Interactive learning system for preschool education

Abstract

The invention provides an interactive learning system for preschool education, which relates to the technical field of preschool education and comprises a behavior recognition module, an emotion recognition module, an analysis operation module and a self-adaptive adjustment module. The behavior recognition module captures operation behavior data and outputs behavior characteristic parameters; the emotion recognition module obtains action, eye movement and voice emotion data; the analysis operation module performs feature extraction and classification on the behavior data and the emotion data, and outputs a learning state feature tag; the adaptive adjustment module adjusts the learning content according to the state feature tag; the interactive learning system can comprehensively monitor the learning concentration degree and the emotional state of the child in real time, and by dynamically adjusting the learning content and difficulty, the positive emotion and the concentration state of the child are effectively promoted, the learning effect is improved, learning time management and content recommendation are optimized, and the learning efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of preschool education technology, and in particular to an interactive learning system for preschool education. Background Technology

[0002] In the preschool education stage, traditional teaching methods mainly rely on teacher lectures, with children passively receiving knowledge. This model struggles to meet children's individualized learning needs. With the continuous development of information technology and artificial intelligence, digital and intelligent educational methods are gradually being introduced into the preschool education field, but many problems still exist.

[0003] On the one hand, existing educational technologies rely heavily on teachers' subjective observation and assessment when monitoring children's learning status, lacking comprehensive and objective data support, making it difficult to accurately grasp children's learning focus and emotional state. On the other hand, although some systems have interactive learning functions, they are insufficient in adjusting learning content and difficulty in real time according to children's learning status, failing to effectively improve children's learning outcomes and enthusiasm. Therefore, it is necessary to provide an interactive learning system for preschool education to solve the above-mentioned technical problems. Summary of the Invention

[0004] This invention provides an interactive learning system for preschool education, aiming to address the shortcomings of existing preschool education technologies and achieve comprehensive monitoring of children's learning process, personalized teaching, and data security protection.

[0005] To solve the above-mentioned technical problems, the present invention provides an interactive learning system for preschool education, comprising: The behavior recognition module is used to capture children's operational behavior data during the learning process, including reading speed, click frequency, number of times answers are modified, number of times the learning app is exited / paused, cumulative inactivity time, and cumulative continuous learning time; it analyzes the operational behavior data of children during the learning process to obtain attention-related parameters; and outputs behavioral feature parameters based on the attention-related parameters. The emotion recognition module is used to capture action and behavior data and eye-tracking data through the camera, and to analyze emotional responses in speech through natural language processing. The analysis and computation module is used to extract and classify features from behavioral and emotional data using convolutional neural networks and support vector machines, and output learning state feature labels, which include focus, anxiety, and confusion. An adaptive adjustment module is used to dynamically adjust the type and difficulty of learning content based on the feature labels, or to recommend personalized learning content. The security and privacy protection module is used to encrypt children's data and detect illegal information in learning content through natural language processing.

[0006] Preferably, the CNN network of the analysis and operation module includes an input layer, a convolutional layer, a pooling layer, and a fully connected layer, which is used to extract features from behavioral and emotional data and output learning state feature labels. The SVM is based on the radial basis kernel function. The feature vectors are classified to generate learning state feature labels, where For the input vector, The kernel function parameters are used; feature labels include focus, anxiety, and confusion.

