Individualized learning path generation system and method based on brain-computer interface
By integrating smart wearable devices with non-invasive brain-computer interfaces, EEG and eye movement data are collected in real time, and the difficulty of teaching content is dynamically adjusted, the problem of low efficiency of traditional personalized learning methods is solved, the timeliness and accuracy of personalized teaching are achieved, and learning effects are improved.
Patent Information
- Application Number
- CN202510915475.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-03
- Publication Date
- 2025-09-30
AI Technical Summary
Existing personalized learning methods rely on students' active participation in data collection, which leads to low efficiency. Teachers are unable to accurately judge students' understanding of teaching content in a timely manner, which affects teaching effectiveness.
By integrating non-invasive brain-computer interfaces into smart wearable devices, EEG data and facial images are collected in real time. Combined with eye movement data, the difficulty of teaching content can be dynamically adjusted to generate personalized learning paths.
It is possible to timely and accurately judge the learner's level of understanding without the need for active student participation, dynamically adjust the teaching content, reduce learning frustration, and improve learning efficiency and results.
Smart Images
Figure CN120723073A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of computer technology, and in particular to a system and method for generating a personalized learning path based on a brain-computer interface. Background Art
[0002] In modern education, personalized learning is becoming an increasingly important means of improving teaching effectiveness. The goal is to provide customized instructional content to each student based on their individual learning abilities. For example, in remote learning environments, it's difficult for teachers to understand each student's understanding of the instructional content in real time. Therefore, a technology that can automatically adjust instructional content to suit individual student needs is needed.
[0003] In related technologies, personalized learning relies primarily on traditional data collection methods, such as online tests and questionnaires, to assess students' understanding of the teaching content. These methods often require active student participation, are time-consuming, and inefficient. This prevents teachers from accurately assessing students' understanding of the teaching content in a timely manner, thus hindering the effectiveness of personalized teaching. Summary of the Invention
[0004] The present embodiments provide a personalized learning path generation system based on a brain-computer interface. To provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is provided below. This summary is not intended to be a comprehensive review, identify key or important elements, or delineate the scope of protection for these embodiments. Its sole purpose is to present some concepts in a simplified form, serving as a prelude to the detailed description that follows.
[0005] In a first aspect, an embodiment of the present application provides a personalized learning path generation system based on a brain-computer interface, the system comprising: A smart wearable device and a client integrated with a non-invasive brain-computer interface, wherein the smart wearable device and the client communicate with each other; The client is configured to respond to a teaching instruction and obtain a course identifier indicated by the teaching instruction; obtain an initial learning path pre-set for the course identifier, the initial learning path being used to represent a system-default sequence of target knowledge points for the course identifier and a difficulty level of each target knowledge point in the sequence; determine a teaching video corresponding to the initial learning path from a teaching video library corresponding to the course identifier and play the teaching video; and during the playback process, capture a facial image of a target subject in real time using a camera; Smart wearable devices are used to collect the target object's raw EEG data in real time and send it to the client; The client is also used to receive the original EEG data sent by the smart wearable device and store the facial images and original EEG data at each moment; when the playback time of the teaching video is equal to the preset time, the difficulty level of each target knowledge point is adjusted based on the stored facial images and original EEG data to obtain the target learning path of the target object.
[0006] In the second aspect, a method for generating a personalized learning path based on a brain-computer interface is applied to a client, and the method includes: In response to the teaching instruction, obtaining a course identifier indicated by the teaching instruction; Obtaining an initial learning path pre-set for the course identifier, where the initial learning path is used to represent a system default sequence of target knowledge points for the course identifier and a difficulty level of each target knowledge point in the sequence; From the teaching video library corresponding to the course identifier, determine the teaching video corresponding to the initial learning path and play it; During the playback process, the face image of the target object is captured in real time through the camera; Receive the original EEG data sent by the smart wearable device and store the facial image and original EEG data at each moment; When the playback time of the teaching video is equal to the preset time, the difficulty level of each target knowledge point is adjusted based on the stored facial image and original EEG data to obtain the target learning path of the target object.
[0007] In the embodiments of the present application, on the one hand, the EEG data of the target object is collected in real time by integrating a smart wearable device with a non-invasive brain-computer interface. The EEG data can reflect the electrical activity of the central nervous system and has the characteristics of high temporal resolution. It can represent the cognitive state of the learner in real time. By combining the EEG data with facial images that can reflect changes in eye contact, it can timely and accurately judge the student's understanding of the teaching content. Without the need for students to actively participate in the survey, the difficulty of the teaching content can be dynamically adjusted, and customized teaching content can be provided to students in a timely manner to achieve personalized teaching. On the other hand, by dynamically adjusting the teaching content, learners can avoid getting into trouble in learning, reduce frustration in the learning process, and thus improve learning efficiency and learning outcomes.
[0008] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.
[0010] Figure 1This is a schematic diagram of the system structure of a brain-computer interface-based personalized learning path generation system provided in an embodiment of the present application; Figure 2 This is a learning path diagram provided by an embodiment of the present application; Figure 3 This is a schematic diagram of the data structure in a teaching video library corresponding to a course identifier provided in an embodiment of the present application; Figure 4 This is a flowchart of a method for generating a personalized learning path based on a brain-computer interface provided in an embodiment of the present application; Figure 5 1 is a flow chart of a method for training an eye feature point analysis model provided in an embodiment of the present application; Figure 6 This is a structural diagram of a personalized learning path generation device based on a brain-computer interface provided in an embodiment of the present application; Figure 7 This is a structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0011] The following description and the drawings sufficiently illustrate specific embodiments of the application to enable those skilled in the art to practice them.
[0012] It should be clear that the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.
