' reading health. comfortable δ represents the proportion of user reading comfort factors. stable δ is the user's reading stability factor. health δ comfortable and δ stable These are constant parameters ranging from 0 to 1.

[0028] In this embodiment, the user's reading speed and the user's reading expectation value at the next moment are input into the user's reading quality evaluation function H. evaluate Standardization processes were performed previously.

[0029] In this embodiment, an empirical test was conducted on a provincial smart education cloud platform, covering six standard texts from junior high school Chinese courses, including "The Back View" and "Suzhou Gardens," with a total of 213 students recruited (aged 12-15, male-to-female ratio 1.07:1). Hardware configuration: a Chromebook equipped with a TOF sensor (30 Hz line-of-sight sampling rate), a smart bracelet with an integrated PPG module (50 Hz heart rate sampling rate), and a 1080p USB camera (98.2% blink recognition accuracy, <0.5% false detection rate).

[0030] In the data preprocessing stage, segments with signal gaps exceeding 5 seconds were removed, ultimately yielding 18,642 valid time-series samples (each containing 10 input frames + 1 prediction frame). After 200 iterations, the SSO algorithm determined the optimal number of hidden layer nodes for ELMAN to be 67. Comparative experiments show that SSO-ELMAN reduces the RMSE by 32.7%, 18.4%, and 9.6% compared to traditional BPNN, LSTM, and GRU, respectively.

[0031] Table 1: Comparison of prediction accuracy of different models for physiological parameters at the next time step

[0032] H evaluate The scatter distribution of the values ​​and expert human ratings (out of 5) showed a strong positive correlation (Pearson r = 0.892, p < 0.001); when H evaluate When the score is ≥4.2, the accuracy rate of students' summarizing the main idea of ​​the text reaches 91.3%, which is significantly higher than H. evaluate <3.0 group 62.7%.

[0033] When a teacher was lecturing on the expository text "Chinese Stone Arch Bridges," the system monitored student A's H in real time. evaluate If the value remains below 2.8 in segment 3 (corresponding to excessive viewing distance, a sharp drop in blinking frequency, and low heart rate), an automatic voice prompt will be sent: "Please adjust your posture, move closer to the screen, and try silently reading the key sentences." After intervention, H evaluate It rebounded to 3.9, with its accuracy rate in answering after-class questions increasing by 37%.

[0034] In another case, the system found that although student B's reading speed reached 380 words per minute (exceeding the threshold), G... expectWith a score of only 0.31 (significant fluctuations in visual distance, blinking >25 times / min, heart rate <68 bpm), the student was determined to be in a "pseudo-high-speed - low comprehension" state. The slow-speed intensive reading guidance mode was then activated, which improved the student's final comprehension score by 2.4 points (out of 5).

[0035] In this embodiment, a dynamic quality evaluation closed loop is established. evaluate The scoring function is not static, but evolves in real time as the reading progresses, thus constraining the reasonableness of the speed (v). look ), and anchored cognitive readiness (G) expect ).

[0036] This embodiment boasts high engineering feasibility. All sensors are consumer-grade hardware (TOF modules are low-cost, and PPG wristbands are already widely used), the algorithm is lightweight (SSO-ELMAN single prediction time is <8ms, meeting 60 fps real-time performance), and no individual calibration is required—the normalization strategy and SSO global optimization ensure cross-user generalization capability.

[0037] Example 2: Based on the method for evaluating the quality of online reading for students in Example 1, the present invention will be further explained and described below.

[0038] like Figure 1 As shown, a method for assessing the quality of students' online reading is provided, the method comprising: Q1. During the online reading process, data on the user's viewing distance, blink frequency, and heart rate at the same time are collected and normalized to obtain the normalized data on the user's viewing distance, blink frequency, and heart rate at the same time. Q2. Based on the normalized data of the user's viewing distance, blink frequency, and heart rate at the same moment, the ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm is used to predict the user's viewing distance, blink frequency, and heart rate at the next moment, so as to obtain the data of the user's viewing distance, blink frequency, and heart rate at the next moment. Q3. Based on the user's viewing distance, blink frequency, and heart rate data at the next moment, construct the user's reading expectation function G. expect It represents the user's reading expectation value at the next moment and obtains the data information of the user's reading expectation value at the next moment; Q4. Based on the user's reading expectation value at the next moment, and combined with the user's reading speed, construct the user's reading quality evaluation function H. evaluate The system assesses the user's reading quality and obtains data on the assessment value of the user's reading quality.

