Student learning monitoring method and system based on big data
By acquiring multi-dimensional data from online education platforms and utilizing learning behavior and attention prediction models, a learning feedback model is constructed, which solves the problem of existing assessments being too simplistic and enables comprehensive evaluation and personalized feedback of students' learning.
Patent Information
- Application Number
- CN202511058804.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-30
- Publication Date
- 2025-11-14
AI Technical Summary
Existing online education platforms have limited assessment dimensions, making it difficult for teachers or parents to provide comprehensive and accurate guidance on students' learning progress.
By acquiring multi-dimensional data on students on online education platforms, including clicks, touch screens, eye movements, voice, and duration data, we use learning behavior prediction models and attention prediction models to conduct multi-angle evaluations. We then combine random forest models or decision trees to construct learning feedback models and output learning feedback and suggestions.
It enables a comprehensive and accurate assessment of students' learning behavior and attention levels, provides personalized learning feedback and suggestions, and improves the guidance effectiveness of online education platforms.
Smart Images
Figure CN120951279A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of educational data mining technology, and in particular to a student learning monitoring method and system based on big data. Background Technology
[0002] With the continuous development of society, school teaching has gradually shifted from single classroom teaching to a teaching method that combines online teaching and course teaching, thus integrating multimedia online education platforms into daily teaching.
[0003] Online education platforms not only create an open learning environment for students, but also provide teachers with richer teaching methods for their daily teaching, enabling students to gain a more solid understanding and mastery of the teaching content through interaction with the online education platform.
[0004] Currently, existing online education platforms assess students' learning after they finish their studies. However, the assessment dimensions and data are relatively limited, making it difficult for teachers or parents to provide appropriate guidance based on the assessment results. Summary of the Invention
[0005] This invention provides a student learning monitoring method and system based on big data to address the problem of single-assessment methods in current online education platforms.
[0006] In a first aspect, embodiments of the present invention provide a student learning monitoring method based on big data, including:
[0007] Acquire data information of target students when using the online education platform. The data information includes test data, touch screen data, eye movement data, voice data, and duration data between clicking into the target learning module and exiting the target learning module. The target learning module can be any learning module.
[0008] Test data, voice data, and duration data are input into a pre-built learning behavior prediction model to evaluate learning behavior and obtain learning behavior evaluation data.
[0009] Touchscreen data and eye movement data are input into a pre-built attention prediction model to evaluate attention and obtain learning attention evaluation data.
[0010] Learning behavior assessment data and learning attention assessment data are input into a pre-built learning feedback model, which outputs learning feedback and suggestions; the learning feedback model is built based on a random forest model or a decision tree.
[0011] In one possible implementation, the learning behavior prediction model includes multiple sub-learning prediction modules, each of which corresponds to a learning module.
[0012] Test data, voice data, and duration data are input into a pre-built learning behavior prediction model to evaluate learning behavior, resulting in learning behavior evaluation data, including:
[0013] The test data corresponding to the target learning module is input into the written score evaluation unit in the target sub-learning prediction module to obtain the written learning score;
[0014] The speech data corresponding to the target learning module is input into the speech evaluation unit in the target sub-learning prediction module to obtain the speech learning score;
[0015] Input the duration data corresponding to the target learning module into the duration evaluation unit in the target sub-learning prediction module to obtain the learning proficiency score;
[0016] Learning behaviors are assessed based on written learning scores, speech learning scores, and learning proficiency scores, resulting in learning behavior assessment data.
[0017] In one possible implementation, learning behavior is assessed based on written learning scores, spoken learning scores, and learning proficiency scores to obtain learning behavior assessment data, including:
[0018] The written learning score, the voice learning score, and the learning proficiency score are input into the feature interaction graph neural network module in the learning behavior prediction model to obtain learning behavior assessment data.
[0019] In one possible implementation, the attention prediction model includes multiple sub-attention prediction modules, each of which corresponds to a learning module.
[0020] Touchscreen data and eye movement data are input into a pre-built attention prediction model to evaluate attention levels, resulting in learning attention evaluation data, including:
[0021] The touch screen data corresponding to the target learning module is input into the touch screen analysis unit in the target sub-attention prediction module to obtain the first attention feature; where the touch screen data is an image of the moment a finger touches the online education platform;
[0022] The eye motion data corresponding to the target learning module is input into the attention unit in the target sub-attention prediction module to obtain the second attention feature; wherein, the eye motion data is the eye motion trajectory image;
[0023] Based on the first and second attention features, attention is evaluated to obtain learning attention evaluation data.