[0013] When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Instead, they are merely examples of systems and methods consistent with certain aspects of the present application, as detailed in the appended claims.
[0014] In the description of this application, it should be understood that the terms "first", "second", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances. In addition, in the description of this application, unless otherwise specified, "multiple" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the previous and subsequent associated objects are in an "or" relationship.
[0015] Currently, the implementation of personalized learning mainly relies on traditional data collection methods, such as online tests and questionnaires to assess students' understanding of teaching content.
[0016] The inventors realized that these methods often require active participation from students, have long cycles, and are inefficient, which results in teachers being unable to accurately judge students' understanding of the teaching content in a timely manner, affecting the effectiveness of personalized teaching.
[0017] In order to solve the above problems, in the embodiments of the present application, on the one hand, the EEG data of the target object is collected in real time by integrating a smart wearable device with a non-invasive brain-computer interface. The EEG data can reflect the electrical activity of the central nervous system and has the characteristics of high temporal resolution. It can represent the cognitive state of the learner in real time. By combining the EEG data with facial images that can reflect changes in eye expression, it can timely and accurately judge the degree of students' understanding of the teaching content, and dynamically adjust the difficulty of the teaching content without the need for students to actively participate in the survey, and provide students with customized teaching content in a timely manner to achieve personalized teaching. On the other hand, by dynamically adjusting the teaching content, learners can avoid getting into trouble in learning, reduce frustration in the learning process, and thus improve learning efficiency and learning outcomes. The following uses an exemplary embodiment to explain in detail.
[0018] It should be noted that this system was designed in strict compliance with educational ethics, with special consideration given to the special needs and comfort of autistic users. It uses a non-invasive monitoring method to ensure that users can learn in a natural and stress-free state.
[0019] See Figure 1 , Figure 1 This is a system structure diagram of a personalized learning path generation system based on a brain-computer interface provided in an embodiment of the present application. The system includes: a smart wearable device and a client integrated with a non-invasive brain-computer interface, and the smart wearable device and the client are communicatively connected.
[0020] Among them, smart wearable devices with integrated non-invasive brain-computer interfaces refer to devices that can be worn directly, such as headphones, hats, headbands or other forms of headwear, in which a non-invasive brain-computer interface (BCI) system is embedded to capture signals generated by brain activity. Smart wearable devices are used to monitor and record the user's brain electrical activities in real time, which are related to the user's thinking, emotions, attention or other cognitive functions. The client is a software application or system that can run on a personal computer, tablet computer, smartphone or other computing device. The client is usually responsible for processing the data received from the smart wearable device, performing data analysis, and providing related functions such as user interface viewing. The data transmission between the smart wearable device and the client can be achieved through Bluetooth, Wi-Fi or other wireless technologies, or through USB or other cables.
[0021] In some embodiments of the present application, the client is used to respond to the teaching instruction and obtain the course identifier indicated by the teaching instruction; obtain the initial learning path pre-set for the course identifier, the initial learning path is used to represent the system default target knowledge point sequence regarding the course identifier and the difficulty level of each target knowledge point in the sequence; determine the teaching video corresponding to the initial learning path from the teaching video library corresponding to the course identifier for playback; during the playback process, capture the facial image of the target object in real time through the camera.
[0022] Among them, the teaching instruction refers to the start teaching command issued by the user for the relevant teaching course. The course identifier is an identifier used to distinguish different teaching courses, which can be a course code, name or other unique identifier. The initial learning path is a pre-set learning roadmap for a certain teaching course, including the target knowledge points arranged in sequence and the difficulty level of each knowledge point. It is a learning plan provided by the system by default and is suitable for most learners. For example, for a programming course, the initial learning path may include knowledge points such as "variables and data types", "control structures", and "functions", and each knowledge point is marked with a difficulty level, such as "simple", "medium", and "difficult". The teaching video library is a collection that stores teaching video content, and these videos are associated with different knowledge points. Facial images refer to images captured by a camera and are used to monitor the learner's concentration in real time.
[0023] In some embodiments of the present application, a smart wearable device is used to collect raw EEG data of a target object in real time and send it to a client.
[0024] Among them, non-invasive brain-computer interface technology research uses external devices such as electrodes to directly measure signals emitted by the human body. It is relatively safe and easy to use, suitable for daily consumption. These signals can be used to decode information such as an individual's thoughts and movement intentions, and then convert them into computer instructions.
[0025] In some embodiments of the present application, the client is also used to receive the original EEG data sent by the smart wearable device, and store the facial images and original EEG data at each moment; when the playback time of the teaching video is equal to the preset time, the difficulty level of each target knowledge point is adjusted based on the stored facial images and original EEG data to obtain the target learning path of the target object.
[0026] Among them, the preset duration can be set based on actual conditions, for example, 10 minutes. The target learning path is a learning roadmap optimized for specific learners. It is dynamically adjusted according to the learner's actual performance and response to adapt to the individual's learning style and ability.
[0027] For example Figure 2 As shown, Figure 2 This is a diagram of the learning path, including the initial learning path and the target learning path. The default initial learning path shows Knowledge Points A, B, and C, all with a difficulty level of 3. The target learning path reduces the difficulty of Knowledge Points A, B, and C by one level, resulting in a difficulty level of 2 for each.