[0039] In this embodiment, as Figure 3As shown, this invention provides an evaluation system for students' online reading quality, used to implement the aforementioned evaluation method for students' online reading quality. The system includes: The data acquisition and preprocessing module is used to collect data on the user's visual distance, blink frequency, and heart rate at the same time, and perform normalization processing to obtain the normalized data on the user's visual distance, blink frequency, and heart rate at the same time. The user's viewing distance, blink frequency, and heart rate prediction module is connected to the data acquisition and preprocessing module. It is used to predict the user's viewing distance, blink frequency, and heart rate at the next moment using the ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm, so as to obtain the data information of the user's viewing distance, blink frequency, and heart rate at the next moment. The user's reading expectation calculation module is connected to the user's viewing distance, blink frequency, and heart rate prediction module, and is used to construct the user's reading expectation function G. expect It represents the user's reading expectation value at the next moment and obtains the data information of the user's reading expectation value at the next moment; The user's reading quality assessment module is connected to the user's reading expectation calculation module, and is used to construct the user's reading quality evaluation function H by combining the user's reading speed. evaluate The system assesses the user's reading quality and obtains data on the assessment value of the user's reading quality.

[0040] In this embodiment, the system further includes a user reading quality discrimination module connected to the user reading quality evaluation module. This module is used to set a preset threshold. If the user's reading quality evaluation value is less than the preset threshold, the user's reading quality is considered poor, and a rest reminder is issued. If the user's reading quality evaluation value is greater than the preset threshold, the user's reading quality is considered high, and a refueling reminder is issued.

[0041] In this embodiment, (1) three physiological parameters—visual distance, blink frequency, and heart rate—are collected simultaneously; (2) the temporal neural network structure is optimized using a swarm intelligence-inspired optimization algorithm (SSO); and (3) a cognitive expectation (G) model is constructed. expect ) is an intermediate variable, and the fusion speed adjustment factor (v) look The reading quality evaluation function (H) evaluate Especially G expect The composite structure of the exponential, truncated linear, and sigmoid terms in the function, and the design of its coefficients empirically calibrated by fNIRS, have prominent substantive features and significant progress.

[0042] In this embodiment, the present invention provides a computer-readable storage medium storing a computer program programmed or configured to perform the aforementioned method for assessing the quality of online reading for students.

[0043] In this embodiment, the hardware includes: a color e-ink screen or a high-definition LCD screen with anti-glare and anti-blue light functions, a built-in binocular recognition camera, a built-in speaker, and a stylus; a K12 student-style shell and stand; and a pull-out reading light.

[0044] System software: PAD-side APP, pre-installed on the hardware, allows registration and login, downloading classic works, setting reading plans, and powerful reading comprehension tutoring functions, such as pronunciation of new words, word and sentence comprehension, classic videos, and animated mnemonic devices; Reading quality assessment: Data collection and preprocessing module, user viewing distance, blink frequency and heart rate prediction module, user reading expectation value calculation module, user reading quality assessment module and user reading quality discrimination module, used for viewing distance reminders, attention analysis, facial expression analysis and other functions.

[0045] Home-School App: Parents and teachers can download and install the app, bind the student's account, and then view the student's usage status in real time. They can also receive reading reports from the student in a timely manner, assign reading tasks, set reading plans, goals, and rewards for the student.

[0046] Any references to memory, storage, database, or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0047] In summary, this invention not only evaluates users' reading quality from multiple aspects to ensure the accuracy of the reading quality assessment, but also assesses and reminds users of their reading quality in advance, thereby improving users' enthusiasm and stability in reading.

[0048] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for evaluating the quality of students' online reading, characterized in that, The method includes: Q1. During the online reading process, data on the user's viewing distance, blink frequency, and heart rate at the same time are collected and normalized to obtain the normalized data on the user's viewing distance, blink frequency, and heart rate at the same time. Q2. Based on the normalized data of the user's viewing distance, blink frequency, and heart rate at the same moment, the ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm is used to predict the user's viewing distance, blink frequency, and heart rate at the next moment, so as to obtain the data of the user's viewing distance, blink frequency, and heart rate at the next moment. Q3. Based on the user's viewing distance, blink frequency, and heart rate data at the next moment, construct the user's reading expectation function G. expect It represents the user's reading expectation value at the next moment and obtains the data information of the user's reading expectation value at the next moment; Q4. Based on the user's reading expectation value at the next moment, and combined with the user's reading speed, construct the user's reading quality evaluation function H. evaluate The system assesses the user's reading quality and obtains data on the assessment value of the user's reading quality.

2. The method for evaluating the quality of online reading for students according to claim 1, characterized in that, The method further includes: Q5. Based on the user's reading quality assessment data, a preset threshold is set. If the user's reading quality assessment value is less than the preset threshold, the user's reading quality is considered poor, and a rest reminder is issued. If the user's reading quality assessment value is greater than the preset threshold, the user's reading quality is considered high, and a refueling reminder is issued.