[0024] In one possible implementation, attention is evaluated based on a first attention feature and a second attention feature to obtain learning attention evaluation data, including:
[0025] The first and second attention features are input into the attention model to obtain the attention assessment data; the attention model is based on a neural network.
[0026] In one possible implementation, the touchscreen data corresponding to the target learning module is input into the touchscreen analysis unit in the target sub-attention prediction module to obtain the first attention feature, including:
[0027] The number of valid touch images is determined based on the standard touch images in the touch analysis unit and the touch data corresponding to the target learning module.
[0028] The primary attention feature is determined based on the number of valid touchscreen images and the proportion of touchscreen data corresponding to the target learning module.
[0029] In one possible implementation, the eye movement data corresponding to the target learning module is input into the attention unit in the target sub-attention prediction module to obtain the second attention feature, including:
[0030] Eye movement features are extracted from the eye movement data corresponding to the target learning module to obtain the fixation time;
[0031] The second attention feature is determined based on the proportion of standard gaze time in the attention unit of the target sub-attention prediction module.
[0032] In one possible implementation, attention is evaluated based on a first attention feature and a second attention feature to obtain learning attention evaluation data, including:
[0033] Based on the first attention feature, the second attention feature, and the preset normal distribution, the learning attention assessment data is determined.
[0034] In one possible implementation, learning attention evaluation data is determined based on a first attention feature, a second attention feature, and a preset normal distribution, including:
[0035] A first mean is determined based on the first and second attention features;
[0036] Based on the first mean and the mean of the preset normal distribution, the learning attention assessment data is determined.
[0037] Secondly, embodiments of the present invention provide a student learning monitoring system based on big data, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect or any possible implementation of the first aspect.
[0038] In this embodiment of the invention, to provide more comprehensive and accurate suggestions to students using online education platforms, data information on the target students' use of the online education platform is first acquired. This data includes test data, touch screen data, eye movement data, voice data, and duration data from the time they click to enter the target learning module to the time they exit the module. Then, by inputting the test data, voice data, and duration data into a pre-built learning behavior prediction model, learning behavior assessment data is obtained. By inputting the touch screen data and eye movement data into a pre-built attention prediction model, learning attention assessment data is obtained. Finally, the learning behavior assessment data and learning attention assessment data are input into a pre-built learning feedback model to output learning feedback and suggestions. Therefore, feedback and suggestions can be provided not only regarding learning itself but also regarding the level of attention during learning. Attached Figure Description
[0039] Figure 1 This is a flowchart illustrating the implementation of the student learning monitoring method based on big data provided in this embodiment of the invention.
[0040] Figure 2 This is a schematic diagram of the structure of the student learning monitoring device based on big data provided in an embodiment of the present invention;
[0041] Figure 3 This is a schematic diagram of the structure of a student learning monitoring system based on big data provided in an embodiment of the present invention. Detailed Implementation
[0042] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
[0043] With the continuous development of artificial intelligence, analyzing students' learning behaviors and outcomes using AI technology has become a common practice. However, as mentioned in the background section, although educational assessments have been introduced into existing online education platforms, the limited data and singular assessment dimensions mean that the results cannot comprehensively reflect students' learning progress.
[0044] In order to reflect students' learning behavior in a multi-dimensional and accurate manner, this invention provides a student learning monitoring method and system based on big data.
[0045] See Figure 1The document illustrates a flowchart of the implementation of a big data-based student learning monitoring method provided in an embodiment of the present invention, detailed below:
[0046] S110. Obtain data information about target students when using online education platforms.
[0047] Since online education platforms have multiple learning modules, in order to more accurately predict students' learning behavior and to accurately evaluate each module, it is necessary to evaluate each learning module separately. Therefore, the data information in this invention includes test data, touch screen data, eye movement data, voice data, and duration data from clicking to enter the target learning module to exiting the target learning module. It should be noted that the target learning module can be any learning module.
[0048] S120. Input the test data, voice data, and duration data into the pre-built learning behavior prediction model to evaluate the learning behavior and obtain learning behavior evaluation data.
[0049] In some embodiments, the learning behavior prediction model includes multiple sub-learning prediction modules, each of which corresponds to a learning module.
[0050] In this embodiment, firstly, the test data corresponding to the target learning module is input into the written score evaluation unit in the target sub-learning prediction module to obtain the written learning score.