[0007] As a preferred approach, analyzing children's operational behavior data during the learning process yields attention-related parameters, specifically: Obtain a child's reading speed R during the learning process and compare it with the average reading speed for their current age group. ; Compare reading speed R with the average reading speed of the age group The comparison was performed using the following formula: 00, to obtain a reading activity score. ;like A reading speed greater than the set reading speed threshold indicates that the child's reading activity level is above average, while a reading speed less than the set reading speed threshold indicates that it is below average. The system obtains the child's click frequency (C) and the number of times the answer is modified (M) during the learning process. Based on these metrics, it calculates the child's activity level during the operation and answering phases using a pre-defined calculation formula. The operational activity score was calculated. ,in , These are the average click frequency and average number of answer modifications for the same age group, respectively. The system retrieves the number of times a child exits / pauses their learning app (E) and the cumulative time of inactivity (U) during the learning process, using a preset formula. 00 The learning stability score S is calculated; where , These represent the average number of times an action is canceled / paused and the average cumulative time without action, respectively, for the same age group. Obtain the child's cumulative continuous learning time L during the learning process using a preset formula. 00 calculates the sustained focus score P, where This indicates the average cumulative continuous learning time for the same age group; Taking into account the child's reading activity score, performance activity score, learning stability score, and sustained focus score, the child's learning focus score (FS) for this learning process is calculated using the formula: ;in, , , Represents the weighting coefficient, and + ; Reading activity score, operational activity score, learning stability score, sustained focus score, and learning focus score are recorded as attention-related parameters.

[0008] Preferably, a visualization module is also included for generating: A pie chart showing the time spent on each subject category displays the percentage of time children dedicate to each category. The calculation is based on the time (Ti) children spend on each subject. generate; The time-feature label map shows the peak time of daily focus. By plotting time on the horizontal axis and learning state feature labels on the vertical axis, it records the generation of learning states corresponding to different time periods. The focus rate-study day variation chart shows the long-term focus rate trend. It quantifies the proportion of focus characteristic tags in the total study time on each study day, i.e., the focus rate. XF1 and XF2 represent the duration of the focus tag and the total learning time, respectively. The graph is generated with the learning days on the horizontal axis and the focus rate on the vertical axis.

[0009] Preferably, the adaptive adjustment module includes: The difficulty adjustment unit lowers the learning content difficulty level by a preset rule when an anxiety tag is detected; and increases the learning content difficulty when a focus tag is detected and the current difficulty level is below the highest level. The recommendation unit calculates interest preference scores based on the frequency of children's attention tags (Fi) across different learning categories. The system prioritizes categories with the highest interest and preference scores; these categories include fables, science, pinyin (phonetic transcription of Chinese characters), sports, handicrafts, and music.

[0010] Preferably, the security and privacy protection module includes: The data encryption unit uses the AES encryption algorithm to encrypt and store children's operational data, emotional data, and learning records. The illegal content detection unit identifies and filters inappropriate information in the learning content based on a natural language processing model.

[0011] As a preferred approach, the method involves identifying and filtering inappropriate information in the learning content based on a natural language processing model, specifically as follows: Acquire children's learning content; preprocess the learning content; preprocessing includes text cleaning and word segmentation; Key text features are extracted from the preprocessed text information, and the bag-of-words model and Naive Bayes model are used for classification. If the information is determined to contain inappropriate information, the corresponding words are replaced to achieve filtering.

[0012] Preferably, the difficulty adjustment unit of the adaptive adjustment module dynamically adjusts the difficulty of subsequent learning content based on the child's accuracy and completion time on the current difficulty level of the learning content, specifically as follows: After a child completes the learning content at the current difficulty level, the number of questions the child answered correctly (nc) and the total number of questions (nt) are collected. The accuracy rate (Acc) is then calculated based on this data. 00%; Next, obtain the time it takes for the child to complete the learning content, i.e., the completion time t; Set accuracy threshold and completion time threshold And the corresponding acceptable threshold range, which is compared with the accuracy and completion time: When a child's accuracy rate is higher than the set accuracy rate threshold and the completion time is less than the set completion time threshold, it means that the child is learning relatively easily at the current level. At this time, the difficulty of the learning content should be increased quickly. If a child's accuracy rate is lower than the set accuracy rate threshold and the completion time is greater than the set completion time threshold, it indicates that the child is having difficulty learning at the current level. In this case, it is necessary to appropriately reduce the difficulty of the learning content and provide additional pre-set tutoring resources. If a child's accuracy rate is within the acceptable threshold range and their completion time is within the acceptable threshold range, then the current difficulty level will be maintained.