[0028] For example, a user selects a course on the client, such as "Python Programming Basics." The client receives the user's selection and identifies the course as "Python 101." Based on the course ID "Python 101," the system retrieves the initial learning path for the course from the database. The initial learning path is a system-default sequence of knowledge points, for example: Knowledge Point 1: Introduction to Python (Difficulty Level: 1), Knowledge Point 2: Variables and Data Types (Difficulty Level: 2), Knowledge Point 3: Control Flow (Difficulty Level: 3), and Knowledge Point 4: Functions (Difficulty Level: 4). The difficulty level of each knowledge point is determined based on a system-preset difficulty level. For example, 1 represents the easiest and 5 represents the most difficult. The client then searches the video library for the first knowledge point, "Introduction to Python," according to the sequence of the initial learning path, and begins playing it. While the video is playing, the client captures the user's facial image in real time using the front-facing camera. Simultaneously, the user wears a smart wearable device (such as an EEG headband). The device sends the collected raw EEG data to the client and stores it synchronously with the facial image. Assuming the preset duration is 10 minutes, the system will continuously collect facial images and EEG data during these 10 minutes. If, while watching "Introduction to Python," the EEG data indicates a high level of concentration, and the facial image shows focused attention, the system may adjust the difficulty level of that knowledge point from 1 to 0.5 (indicating easier). If, while watching "Control Flow," the EEG data and facial image indicate a lack of understanding, the system may adjust the difficulty level of that knowledge point from 3 to 2 (indicating easier). After these adjustments, the system generates a target learning path for that user.
[0029] The teaching video library corresponding to the course identifier includes a mapping relationship between each teaching knowledge point and a teaching video set of each teaching knowledge point; each target knowledge point is marked with a first difficulty level.
[0030] In some embodiments of the present application, the specific process of determining the teaching video corresponding to the initial learning path from the teaching video library corresponding to the course identifier for playback includes: obtaining a target teaching video set corresponding to each target knowledge point from the mapping relationship, each target teaching video in the target teaching video set being marked with a second difficulty level; matching the first difficulty level marked for each target knowledge point with the second difficulty level marked for each target teaching video in the target teaching video set corresponding to each target knowledge point, so as to identify the target teaching video corresponding to the second difficulty level that is the same as the first difficulty level carried by each target knowledge point, as the teaching video corresponding to the initial learning path; and playing the teaching video corresponding to the initial learning path.
[0031] The second difficulty level is the original difficulty level of each teaching video in the teaching video library, and the first difficulty level is the default difficulty level of each target knowledge point in the initial learning path.
[0032] Among them, the data structure of the teaching video library corresponding to the course identifier is as follows: Figure 3 As shown, a mapping relationship exists between each teaching knowledge point (knowledge point A, knowledge point B, knowledge point C) and the set of teaching videos for each teaching knowledge point (for example, the set of teaching videos for knowledge point A is teaching videos 1, 2, and 3). If the target knowledge points in the initial learning path are knowledge points B and knowledge point C, then this mapping relationship can be used to obtain teaching videos 4, 5, and 6, as well as teaching videos 7, 8, and 9 corresponding to knowledge points B and C. Knowledge points B and C in the initial learning path are marked with a first difficulty level. If the first difficulty levels marked for knowledge points B and C in the initial learning path are level 2 and level 3, respectively, then the video corresponding to knowledge point B in the initial learning path is teaching video 5, and the video corresponding to knowledge point C in the initial learning path is teaching video 9.
[0033] In an embodiment of the present application, in the process of generating a teaching video library corresponding to a course identifier, first, multiple knowledge points set by the user for the course identifier are obtained, a teaching video set set by the user for each knowledge point is received, a mapping relationship is established between each knowledge point and the teaching video set set for each knowledge point, and a teaching video library corresponding to the course identifier is obtained, in which each teaching video in the teaching video set is marked with a first difficulty level.
[0034] In some embodiments of the present application, the specific process of adjusting the difficulty level of each target knowledge point based on the stored facial image and the original EEG data includes: analyzing the eye movement data of the target object through the stored facial image, the eye movement data of the target object including the number of gaze points, gaze duration and blinking frequency of the target object; calculating the concentration of the target object based on the number of gaze points, gaze duration and blinking frequency; performing spectral analysis on the stored original EEG data, and calculating the power spectrum density of different frequency bands as the spectrum analysis results; quantifying the recognition of the target object when watching the teaching video based on the spectrum analysis results; adjusting the difficulty level of each target knowledge point based on the concentration and recognition; wherein, the concentration calculation formula is: concentration = α×number of gaze points + β×average gaze duration-γ×blinking frequency; α, β and γ are weight coefficients.
[0035] Eye movement data refers to the user's eye movement information, including the number of fixations, duration of fixations, and blink frequency, obtained by analyzing facial images captured by a camera. Concentration refers to the degree to which an individual focuses their attention on specific content during the learning process. Spectral analysis is a signal processing technique used to analyze the distribution of signals across different frequency components. Spectral analysis of EEG data can extract the power spectral density of different frequency bands, thereby quantifying the learner's cognitive state. Cognition refers to the comprehensive reflection of an individual's perception, understanding, judgment, and thinking abilities regarding information during the learning process.
[0036] In some embodiments of the present application, the specific process of analyzing the eye movement data of the target object through the stored facial images includes: segmenting the eye area images of the target object at each moment within a preset time length from the stored facial images; inputting the eye area images at each moment into a pre-trained eye feature point analysis model, and outputting the eye feature points of the target object at each moment, the eye feature points including the pupil position and the degree of eyelid opening and closing; analyzing the movement trajectory of the eye feature points at each moment through the pupil position and the degree of eyelid opening and closing to determine the target object's gaze point sequence; counting the gaze duration of each gaze point in the gaze point sequence; counting the target object's blinking frequency within the preset time length; and using the number of gaze points in the gaze point sequence, the gaze duration of each gaze point, and the blinking frequency as the target object's eye movement data.