3. The method for evaluating the quality of online reading for students according to claim 1, characterized in that, In step Q2, the prediction of the user's visual distance, blink rate, and heart rate at the next moment using the ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm includes: Q21. Input the normalized data information of the user's viewing distance, blink frequency and heart rate at the same time into the ELMAN neural network regression prediction model, initialize the weights and biases of the model, and obtain the data information of the weights and biases of the initialized ELMAN neural network regression prediction model. Q22. Based on the weight and bias parameter data of the initialized ELMAN neural network regression prediction model, the sardine population is initialized, the individual parameters and the maximum number of iterations K are determined, and the data information of the initialized sardine population is obtained. Q23. Based on the data of the initialized sardine population, establish a location update function P for the sardine population. ， in, Let represent the position of the i-th sardine individual in the k+1th iteration. Let represent the position of the i-th sardine individual in the population at time k. ω is the weight coefficient of the i-th sardine individual in the k+1th generation. min ω is the lower bound of the weighting coefficients. max Δk is the upper limit of the weight coefficient, a is the decay coefficient, K is the maximum number of iterations, and Δk is the search step size. For the parameters of the i-th sardine population in k iterations, iterative optimization is performed to optimize the weights and bias parameters of the initialized ELMAN neural network regression prediction model, resulting in the optimized weights and bias parameters of the ELMAN neural network regression prediction model. Q24. Based on the weights and bias parameters of the optimized ELMAN neural network regression prediction model, obtain the optimized ELMAN neural network regression prediction model, and input the normalized data information of the user's viewing distance, blink frequency, and heart rate at the same time moment to predict the user's viewing distance, blink frequency, and heart rate at the next time moment, and obtain the data information of the user's viewing distance, blink frequency, and heart rate at the next time moment.

4. The method for evaluating the quality of online reading for students according to claim 3, characterized in that: The ELMAN neural network regression prediction model is a local feedback recurrent neural network, including an input layer, a hidden layer, a continuation layer, and an output layer. The model is defined as net = newelm((input), [11, 1], {"tansig", "purelin"}, "traingdx"), where input is the training set data, tansig is the activation function of the hidden layer, purelin is the activation function of the output layer, and traingdx is the gradient descent function.

5. The method for evaluating the quality of online reading for students according to claim 1, characterized in that: The user's reading expectation function G expect for, ， Among them, g t h represents the user's line-of-sight distance at the next moment. t r represents the user's blink frequency at the next moment. t Let λ be the user's heart rate at the next moment, and let ρ and η be any constant parameters between 0 and 1.

6. The method for evaluating the quality of online reading for students according to claim 5, characterized in that: The constraints on the constant parameters λ, ρ, and η are as follows: 。 7. The method for evaluating the quality of online reading for students according to claim 1, characterized in that: The user's reading quality evaluation function H evaluate for, ， d health +d comfortable +d stable =1, Among them, v look For users' reading speed, δ health δ is a factor representing the proportion of users' reading health. comfortable δ represents the proportion of user reading comfort factors. stable δ is the user's reading stability factor. health δ comfortable and δ stable These are constant parameters ranging from 0 to 1.

8. The method for evaluating the quality of online reading for students according to claim 1, characterized in that: The user's reading speed and the user's reading expectation for the next moment are fed into the user's reading quality evaluation function H. evaluate Standardization processes were performed previously.

9. A system for evaluating the quality of students' online reading, characterized in that, The system is used to implement the online reading quality assessment method for students as described in any one of claims 1-8, the system comprising: The data acquisition and preprocessing module is used to collect data on the user's visual distance, blink frequency, and heart rate at the same time, and perform normalization processing to obtain the normalized data on the user's visual distance, blink frequency, and heart rate at the same time. The user's viewing distance, blink frequency, and heart rate prediction module is connected to the data acquisition and preprocessing module. It is used to predict the user's viewing distance, blink frequency, and heart rate at the next moment using the ELMAN neural network regression prediction algorithm based on the sardine optimization algorithm, so as to obtain the data information of the user's viewing distance, blink frequency, and heart rate at the next moment. The user's reading expectation calculation module is connected to the user's viewing distance, blink frequency, and heart rate prediction module, and is used to construct the user's reading expectation function G. expect It represents the user's reading expectation value at the next moment and obtains the data information of the user's reading expectation value at the next moment; The user's reading quality assessment module is connected to the user's reading expectation calculation module, and is used to construct the user's reading quality evaluation function H by combining the user's reading speed. evaluate The system assesses the user's reading quality and obtains data on the assessment value of the user's reading quality.

10. The online reading quality assessment system for students according to claim 9, characterized in that, The system also includes a user reading quality discrimination module, which is connected to the user reading quality evaluation module. This module is used to set a preset threshold. If the user's reading quality evaluation value is less than the preset threshold, the user's reading quality is considered poor, and a rest reminder is issued. If the user's reading quality evaluation value is greater than the preset threshold, the user's reading quality is considered high, and a refueling reminder is issued.

Images