[0013] As a preferred option, a historical attention management module and an attention storage module are also included; The attention storage module is used to record and store the attention-related parameters corresponding to the time range from the child's first learning moment to the current moment as historical records; The historical attention management module is used to analyze and process attention-related parameters in the historical records. Specifically, it obtains the time of each learning session of the child, calculates the difference between adjacent learning times to obtain the learning interval, and numbers the learning intervals in the historical records. Construct a learning attention line graph by inputting the historical numbers and learning attention scores into the line graph, and marking the position of the learning attention score in the line graph as the construction point; connect adjacent construction points to obtain construction lines; and form the learning attention line graph by combining all the construction lines. Identify any construction line, take the midpoint of the construction line as the starting point, draw a perpendicular line to the starting point with the learning interval as the height, and connect the vertex of the perpendicular line to the two construction points of the corresponding construction line to obtain the learning attention change pattern. The inter-learning influence surface value is obtained by calculating the change graph of learning attention; the inter-learning influence surface value is used to determine the influence between the learning interval of adjacent learning and the learning attention score of children. Then, the inter-learning influence value is obtained by using the standard deviation formula to calculate the inter-learning influence surface value in the learning attention line graph.

[0014] Compared with related technologies, the interactive learning system for preschool education provided by this invention has the following beneficial effects: This invention's interactive learning system can monitor children's learning focus and emotional state in real time and comprehensively, providing precise data support for personalized education. By dynamically adjusting learning content and difficulty, it effectively promotes children's positive emotions and focus, improving learning outcomes. It also ensures secure storage and privacy protection of children's learning data, giving parents and teachers peace of mind. The visualization function helps parents and teachers understand children's learning patterns, facilitating scientific educational guidance. The historical attention management module analyzes historical data such as learning intervals, helping to optimize learning time management and content recommendations, further improving learning efficiency. Attached Figure Description

[0015] Figure 1 A schematic diagram of an interactive learning system for preschool education provided by the present invention; Figure 2 The learning focus line graph provided for this invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] The terminology used in this disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The singular forms “group,” “class,” and “the” as used in this disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0018] It should be understood that although the terms first, second, third, etc., may be used in this disclosure to describe various information, such information should not be limited to these terms. These terms are used only to distinguish information of the same type from one another. For example, without departing from the scope of this disclosure, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0019] Please refer to the following: Figures 1-2 An interactive learning system for preschool education, comprising: The behavior recognition module is used to capture children's operational behavior data during the learning process, including reading speed, click frequency, number of times answers are modified, number of times the learning app is exited / paused, cumulative inactivity time, and cumulative continuous learning time; it analyzes the operational behavior data of children during the learning process to obtain attention-related parameters; and outputs behavioral feature parameters based on the attention-related parameters. The emotion recognition module is used to capture action behavior data and eye tracking data through the camera, and to analyze voice emotional responses through natural language processing. The action behavior data includes the number of times the person leaves their seat, the number of times they look around, and the number of times they ask for help. The eye tracking data includes the duration of the gaze focus and the amplitude / frequency of the gaze movement. The voice emotional responses include engagement, anxiety, and confusion. The analysis and computation module is used to extract and classify features from behavioral and emotional data using convolutional neural networks and support vector machines, and output learning state feature labels, including focus, anxiety, and confusion. The adaptive adjustment module is used to dynamically adjust the type and difficulty of learning content based on feature labels, or recommend personalized learning content. The security and privacy protection module is used to encrypt children's data and detect illegal information in learning content through natural language processing.

[0020] It should be noted that through the collaborative work of the aforementioned modules, the system achieves comprehensive monitoring, precise analysis, intelligent adjustment, and security assurance functions, providing a personalized, efficient, and safe learning environment for preschool children and promoting their learning and development. Through behavior recognition and emotion recognition modules, it achieves comprehensive, real-time monitoring of children's learning status, providing accurate basis for subsequent teaching adjustments. The adaptive adjustment module dynamically adjusts learning content based on the learning status feature tags output by the analysis and calculation module. When anxiety is detected in a child, the difficulty of the learning content is reduced; if the child is focused and the current difficulty level allows, the difficulty is increased. Simultaneously, based on the frequency of the child's focus tags across different learning categories, personalized learning content is recommended to maintain the child's learning interest and enthusiasm, improve learning outcomes, and achieve personalized education.