[0051] Next, the speech data corresponding to the target learning module is input into the speech evaluation unit in the target sub-learning prediction module to obtain the speech learning score.
[0052] Then, the duration data corresponding to the target learning module is input into the duration evaluation unit in the target sub-learning prediction module to obtain the learning proficiency score.
[0053] Finally, learning behavior was evaluated based on written learning scores, speech learning scores, and learning proficiency scores, resulting in learning behavior evaluation data.
[0054] Specifically, test data mainly refers to the corresponding test data given to students based on their answers when using the target learning module, such as test results and test levels.
[0055] Voice data refers to the voice data that students submit when using the target learning module to learn, based on the needs of answering questions, such as reading aloud and answering questions.
[0056] Duration data refers to the total time from clicking to enter the target learning module to exiting the target learning module.
[0057] The written assessment unit includes all standard data in the target learning module that requires a score or grade. The written assessment unit compares the standard data for each sub-unit's score or grade with the obtained test data to obtain the written learning score.
[0058] The speech evaluation unit includes all the standard speech data required from the target learning module. It compares the standard speech data of each sub-unit with the acquired speech data to obtain a speech learning score. Comparisons can be made based on factors such as completeness, accuracy, and speech standardization to obtain a comprehensive speech learning score.
[0059] The duration evaluation unit includes the time threshold for the slowest time of each sub-unit with duration requirements in the target learning module. The unit also decomposes the obtained duration data to obtain the duration of each sub-unit. Next, based on the time threshold for the slowest time of each sub-unit and the decomposed duration of each sub-unit, the unit calculates the proficiency of each sub-unit. Finally, a learning proficiency score is obtained based on the proficiency of all sub-units.
[0060] In this embodiment, the written learning score, the voice learning score, and the learning proficiency score are input into the feature interaction graph neural network module in the learning behavior prediction model to obtain learning behavior assessment data.
[0061] Feature Interaction Graph Neural Network (Fi-GNN) represents features with multiple feature domains as a graph structure, thereby leveraging graph neural networks to capture the structural relationships between different features and providing excellent model interpretability. Fi-GNN first maps the input sparse vector containing multiple class features into sparse one-hot embedding vectors, and then obtains embedding vectors unique to each feature domain through embedding layers. For each feature containing multiple feature domains, we represent it as a feature graph. In this feature graph, each node corresponds to a feature domain, and different feature domains can interact through edges between them. To model the interaction between any two feature domains, there is an edge connecting any two points in this graph; therefore, modeling the interaction can be transformed into modeling node interactions on the feature graph. Node interactions are modeled by inputting the feature graph into the proposed Fi-GNN. An attention layer is applied to the output of the Fi-GNN for prediction.
[0062] Feature interaction graph neural networks are mainly based on GGNN and feature maps to model the interaction between feature domain pairs.
[0063] When constructing this Feature Interaction Graph Neural Network (Fi-GNN), the first step is to obtain a dataset containing test data, speech data, and duration data as the raw dataset. Then, the raw dataset is preprocessed to obtain the graph adjacency matrix and learning behavior evaluation data. Next, feature interactions are performed using the preprocessed raw dataset, and graph propagation is performed using the resulting adjacency matrix. The learned features obtained after graph propagation are then used to predict the learning behavior evaluation data, thereby training the Feature Interaction Graph Neural Network.
[0064] S130. Input the touch screen data and eye movement data into the pre-built attention prediction model to evaluate the attention and obtain learning attention evaluation data.
[0065] In some embodiments, the attention prediction model includes multiple sub-attention prediction modules, each of which corresponds to a learning module.
[0066] First, the touch screen data corresponding to the target learning module is input into the touch screen analysis unit in the target sub-attention prediction module to obtain the first attention feature.
[0067] Next, the eye movement data corresponding to the target learning module is input into the attention unit in the target sub-attention prediction module to obtain the second attention feature.
[0068] Finally, based on the first and second attention features, the attention was evaluated to obtain learning attention evaluation data.
[0069] In this embodiment, the touchscreen data is an image of the moment a finger touches the online education platform. Students touch the screen when using the online education platform; some touchscreen operations are necessary for learning, while others are unnecessary and can distract them.
[0070] In some embodiments, the touchscreen analysis unit stores standard touchscreen images that must be touched in the learning target module. The number of valid touchscreen images can be determined based on the standard touchscreen images in the touchscreen analysis unit and the touchscreen data corresponding to the target learning module. Then, based on the ratio of the number of valid touchscreen images to the touchscreen data corresponding to the target learning module, the first attention feature is determined.