[0037] For example, while students watch instructional videos, the platform uses its front-facing camera to capture facial images at a certain frequency (e.g., 50Hz). Assuming the preset duration is 10 minutes, the system continuously collects image data during this time. From the stored facial images, the system uses image segmentation techniques to extract images of the student's eye region. For example, a deep learning model (such as a convolutional neural network) is used to identify and segment the eye region. The segmented eye region images are then fed into a pre-trained eye feature point analysis model. This model outputs the student's eye features at each moment, including pupil location and eyelid opening and closing. For example, the model can identify the pupil's center coordinates and the eyelid opening and closing angle. By analyzing changes in pupil position, the system can track the student's eye movements. For example, when the pupil position continuously moves within the image, the system records these changes and determines the student's gaze sequence. The system also calculates the duration of each gaze point in the sequence. For example, the system records the start and end times of each gaze point and calculates the duration of each gaze point. Furthermore, the system counts the student's blink frequency within the preset duration. The above analysis results were integrated into eye movement data: number of fixations (counting the total number of times students fixed their gazes within 10 minutes), fixation duration (calculating the average duration of each fixation point), and blink frequency (counting the number of blinks within 10 minutes).
[0038] In some embodiments of the present application, the specific process of generating a pre-trained eye feature point analysis model includes: collecting multiple historical facial images of the same user browsing client content in a natural light environment; segmenting the historical eye area image of each historical facial image; using a labeling tool to perform feature point labeling on the historical eye area image of each historical facial image to mark the pupil position and eyelid opening and closing degree, thereby obtaining a model training sample; using a neural network to create an eye feature point analysis model; inputting the model training sample into the eye feature point analysis model and outputting the model loss value; when the model loss value reaches the minimum, generating a pre-trained eye feature point analysis model. The trained eye feature point analysis model can analyze the user's eye movement data in real time, such as the number of fixations, fixation duration, and blinking frequency. These data can further provide important basis for applications such as personalized learning and human-computer interaction.
[0039] Historical facial images refer to multiple facial images collected and stored by the system when users browse client content in natural light environments. These images are used for subsequent analysis and model training. Historical eye region images are images of the eye region segmented from historical facial images. The segmentation process is typically performed using image processing techniques (such as OpenCV or Dlib). Labeling tools (such as LabelMe or X-AnyLabeling) can be used to label eye region images, marking feature points such as pupil position and eyelid opening. Models constructed using neural networks (such as convolutional neural networks (CNNs)) are used to analyze eye feature points (such as pupil position and eyelid opening).
[0040] By using the embodiment of the present application, the platform can use the user's historical data to train an efficient eye feature point analysis model, providing technical support for the generation of personalized learning paths.
[0041] In some embodiments of the present application, spectral analysis is performed on the stored original EEG data, and the power spectral density of different frequency bands is calculated. The specific process of the spectral analysis result includes: preprocessing the stored original EEG data to obtain the EEG data to be analyzed in each time window, and the preprocessing includes removing DC bias, removing artifacts and window segmentation; performing fast Fourier transform on the EEG data to be analyzed in each time window to decompose the EEG data to be analyzed into sine waves of different frequencies to obtain a complex representation of the frequency domain signal; determining the frequency axis of the spectrum according to the sampling frequency and the number of data points of the smart wearable device; determining the frequency value of each frequency component in the complex representation of the frequency domain signal through the frequency axis of the spectrum, the frequency axis provides a frequency reference for the complex representation of the frequency domain signal, and the frequency position of each frequency component can be determined; calculating the square of the modulus of the frequency value of each frequency component to obtain the power spectral density of different frequency bands; and using the power spectral density of different frequency bands as the spectral analysis result.
[0042] The Fast Fourier Transform (FFT) is an efficient algorithm for converting time-domain signals into frequency-domain signals. Using the FFT, EEG data can be decomposed into sinusoidal waves of varying frequencies. The power spectral density (PSD) describes the distribution of signal power in the frequency domain. The frequency axis is the frequency reference for frequency-domain signals and is used to determine the frequency position of each frequency component.
[0043] For example, assume EEG data is collected from a smart wearable device at a sampling frequency of 250 Hz and for a one-minute acquisition time. Preprocess the raw EEG data by removing low-frequency drift through a high-pass filter, using independent component analysis (ICA) to remove eye movement artifacts and myoelectric interference, and segmenting the data into multiple time windows, for example, each window is 2 seconds long with a 1-second overlap. Perform an FFT on the EEG data within each time window to convert the time-domain signal into the frequency domain. Based on the sampling frequency (250 Hz) and the number of data points (for example, 500 points per time window), the frequency axis is calculated: frequency axis = (number of data points / sampling frequency) × frequency index.
[0044] In some embodiments of the present application, based on the spectrum analysis results, the specific process of quantifying the target object's cognition when watching the teaching video includes: determining the beta wave related to alertness and cognitive activities; determining the gamma wave related to high-order cognitive activities; obtaining the beta wave power corresponding to the beta wave and the gamma wave power corresponding to the gamma wave from the spectrum analysis results; weightedly summing the beta wave power and the gamma wave power to obtain a cognitive load score, which is used to characterize the target object's degree of understanding of the content when watching the teaching video; and using cognitive load as the target object's cognition when watching the teaching video.
[0045] By adopting the embodiment of the present application, eye movement analysis and EEG spectrum analysis are performed to achieve a comprehensive assessment of learning status from two dimensions: visual attention and brain cognition. This comprehensive analysis method can more accurately reflect the learner's actual learning situation.
[0046] In some embodiments of the present application, the specific process of adjusting the difficulty level of each target knowledge point according to concentration and cognition includes: when the concentration is greater than or equal to a preset concentration threshold and the cognition is less than the preset cognition threshold, reducing the difficulty level of each target knowledge point according to a preset step size; or, when the concentration is greater than or equal to the preset concentration threshold and the cognition is greater than or equal to the preset cognition threshold, determining not to adjust the initial learning path; or, when the concentration is less than the preset concentration threshold and the cognition is greater than or equal to the preset cognition threshold, increasing the difficulty level of each target knowledge point according to a preset step size; or, when the concentration is less than the preset concentration threshold and the cognition is less than the preset cognition threshold, determining not to adjust the initial learning path.