[0021] In this application, the CNN network of the analysis and operation module includes an input layer, a convolutional layer, a pooling layer, and a fully connected layer, which is used to extract features from behavioral and emotional data and output learning state feature labels. SVM is based on radial basis kernel function. The feature vectors are classified to generate learning state feature labels, where For the input vector, The kernel function parameters are used; feature labels include focus, anxiety, and confusion.

[0022] It should be noted that by combining CNN and SVM in the analysis and computation module, and through complex data processing and classification, accurate analysis of children's learning status is achieved. The multi-layered structure of the CNN network endows it with powerful data feature extraction capabilities. The input layer receives behavioral and emotional data, and the convolutional layer uses convolutional kernels to perform convolution operations on the data, which can capture local features in the data; for example, when analyzing behavioral data, features such as patterns and rhythms of operational behaviors can be extracted; for image information in emotional data, features such as facial expression changes and eye movement trajectories can be identified; the pooling layer further filters and integrates the features output by the convolutional layer, reducing data dimensionality and improving computational efficiency while retaining key features; the fully connected layer integrates all local features to form a vector that comprehensively reflects the data features. SVM classifies feature vectors output by CNN networks based on radial basis function (RBF) kernels. RBF kernels map low-dimensional feature vectors to high-dimensional spaces, making data that might be linearly inseparable in low-dimensional spaces linearly separable in high-dimensional spaces. This characteristic allows SVM to more accurately classify children's learning states, generating accurate learning state feature labels such as focus, anxiety, and confusion. Compared to other simple classification methods, SVM performs better in handling complex nonlinear classification problems, effectively improving the accuracy of judging children's learning states and providing a reliable decision-making basis for the adaptive adjustment module.

[0023] In this application, action behavior data during children's learning process are analyzed to obtain attention-related parameters, specifically: Obtain a child's reading speed R during the learning process and compare it with the average reading speed for their current age group. ; Compare reading speed R with the average reading speed of the age group The comparison was performed using the following formula: 00, to obtain a reading activity score. ;like A reading speed greater than the set reading speed threshold indicates that the child's reading activity level is above average, while a reading speed less than the set reading speed threshold indicates that it is below average. The system obtains the child's click frequency (C) and the number of times the answer is modified (M) during the learning process. Based on these metrics, it calculates the child's activity level during the operation and answering phases using a pre-defined calculation formula. The operational activity score was calculated. ,in , These are the average click frequency and average number of answer modifications for the same age group; reading speed, click frequency, and number of answer modifications can reflect the degree of active participation of children in the learning process; The system retrieves the number of times a child exits / pauses their learning app (E) and the cumulative time of inactivity (U) during the learning process, using a preset formula. 00 The learning stability score S is calculated; where , These represent the average number of exits / pauses and the average cumulative time without operation for the same age group, respectively; the fewer the number of exits / pauses (E) and the shorter the cumulative time without operation (U), the higher the learning stability. Obtain the child's cumulative continuous learning time L during the learning process using a preset formula. 00 calculates the sustained focus score P, where This indicates the average cumulative continuous learning time for the same age group; Taking into account the child's reading activity score, performance activity score, learning stability score, and sustained focus score, the child's learning focus score (FS) for this learning process is calculated using the formula: ;in, , , Represents the weighting coefficient, and + ; Reading activity score, operational activity score, learning stability score, sustained focus score, and learning focus score are recorded as attention-related parameters.