[0071] In some embodiments, the student's eye movements can also be monitored to reflect their attention during the learning of the learning objective module.
[0072] In this embodiment, eye movement features can be extracted from the eye movement data corresponding to the target learning module to obtain the fixation time. Then, based on the fixation time and the proportion of the standard fixation time in the attention unit of the target sub-attention prediction module, a second attention feature is determined.
[0073] In this embodiment, eye movement data can be videos or images captured during the learning target learning module.
[0074] For example, eye-tracking software can be used to process the acquired eye movement data, extract eye movement features, and obtain fixation time. Alternatively, the OpenPose system can be set up on an online education platform.
[0075] OpenPose is a deep learning-based human pose estimation system that uses convolutional neural networks (CNNs) and part affinity fields (PAFs) to detect the poses of multiple people in images or videos in real time, including key points such as the body, hands, and eyes.
[0076] In some embodiments, the first attention feature and the second attention feature can also be input into the attention model to obtain learning attention assessment data.
[0077] In this embodiment, the attention model can be built based on a neural network. The neural network can be trained using pre-acquired first and second attention features, as well as the corresponding learning attention evaluation data.
[0078] In some embodiments, learning attention assessment data can also be determined based on a first attention feature, a second attention feature, and a preset normal distribution.
[0079] In this embodiment, the preset normal distribution can be constructed based on the attention data obtained in advance, which will not be elaborated here.
[0080] In this embodiment, a first mean can be determined first based on a first attention feature and a second attention feature. Then, learning attention assessment data can be determined based on the first mean and the mean of a preset normal distribution.
[0081] Specifically, the first mean can be determined based on the average of the first attention feature and the second attention feature.
[0082] S140. Input the learning behavior assessment data and learning attention assessment data into the pre-built learning feedback model, and output learning feedback and suggestions.
[0083] Among them, the learning feedback model is built based on the random forest model or decision tree.
[0084] Decision tree models are predictive models based on a tree structure. They approximate the target variable by dividing the dataset into multiple subsets.
[0085] Random forest is a prediction model based on ensemble learning. It obtains the final prediction result by constructing multiple decision tree models and averaging their results in a weighted manner.
[0086] In some embodiments, the learning feedback model can output assessment results for two aspects of a student's learning and attention span. The learning aspect mainly refers to the student's proficiency in knowledge and their thinking ability, providing suggestions on which areas of knowledge are weaker and which need improvement. Attention span refers to the duration of attention during the learning process, along with related suggestions based on that attention span.
[0087] This invention provides a big data-based student learning monitoring method. To offer more comprehensive and accurate suggestions to students using online education platforms, it first acquires data on the target student's usage of the platform. This data includes test data, touchscreen data, eye movement data, voice data, and duration data from entering to exiting the target learning module. Then, by inputting the test data, voice data, and duration data into a pre-built learning behavior prediction model, learning behavior assessment data is obtained. Similarly, by inputting the touchscreen data and eye movement data into a pre-built attention prediction model, learning attention assessment data is obtained. Finally, the learning behavior assessment data and learning attention assessment data are input into a pre-built learning feedback model, outputting learning feedback and suggestions. Therefore, feedback and suggestions can be provided not only regarding learning itself but also regarding the level of attention.
[0088] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.
[0089] The following are device embodiments of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.
[0090] Figure 2 The diagram shows a schematic representation of a student learning monitoring device based on big data provided in an embodiment of the present invention. For ease of explanation, only the parts relevant to the embodiment of the present invention are shown, and are described in detail below:
[0091] like Figure 2As shown, the student learning monitoring device 200 based on big data includes:
[0092] The acquisition module 210 is used to acquire data information of the target student when using the online education platform. The data information includes test data, touch screen data, eye movement data, voice data and duration data between clicking to enter the target learning module and exiting the target learning module. The target learning module can be any learning module.
[0093] The first assessment module 220 is used to input test data, voice data and duration data into a pre-built learning behavior prediction model to assess learning behavior and obtain learning behavior assessment data.
[0094] The second evaluation module 230 is used to input touch screen data and eye movement data into a pre-built attention prediction model to evaluate attention and obtain learning attention evaluation data.
[0095] The feedback module 240 is used to input learning behavior assessment data and learning attention assessment data into a pre-built learning feedback model and output learning feedback and suggestions; wherein, the learning feedback model is built based on a random forest model or a decision tree.