[0047] Among them, if the concentration is greater than or equal to the preset concentration threshold and the cognition is less than the preset cognition threshold, it means that the target object is paying close attention and is listening carefully to the class, but cannot fully understand the content of the teaching video. This may be caused by the high difficulty level of the teaching video. At this time, it is necessary to reduce the difficulty level of each target knowledge point according to the preset step size.
[0048] Among them, if the concentration is greater than or equal to the preset concentration threshold and the cognition is greater than or equal to the preset cognition threshold, it means that the target object is very focused and can fully understand the content of the teaching video. At this time, the initial learning path does not need to be adjusted.
[0049] Among them, if the concentration is less than the preset concentration threshold and the recognition is greater than or equal to the preset recognition threshold, it means that the target object is not paying attention, but cannot fully understand the content of the teaching video. This may be caused by the low difficulty level of the teaching video. At this time, it is necessary to increase the difficulty level of each target knowledge point according to the preset step size.
[0050] If the concentration level is lower than the preset concentration threshold and the awareness level is lower than the preset awareness threshold, it means that the target student is not paying attention and does not understand the content of the teaching video. This may be because the target student is not listening to the class. In this case, the initial learning path will not be adjusted. An alarm message can be generated to remind students to concentrate on listening and learning.
[0051] For example, the concentration threshold is 2.0. The awareness threshold is 60%. The difficulty adjustment step size is ±1 level per knowledge point. Suppose that within a certain time period, the system analyzes the following data: concentration: 2.5, awareness: 55%. This indicates that the student's concentration is ≥ 2.0 and their awareness is < 60%. This indicates that the student is focused but has difficulty understanding the current knowledge point, possibly because the knowledge point is too difficult. Lower the difficulty level of the current knowledge point. Lower the difficulty level of all knowledge points from level 3 to level 2. Regenerate the target learning path.
[0052] In the embodiment of the present application, the above strategy can accurately judge the degree of students' understanding of the teaching content, and timely adjust the initial learning path.
[0053] In an embodiment of the present application, after adjusting the difficulty level of each target knowledge point and obtaining the target learning path of the target object, it is necessary to regenerate the teaching video. The specific process includes: obtaining the third difficulty level carried by each target knowledge point in the target learning path; matching the third difficulty level carried by each target knowledge point with the second difficulty level carried by each target teaching video in the target teaching video set corresponding to each target knowledge point, so as to identify the target teaching video corresponding to the second difficulty level that is the same as the third difficulty level carried by each target knowledge point, as the teaching video corresponding to the target learning path; playing the teaching video corresponding to the target learning path, and continuing to execute the step of receiving the original EEG data sent by the smart wearable device until the concentration is greater than or equal to the preset concentration threshold and the cognition is greater than or equal to the preset cognition threshold.
[0054] For example, the data structure of the teaching video library corresponding to the course identifier is as follows: Figure 3 As shown, a mapping relationship exists between each teaching knowledge point (Knowledge Point A, Knowledge Point B, Knowledge Point C) and its associated teaching video set (for example, the teaching video set for Knowledge Point A is Teaching Videos 1, 2, and 3). If the target knowledge points in the target learning path are Knowledge Point B and Knowledge Point C, then this mapping relationship can be used to obtain Teaching Videos 4, 5, and 6, as well as Teaching Videos 7, 8, and 9, corresponding to Knowledge Points B and C. In the adjusted target learning path, Knowledge Points B and C have difficulty levels. If the difficulty levels marked for Knowledge Points B and C in the target learning path are adjusted to Level 1 and Level 2, then the video corresponding to Knowledge Point B in the initial learning path is Teaching Video 4, and the video corresponding to Knowledge Point C in the initial learning path is Teaching Video 8.
[0055] In this embodiment of the application, the system can dynamically adjust the learning path based on the target subject's performance while watching the instructional video, ensuring that the target subject always learns at a difficulty level that suits them. At the same time, by providing each learner with a personalized learning path and instructional video, it helps narrow the learning gap between different learners and promote educational equity.
[0056] In the embodiments of the present application, on the one hand, the EEG data of the target object is collected in real time by integrating a smart wearable device with a non-invasive brain-computer interface. The EEG data can reflect the electrical activity of the central nervous system and has the characteristics of high temporal resolution. It can represent the cognitive state of the learner in real time. By combining the EEG data with facial images that can reflect changes in eye contact, it can timely and accurately judge the student's understanding of the teaching content. Without the need for students to actively participate in the survey, the difficulty of the teaching content can be dynamically adjusted, and customized teaching content can be provided to students in a timely manner to achieve personalized teaching. On the other hand, by dynamically adjusting the teaching content, learners can avoid getting into trouble in learning, reduce frustration in the learning process, and thus improve learning efficiency and learning outcomes.
[0057] See Figure 4 , provides a flowchart of a method for generating a personalized learning path based on a brain-computer interface for an embodiment of the present application. Figure 4 As shown, the detection method of the embodiment of the present application may include the following steps: S101, in response to a teaching instruction, obtaining a course identifier indicated by the teaching instruction; S102, obtaining an initial learning path pre-set for the course identifier, where the initial learning path is used to represent a system default sequence of target knowledge points for the course identifier and a difficulty level of each target knowledge point in the sequence; S103, determining a teaching video corresponding to the initial learning path from a teaching video library corresponding to the course identifier and playing the teaching video; S104, during the playback process, capturing a face image of the target object in real time through a camera; S105, receiving the original EEG data sent by the smart wearable device, and storing the facial image and the original EEG data at each moment; S106, when the playing time of the teaching video is equal to the preset time, the difficulty level of each target knowledge point is adjusted based on the stored facial image and original EEG data to obtain the target learning path of the target object.