[0024] This application also includes a visualization module for generating: A pie chart showing the time spent on each subject category displays the percentage of time children dedicate to each category. The calculation is based on the time (Ti) children spend on each subject. Generate a learning category time pie chart to intuitively display the percentage of time children spend on different learning categories, allowing parents and teachers to quickly understand children's investment in each category of learning, making it easier to rationally plan learning time and optimize learning arrangements; The Time-Feature Label Chart displays the peak periods of daily focus. By plotting time on the horizontal axis and learning status feature labels on the vertical axis, it records the learning status generated in different time periods. The Time-Feature Label Chart clearly presents the peak periods of daily focus, helping parents and teachers to identify the times when children's learning efficiency is higher, thereby scientifically arranging learning tasks and improving learning outcomes. The focus rate-study day variation chart shows the long-term focus rate trend. It quantifies the proportion of focus characteristic tags in the total study time on each study day, i.e., the focus rate. XF1 and XF2 represent the duration of the focus tag and the total learning time, respectively. The graph is generated with the learning day as the horizontal axis and the focus rate as the vertical axis. The focus rate-learning day change graph shows the long-term focus rate trend, which makes it easy to observe the long-term fluctuations in children's learning focus, identify problems in time and adjust teaching strategies, and help children develop good learning habits.

[0025] In this application, the adaptive adjustment module includes: The difficulty adjustment unit lowers the learning content difficulty level by a preset rule when an anxiety tag is detected; and increases the learning content difficulty when a focus tag is detected and the current difficulty level is below the highest level. The recommendation unit calculates interest preference scores based on the frequency of children's attention tags (Fi) across different learning categories. The system prioritizes categories with the highest interest and preference scores; these categories include fables, science, pinyin (phonetic transcription of Chinese characters), sports, handicrafts, and music.

[0026] It should be noted that the difficulty adjustment unit adjusts the difficulty of the learning content in real time according to the child's learning status. The difficulty is reduced when the child is anxious to alleviate learning pressure; the difficulty is increased when the child is focused and there is room for improvement to stimulate learning potential, ensuring that the learning difficulty always matches the child's learning ability and maintains their learning enthusiasm and focus. The recommendation unit calculates interest preference scores based on the frequency of children's focus tags in different learning categories, accurately recommends learning categories with high interest, meets personalized learning needs, helps children learn in depth in areas of interest, and cultivates learning interest and long-term focused learning habits.

[0027] In this application, the security and privacy protection module includes: The data encryption unit uses the AES encryption algorithm to encrypt and store children's operational data, emotional data, and learning records. It should be noted that the data encryption unit uses the AES encryption algorithm to encrypt and store children's related data to prevent data leakage, protect children's privacy, and allow parents and children to use the system with peace of mind. The content violation detection unit uses a natural language processing model to identify and filter inappropriate information in learning content, creating a healthy and safe learning environment for children and preventing inappropriate content from having a negative impact on them.

[0028] In this application, the identification and filtering of inappropriate information in the learning content is based on a natural language processing model, specifically as follows: Acquire children's learning content; preprocess the learning content; preprocessing includes text cleaning and word segmentation; text cleaning removes HTML tags, special characters (such as garbled text, invisible characters), punctuation marks, etc. from the text while retaining text information; word segmentation breaks down continuous text into individual words or phrases. Key text features are extracted from the preprocessed text information, and the bag-of-words model and Naive Bayes model are used for classification. If the information is determined to contain inappropriate information, the corresponding words are replaced to achieve filtering.

[0029] In this application, the difficulty adjustment unit of the adaptive adjustment module dynamically adjusts the difficulty of subsequent learning content based on the child's accuracy and completion time on the current difficulty level of the learning content. Specifically: After a child completes the learning content at the current difficulty level, the number of questions the child answered correctly (nc) and the total number of questions (nt) are collected. The accuracy rate (Acc) is then calculated based on this data. 00%; Next, obtain the time it takes for the child to complete the learning content, i.e., the completion time t; Set accuracy threshold and completion time threshold And the corresponding acceptable threshold range, which is compared with the accuracy and completion time: When a child's accuracy rate is higher than the set accuracy rate threshold and the completion time is less than the set completion time threshold, it means that the child is learning relatively easily at the current level. At this time, the difficulty of the learning content should be increased quickly. If a child's accuracy rate is lower than the set accuracy rate threshold and the completion time is greater than the set completion time threshold, it indicates that the child is having difficulty learning at the current level. In this case, it is necessary to appropriately reduce the difficulty of the learning content and provide additional pre-set tutoring resources. If a child's accuracy rate is within the acceptable threshold range and their completion time is within the acceptable threshold range, then the current difficulty level will be maintained.