[0096] In one possible implementation, the learning behavior prediction model includes multiple sub-learning prediction modules, each of which corresponds to a learning module.
[0097] The first assessment module 220 is used to input the test data corresponding to the target learning module into the written score evaluation unit in the target sub-learning prediction module to obtain the written learning score;
[0098] The speech data corresponding to the target learning module is input into the speech evaluation unit in the target sub-learning prediction module to obtain the speech learning score;
[0099] Input the duration data corresponding to the target learning module into the duration evaluation unit in the target sub-learning prediction module to obtain the learning proficiency score;
[0100] Learning behaviors are assessed based on written learning scores, speech learning scores, and learning proficiency scores, resulting in learning behavior assessment data.
[0101] In one possible implementation, the first assessment module 220 is used to input the written learning score, the speech learning score, and the learning proficiency score into the feature interaction graph neural network module in the learning behavior prediction model to obtain learning behavior assessment data.
[0102] In one possible implementation, the attention prediction model includes multiple sub-attention prediction modules, each of which corresponds to a learning module.
[0103] The second evaluation module 230 is used to input the touch screen data corresponding to the target learning module into the touch screen analysis unit in the target sub-attention prediction module to obtain the first attention feature; wherein, the touch screen data is an image of the moment when a finger touches the online education platform;
[0104] The eye motion data corresponding to the target learning module is input into the attention unit in the target sub-attention prediction module to obtain the second attention feature; wherein, the eye motion data is the eye motion trajectory image;
[0105] Based on the first and second attention features, attention is evaluated to obtain learning attention evaluation data.
[0106] In one possible implementation, the second evaluation module 230 is used to input the first attention feature and the second attention feature into the attention model to obtain the learned attention evaluation data; wherein, the attention model is based on a neural network.
[0107] In one possible implementation, the second evaluation module 230 is used to determine the number of valid touch screen images based on the standard touch screen images in the touch screen analysis unit and the touch screen data corresponding to the target learning module.
[0108] The primary attention feature is determined based on the number of valid touchscreen images and the proportion of touchscreen data corresponding to the target learning module.
[0109] In one possible implementation, the second evaluation module 230 is used to extract eye movement features from the eye movement data corresponding to the target learning module to obtain the fixation time.
[0110] The second attention feature is determined based on the proportion of standard gaze time in the attention unit of the target sub-attention prediction module.
[0111] In one possible implementation, the second evaluation module 230 is used to determine learning attention evaluation data based on the first attention feature, the second attention feature and a preset normal distribution.
[0112] In one possible implementation, the second evaluation module 230 is used to determine a first mean based on the first attention feature and the second attention feature;
[0113] Based on the first mean and the mean of the preset normal distribution, the learning attention assessment data is determined.
[0114] The student learning monitoring device based on big data provided by this invention aims to offer more comprehensive and accurate suggestions to students using online education platforms. First, it acquires data on the target student's use of the online education platform, including test data, touch screen data, eye movement data, voice data, and duration data from entering to exiting the target learning module. Then, by inputting the test data, voice data, and duration data into a pre-built learning behavior prediction model, learning behavior assessment data is obtained. Similarly, by inputting the touch screen data and eye movement data into a pre-built attention prediction model, learning attention assessment data is obtained. Finally, the learning behavior assessment data and learning attention assessment data are input into a pre-built learning feedback model, outputting learning feedback and suggestions. Therefore, it can provide feedback and suggestions not only on learning itself but also on the level of attention during learning.
[0115] This invention also provides a student learning monitoring system based on big data, such as... Figure 3 As shown, it includes a memory 31 and a processor 30. The memory 31 stores a computer program 32, and the processor 30 executes the computer program to implement the method as described in the above method embodiment.
[0116] In the above embodiments, the descriptions of each embodiment have their own emphasis. Parts not detailed or described in a particular embodiment can be referred to in the relevant descriptions of other embodiments. Unless otherwise specified or in conflict with logic, the terminology and / or descriptions between different embodiments are consistent and can be referenced interchangeably. Technical features in different embodiments can be combined to form new embodiments based on their inherent logical relationships.
[0117] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.