[0058] It should be noted that the specific embodiments of steps S101 to S106 can be found in the detailed description in the system embodiment and will not be repeated here.
[0059] In the embodiments of the present application, on the one hand, the EEG data of the target object is collected in real time by integrating a smart wearable device with a non-invasive brain-computer interface. The EEG data can reflect the electrical activity of the central nervous system and has the characteristics of high temporal resolution. It can represent the cognitive state of the learner in real time. By combining the EEG data with facial images that can reflect changes in eye contact, it can timely and accurately judge the student's understanding of the teaching content. Without the need for students to actively participate in the survey, the difficulty of the teaching content can be dynamically adjusted, and customized teaching content can be provided to students in a timely manner to achieve personalized teaching. On the other hand, by dynamically adjusting the teaching content, learners can avoid getting into trouble in learning, reduce frustration in the learning process, and thus improve learning efficiency and learning outcomes.
[0060] See Figure 5 , is a flowchart of a method for training an eye feature point analysis model according to an embodiment of the present application. Figure 5 As shown, the process includes the following steps: S201, collecting multiple historical facial images of the same user when browsing client content in a natural light environment; S202, segmenting the historical eye region image of each historical face image; S203, using a labeling tool to label feature points of the historical eye region image of each historical face image to mark the pupil position and eyelid opening and closing degree, thereby obtaining a model training sample; S204, creating an eye feature point analysis model using a neural network; S205, inputting the model training sample into the eye feature point analysis model and outputting the model loss value; S206: When the model loss value reaches the minimum, a pre-trained eye feature point analysis model is generated.
[0061] In an embodiment of the present application, when the model loss value reaches a minimum, the step of inputting the model training sample into the eye feature point analysis model can be continued until the model loss value reaches a minimum.
[0062] In an embodiment of the present application, by collecting historical facial images of users browsing client content in a natural light environment, segmenting the eye area and marking feature points, and training to generate a high-precision eye feature point analysis model, it is possible to accurately extract information such as pupil position and eyelid opening and closing degree, provide technical support for eye movement analysis, and thus realize personalized services and intelligent applications.
[0063] The following are device embodiments of the present application, which can be used to implement the method embodiments of the present application. For details not disclosed in the device embodiments of the present application, please refer to the method embodiments of the present application.
[0064] See Figure 6, which shows a schematic diagram of the structure of a personalized learning path generation device based on a brain-computer interface, provided by an exemplary embodiment of the present application. The device can be implemented as all or part of an electronic device through software, hardware, or a combination of both. The device 1 includes a first acquisition module 10, a second acquisition module 20, a determination module 30, a capture module 40, a storage module 50, and an adjustment module 60.
[0065] A first acquisition module 10 is configured to acquire a course identifier indicated by the teaching instruction in response to the teaching instruction; A second acquisition module 20 is configured to pre-set an initial learning path for the course identifier, wherein the initial learning path is used to represent a system default sequence of target knowledge points for the course identifier and a difficulty level of each target knowledge point in the sequence; A determination module 30 is configured to determine and play a teaching video corresponding to the initial learning path from a teaching video library corresponding to the course identifier; The capture module 40 is used to capture the face image of the target object in real time through the camera during the playback process; The storage module 50 is used to receive the original EEG data sent by the smart wearable device and store the facial image and the original EEG data at each moment; The adjustment module 60 is used to adjust the difficulty level of each target knowledge point based on the stored facial image and original EEG data when the playback time of the teaching video is equal to the preset time, so as to obtain the target learning path of the target object.
[0066] It should be noted that the above embodiment of the personalized learning path generation device based on the brain-computer interface provides, when executing the personalized learning path generation method based on the brain-computer interface, only uses the division of the above-mentioned functional modules as an example. In actual application, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the personalized learning path generation device based on the brain-computer interface provided in the above embodiment and the personalized learning path generation method based on the brain-computer interface belong to the same concept. The implementation process is detailed in the method embodiment and will not be repeated here.
[0067] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.
[0068] In the embodiments of the present application, on the one hand, the EEG data of the target object is collected in real time by integrating a smart wearable device with a non-invasive brain-computer interface. The EEG data can reflect the electrical activity of the central nervous system and has the characteristics of high temporal resolution. It can represent the cognitive state of the learner in real time. By combining the EEG data with facial images that can reflect changes in eye contact, it can timely and accurately judge the student's understanding of the teaching content. Without the need for students to actively participate in the survey, the difficulty of the teaching content can be dynamically adjusted, and customized teaching content can be provided to students in a timely manner to achieve personalized teaching. On the other hand, by dynamically adjusting the teaching content, learners can avoid getting into trouble in learning, reduce frustration in the learning process, and thus improve learning efficiency and learning outcomes.
[0069] The present application also provides a computer-readable medium having program instructions stored thereon, which, when executed by a processor, implement the personalized learning path generation method based on brain-computer interface provided by the above-mentioned various method embodiments.
[0070] The present application also provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute the method for generating a personalized learning path based on a brain-computer interface according to each of the above-mentioned method embodiments.
[0071] See Figure 7 , is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 7 As shown, the electronic device 1000 may include: at least one processor 1001 , at least one network interface 1004 , a user interface 1003 , a memory 1005 , and at least one communication bus 1002 .
[0072] The communication bus 1002 is used to implement the connection and communication between these components.