[0030] It should be noted that by dynamically adjusting the difficulty of subsequent learning content based on the child's accuracy and completion time on the current learning content, the learning content can be precisely matched with the child's ability, avoiding being too difficult or too easy, thus improving learning effectiveness. Increasing the difficulty when it is easy and decreasing the difficulty when it is difficult can stimulate the child's learning potential, reduce learning frustration, maintain learning enthusiasm and motivation, and achieve personalized learning.

[0031] In this application, the present invention also includes a historical attention management module and an attention storage module; The attention storage module is used to record and store the attention-related parameters corresponding to the time range from the child's first learning moment to the current moment as historical records; The historical attention management module is used to analyze and process attention-related parameters in the historical records. Specifically, it obtains the time of each child's learning session, calculates the difference between adjacent learning times to obtain the learning interval, and numbers the learning intervals in the historical records. Construct a learning attention line graph by inputting the historical numbers and learning attention scores into the line graph, and marking the position of the learning attention score in the line graph as the construction point; connect adjacent construction points to obtain construction lines; and form the learning attention line graph by combining all the construction lines. Identify any construction line, take the midpoint of the construction line as the starting point, draw a perpendicular line to the starting point with the learning interval as the height, and connect the vertex of the perpendicular line to the two construction points of the corresponding construction line to obtain the learning attention change pattern. The inter-learning influence surface value is obtained by calculating the change graph of learning attention; the inter-learning influence surface value is used to determine the influence between the learning interval of adjacent learning and the learning attention score of children. Then, the inter-learning influence value is obtained by using the standard deviation formula to calculate the inter-learning influence surface value in the learning attention line graph.

[0032] It should be noted that by analyzing the relationship between learning intervals and learning attention scores through the historical attention management module, we can discover the time patterns of children's learning, such as the appropriate length of learning intervals, and provide a basis for optimizing learning arrangements. The learning interval impact value can measure the degree of dispersion of the relationship between learning intervals and attention scores, assess learning stability, and promptly detect fluctuations in learning status.

[0033] Other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of the invention are indicated by the following claims.

[0034] It should be understood that the present invention is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. An interactive learning system for preschool education, characterized in that, include: The behavior recognition module is used to capture children's operational behavior data during the learning process, including reading speed, click frequency, number of times answers are modified, number of times the learning app is exited / paused, cumulative time of no operation, and cumulative time of continuous learning. Analyze children's operational behavior data during the learning process to obtain attention-related parameters; output behavioral characteristic parameters based on the attention-related parameters. The emotion recognition module is used to capture action and behavior data and eye-tracking data through the camera, and to analyze emotional responses in speech through natural language processing. The analysis and computation module is used to extract and classify features from behavioral and emotional data using convolutional neural networks and support vector machines, and output learning state feature labels, which include focus, anxiety, and confusion. An adaptive adjustment module is used to dynamically adjust the type and difficulty of learning content based on the feature labels, or to recommend personalized learning content. The security and privacy protection module is used to encrypt children's data and detect illegal information in learning content through natural language processing.

2. The interactive learning system for preschool education according to claim 1, characterized in that, The CNN network of the analysis and operation module includes an input layer, a convolutional layer, a pooling layer, and a fully connected layer, which is used to extract features from behavioral and emotional data and output learning state feature labels. The SVM is based on the radial basis kernel function. The feature vectors are classified to generate learning state feature labels, where For the input vector, The kernel function parameters are used; feature labels include focus, anxiety, and confusion.