Claims
1. A student learning monitoring method based on big data, characterized in that, include: Acquire data information of target students when using the online education platform. The data information includes test data, touch screen data, eye movement data, voice data, and duration data between clicking into the target learning module and exiting the target learning module. The target learning module can be any learning module. The test data, voice data, and duration data are input into a pre-built learning behavior prediction model to evaluate the learning behavior and obtain learning behavior evaluation data. The touch screen data and eye movement data are input into a pre-built attention prediction model to evaluate attention and obtain learning attention evaluation data. The learning behavior assessment data and the learning attention assessment data are input into a pre-built learning feedback model, which outputs learning feedback and suggestions; wherein, the learning feedback model is built based on a random forest model or a decision tree.
2. The student learning monitoring method based on big data according to claim 1, characterized in that, The learning behavior prediction model includes multiple sub-learning prediction modules, each of which corresponds to a learning module; The step involves inputting the test data, voice data, and duration data into a pre-built learning behavior prediction model to evaluate the learning behavior and obtain learning behavior evaluation data, including: The test data corresponding to the target learning module is input into the written score evaluation unit in the target sub-learning prediction module to obtain the written learning score; The speech data corresponding to the target learning module is input into the speech evaluation unit in the target sub-learning prediction module to obtain a speech learning score; The duration data corresponding to the target learning module is input into the duration evaluation unit in the target sub-learning prediction module to obtain the learning proficiency score; Based on the written learning score, the voice learning score, and the learning proficiency score, the learning behavior is evaluated to obtain learning behavior evaluation data.
3. The student learning monitoring method based on big data according to claim 2, characterized in that, The learning behavior is evaluated based on the written learning score, the speech learning score, and the learning proficiency score to obtain learning behavior evaluation data, including: The written learning score, the voice learning score, and the learning proficiency score are input into the feature interaction graph neural network module in the learning behavior prediction model to obtain learning behavior assessment data.
4. The student learning monitoring method based on big data according to claim 1, characterized in that, The attention prediction model includes multiple sub-attention prediction modules, and each sub-attention prediction module corresponds to a learning module. The process involves inputting the touchscreen data and eye movement data into a pre-built attention prediction model to evaluate attention and obtain learning attention evaluation data, including: The touch screen data corresponding to the target learning module is input into the touch screen analysis unit in the target sub-attention prediction module to obtain the first attention feature; wherein, the touch screen data is an image of the moment when a finger touches the online education platform; The eye motion data corresponding to the target learning module is input into the attention unit in the target sub-attention prediction module to obtain the second attention feature; wherein, the eye motion data is an image of the eye's motion trajectory; Based on the first attention feature and the second attention feature, attention is evaluated to obtain learning attention evaluation data.
5. The student learning monitoring method based on big data according to claim 4, characterized in that, Based on the first attention feature and the second attention feature, attention is evaluated to obtain learning attention evaluation data, including: The first attention feature and the second attention feature are input into the attention model to obtain the learning attention evaluation data; wherein, the attention model is constructed based on a neural network.
6. The student learning monitoring method based on big data according to claim 4 or 5, characterized in that, The step of inputting the touchscreen data corresponding to the target learning module into the touchscreen analysis unit in the target sub-attention prediction module to obtain the first attention feature includes: Based on the standard touch screen images in the touch screen analysis unit and the touch screen data corresponding to the target learning module, the number of valid touch screen images is determined. The first attention feature is determined based on the number of valid touchscreen images and the proportion of touchscreen data corresponding to the target learning module.
7. The student learning monitoring method based on big data according to claim 4 or 5, characterized in that, The step of inputting the eye movement data corresponding to the target learning module into the attention unit in the target sub-attention prediction module to obtain the second attention feature includes: Eye movement features are extracted from the eye movement data corresponding to the target learning module to obtain the fixation time; The second attention feature is determined based on the gaze time and its proportion in the standard gaze time of the attention unit in the target sub-attention prediction module.
8. The student learning monitoring method based on big data according to claim 4, characterized in that, The process of evaluating attention based on the first attention feature and the second attention feature to obtain learning attention evaluation data includes: Based on the first attention feature, the second attention feature, and the preset normal distribution, the learning attention evaluation data is determined.
9. The student learning monitoring method based on big data according to claim 8, characterized in that, The process of determining learning attention assessment data based on the first attention feature, the second attention feature, and a preset normal distribution includes: Based on the first attention feature and the second attention feature, a first mean is determined; Based on the first mean and the mean of the preset normal distribution, the learning attention evaluation data is determined.
10. A student learning monitoring system based on big data, characterized in that, It includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method as described in any one of claims 1 to 9.