[0073] The user interface 1003 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 1003 may also include a standard wired interface and a wireless interface.
[0074] The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).
[0075] The processor 1001 may include one or more processing cores. The processor 1001 utilizes various interfaces and circuits to connect various components within the electronic device 1000. It executes instructions, programs, code sets, or instruction sets stored in the memory 1005, and accesses data stored in the memory 1005 to perform various functions and process data within the electronic device 1000. Optionally, the processor 1001 may be implemented in hardware using at least one of a digital signal processing (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 1001 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing content displayed on the display; and the modem handles wireless communications. It is understood that the modem may also be implemented independently of the processor 1001 and implemented on a separate chip.
[0076] Among them, the memory 1005 may include a random access memory (RAM) or a read-only memory (Read-Only Memory). Optionally, the memory 1005 includes a non-transitory computer-readable storage medium. The memory 1005 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 1005 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 1005 may also be optionally at least one storage system located away from the aforementioned processor 1001. As Figure 7 As shown, the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module, and a personalized learning path generation application based on a brain-computer interface.
[0077] exist Figure 7In the electronic device 1000 shown, the user interface 1003 is mainly used to provide an input interface for the user and obtain user input data; and the processor 1001 can be used to call the brain-computer interface-based personalized learning path generation application stored in the memory 1005 and specifically perform the following operations: In response to the teaching instruction, obtaining a course identifier indicated by the teaching instruction; Obtaining an initial learning path pre-set for the course identifier, where the initial learning path is used to represent a system default sequence of target knowledge points for the course identifier and a difficulty level of each target knowledge point in the sequence; From the teaching video library corresponding to the course identifier, determine the teaching video corresponding to the initial learning path and play it; During the playback process, the face image of the target object is captured in real time through the camera; Receive the original EEG data sent by the smart wearable device and store the facial image and original EEG data at each moment; When the playback time of the teaching video is equal to the preset time, the difficulty level of each target knowledge point is adjusted based on the stored facial image and original EEG data to obtain the target learning path of the target object.
[0078] In the embodiments of the present application, on the one hand, the EEG data of the target object is collected in real time by integrating a smart wearable device with a non-invasive brain-computer interface. The EEG data can reflect the electrical activity of the central nervous system and has the characteristics of high temporal resolution. It can represent the cognitive state of the learner in real time. By combining the EEG data with facial images that can reflect changes in eye contact, it can timely and accurately judge the student's understanding of the teaching content. Without the need for students to actively participate in the survey, the difficulty of the teaching content can be dynamically adjusted, and customized teaching content can be provided to students in a timely manner to achieve personalized teaching. On the other hand, by dynamically adjusting the teaching content, learners can avoid getting into trouble in learning, reduce frustration in the learning process, and thus improve learning efficiency and learning outcomes.
[0079] Those skilled in the art will appreciate that all or part of the processes in the above-described embodiments can be implemented by instructing the relevant hardware through a computer program. The program for generating a personalized learning path based on a brain-computer interface can be stored in a computer-readable storage medium. When executed, the program can include the processes of the embodiments of the above-described methods. The storage medium for the program for generating a personalized learning path based on a brain-computer interface can be a magnetic disk, an optical disk, a read-only memory, or a random access memory.
[0080] The above disclosure is only a preferred embodiment of the present application, and certainly cannot be used to limit the scope of rights of the present application. Therefore, equivalent changes made according to the claims of the present application are still within the scope covered by the present application.
Claims
1. A personalized learning path generation system based on brain-computer interface, characterized in that: The system comprises: An intelligent wearable device and a client integrated with a non-invasive brain-computer interface, wherein the intelligent wearable device and the client are in communication connection; wherein, The client is configured to, in response to a teaching instruction, obtain a course identifier indicated by the teaching instruction; obtain an initial learning path pre-set for the course identifier, the initial learning path being used to represent a system-default sequence of target knowledge points for the course identifier and a difficulty level of each target knowledge point in the sequence; determine a teaching video corresponding to the initial learning path from a teaching video library corresponding to the course identifier and play the teaching video; and during the playback process, capture a facial image of a target subject in real time via a camera; The smart wearable device is used to collect the original EEG data of the target object in real time and send it to the client; The client is also used to receive the original EEG data sent by the smart wearable device, and store the facial image and the original EEG data at each moment; when the playback time of the teaching video is equal to the preset time, based on the stored facial image and the original EEG data, the difficulty level of each target knowledge point is adjusted to obtain the target learning path of the target object.
2. The system according to claim 1, wherein: The teaching video library corresponding to the course identifier includes a mapping relationship between each teaching knowledge point and a teaching video set of each teaching knowledge point; each target knowledge point is marked with a first difficulty level; The determining, from the teaching video library corresponding to the course identifier, the teaching video corresponding to the initial learning path for playback includes: Obtaining, from the mapping relationship, a target teaching video set corresponding to each target knowledge point, wherein each target teaching video in the target teaching video set is marked with a second difficulty level; Matching the first difficulty level marked on each target knowledge point with the second difficulty level marked on each target teaching video in the target teaching video set corresponding to each target knowledge point, so as to identify a target teaching video corresponding to the second difficulty level that is the same as the first difficulty level carried by each target knowledge point as the teaching video corresponding to the initial learning path; Play the instructional video corresponding to the initial learning path.
3. The system according to claim 1, wherein: The adjusting the difficulty level of each target knowledge point based on the stored facial image and the original EEG data includes: Analyzing the eye movement data of the target object using the stored facial image, where the eye movement data of the target object includes the number of fixations, fixation duration, and blink frequency of the target object; Calculating the concentration of the target object according to the number of fixations, fixation duration, and blink frequency; Performing spectrum analysis on the stored raw EEG data and calculating the power spectrum density of different frequency bands as spectrum analysis results; quantifying the target subject's recognition of the teaching video according to the spectrum analysis result; The difficulty level of each target knowledge point is adjusted according to the concentration and the cognition; wherein the concentration calculation formula is: Concentration = α × number of fixations + β × average fixation duration - γ × blink frequency; α, β, and γ are weight coefficients.
4. The system according to claim 3, characterized in that Analyzing the eye movement data of the target object using the stored facial image includes: Segmenting the eye region images of the target object at various moments within a preset time period from the stored facial image; Inputting the eye region images at each moment into a pre-trained eye feature point analysis model, and outputting the eye feature points of the target object at each moment, wherein the eye feature points include pupil position and eyelid opening and closing degree; Analyzing the movement trajectory of the eye feature points at each moment based on the pupil position and eyelid opening and closing degree to determine the gaze point sequence of the target object; Counting the fixation duration of each fixation point in the fixation point sequence; Counting the blinking frequency of the target object within the preset time period; The number of fixation points in the fixation point sequence, the fixation duration of each fixation point, and the blinking frequency are used as the eye movement data of the target object.
5. The system according to claim 4, characterized in that The following steps are used to generate a pre-trained eye landmark analysis model, including: Collecting multiple historical facial images of the same user when browsing the client content in a natural light environment; Segmenting the historical eye region image of each historical face image; Using a labeling tool to label the historical eye region images of each historical facial image with feature points to mark the pupil position and eyelid opening and closing degree, thereby obtaining a model training sample; Use neural network to create eye feature point analysis model; Inputting the model training sample into the eye feature point analysis model, and outputting a model loss value; When the model loss value reaches the minimum, a pre-trained eye feature point analysis model is generated.
6. The system according to claim 3, wherein: The performing of spectrum analysis on the stored raw EEG data and calculating the power spectrum density of different frequency bands as spectrum analysis results includes: Preprocessing the stored raw EEG data to obtain EEG data to be analyzed in each time window, wherein the preprocessing includes removing DC bias, removing artifacts, and window segmentation; Performing a fast Fourier transform on the EEG data to be analyzed in each time window to decompose the EEG data to be analyzed into sine waves of different frequencies to obtain a complex representation of the frequency domain signal; Determining a frequency axis of the spectrum according to a sampling frequency and a number of data points of the smart wearable device; Determining the frequency value of each frequency component in the complex representation of the frequency domain signal using a frequency axis of the spectrum, wherein the frequency axis provides a frequency reference for the complex representation of the frequency domain signal and can determine the frequency position of each frequency component; Calculating the square of the modulus of the frequency value of each frequency component to obtain the power spectrum density of different frequency bands; The power spectrum density of the different frequency bands is used as the spectrum analysis result.
7. The system according to claim 6, characterized in that The step of quantifying the target subject's recognition of the teaching video according to the spectrum analysis result includes: Identify beta waves, which are associated with alertness and cognitive activity; Identify gamma waves that are related to higher-order cognitive activities; Obtaining, from the spectrum analysis result, the beta wave power corresponding to the beta wave and the gamma wave power corresponding to the gamma wave; performing a weighted summation of the beta wave power and the gamma wave power to obtain a cognitive load score, wherein the cognitive load score is used to represent the target subject's degree of comprehension of the content when watching the teaching video; The cognitive load is taken as the cognitive degree of the target object when watching the teaching video.
8. The system according to claim 3, wherein: The adjusting the difficulty level of each target knowledge point according to the concentration and the cognition includes: When the concentration is greater than or equal to a preset concentration threshold and the recognition is less than a preset recognition threshold, the difficulty level of each target knowledge point is reduced according to a preset step size; or In a case where the concentration is greater than or equal to a preset concentration threshold and the awareness is greater than or equal to a preset awareness threshold, determining not to adjust the initial learning path; or When the concentration is less than a preset concentration threshold and the recognition is greater than or equal to a preset recognition threshold, the difficulty level of each target knowledge point is increased according to a preset step size; or When the concentration is less than a preset concentration threshold and the awareness is less than a preset awareness threshold, it is determined not to adjust the initial learning path.
9. The system according to claim 2, wherein: After adjusting the difficulty level of each target knowledge point to obtain the target learning path of the target object, the method further includes: Obtaining a third difficulty level for each target knowledge point in the target learning path; matching the third difficulty level carried by each target knowledge point with the second difficulty level carried by each target teaching video in the target teaching video set corresponding to each target knowledge point, so as to identify a target teaching video corresponding to the second difficulty level that is the same as the third difficulty level carried by each target knowledge point as the teaching video corresponding to the target learning path; Play the teaching video corresponding to the target learning path, and continue to perform the step of receiving the original EEG data sent by the smart wearable device until the concentration is greater than or equal to the preset concentration threshold and the cognition is greater than or equal to the preset cognition threshold.
10. A method for generating a personalized learning path based on a brain-computer interface, characterized in that: Applied to a client, the method includes: In response to the teaching instruction, obtaining a course identifier indicated by the teaching instruction; Obtaining an initial learning path pre-set for the course identifier, the initial learning path being used to represent a system default sequence of target knowledge points for the course identifier and a difficulty level of each target knowledge point in the sequence; Determine the teaching video corresponding to the initial learning path from the teaching video library corresponding to the course identifier and play it; During the playback process, the face image of the target object is captured in real time through the camera; Receiving raw EEG data sent by the smart wearable device, and storing the facial image and the raw EEG data at each moment; When the playing time of the teaching video is equal to the preset time, the difficulty level of each target knowledge point is adjusted based on the stored facial image and the original EEG data to obtain the target learning path of the target object.