3. The interactive learning system for preschool education according to claim 1, characterized in that, By analyzing children's operational behavior data during the learning process, attention-related parameters were obtained, specifically: Obtain a child's reading speed R during the learning process and compare it with the average reading speed for their current age group. ; Compare reading speed R with the average reading speed of the age group The comparison was performed using the following formula: 00, to obtain a reading activity score. ;like A reading speed greater than the set reading speed threshold indicates that the child's reading activity level is above average, while a reading speed less than the set reading speed threshold indicates that it is below average. The system obtains the child's click frequency (C) and the number of times the answer is modified (M) during the learning process. Based on these metrics, it calculates the child's activity level during the operation and answering phases using a pre-defined calculation formula. The operational activity score was calculated. ,in , These are the average click frequency and average number of answer modifications for the same age group, respectively. The system retrieves the number of times a child exits / pauses their learning app (E) and the cumulative time of inactivity (U) during the learning process, using a preset formula. 00 The learning stability score S is calculated; where , These represent the average number of times an action is canceled / paused and the average cumulative time without action, respectively, for the same age group. Obtain the child's cumulative continuous learning time L during the learning process using a preset formula. 00 calculates the sustained focus score P, where This indicates the average cumulative continuous learning time for the same age group; Taking into account the child's reading activity score, performance activity score, learning stability score, and sustained focus score, the child's learning focus score (FS) for this learning process is calculated using the formula: ;in, , , Represents the weighting coefficient, and + ; Reading activity score, operational activity score, learning stability score, sustained focus score, and learning focus score are recorded as attention-related parameters.

4. The interactive learning system for preschool education according to claim 1, characterized in that, It also includes a visualization module for generating: A pie chart showing the time spent on each subject category displays the percentage of time children dedicate to each category. The calculation is based on the time (Ti) children spend on each subject. generate; The time-feature label map shows the peak time of daily focus. By plotting time on the horizontal axis and learning state feature labels on the vertical axis, it records the generation of learning states corresponding to different time periods. The focus rate-study day variation chart shows the long-term focus rate trend. It quantifies the proportion of focus characteristic tags in the total study time on each study day, i.e., the focus rate. XF1 and XF2 represent the duration of the focus tag and the total learning time, respectively. The graph is generated with the learning days on the horizontal axis and the focus rate on the vertical axis.

5. The interactive learning system for preschool education according to claim 1, characterized in that, The adaptive adjustment module includes: The difficulty adjustment unit lowers the learning content difficulty level by a preset rule when an anxiety tag is detected; and increases the learning content difficulty when a focus tag is detected and the current difficulty level is below the highest level. The recommendation unit calculates interest preference scores based on the frequency of children's attention tags (Fi) across different learning categories. The system prioritizes categories with the highest interest and preference scores; these categories include fables, science, pinyin (phonetic transcription of Chinese characters), sports, handicrafts, and music.

6. The interactive learning system for preschool education according to claim 1, characterized in that, The security and privacy protection module includes: The data encryption unit uses the AES encryption algorithm to encrypt and store children's operational data, emotional data, and learning records. The illegal content detection unit identifies and filters inappropriate information in the learning content based on a natural language processing model.

7. The interactive learning system for preschool education according to claim 6, characterized in that, The method involves identifying and filtering inappropriate information in the learning content based on a natural language processing model. Acquire children's learning content; preprocess the learning content; preprocessing includes text cleaning and word segmentation; Key text features are extracted from the preprocessed text information, and the bag-of-words model and Naive Bayes model are used for classification. If the information is determined to contain inappropriate information, the corresponding words are replaced to achieve filtering.

8. The interactive learning system for preschool education according to claim 1, characterized in that, The difficulty adjustment unit of the adaptive adjustment module dynamically adjusts the difficulty of subsequent learning content based on the child's accuracy and completion time at the current difficulty level. Specifically: After a child completes the learning content at the current difficulty level, the number of questions the child answered correctly (nc) and the total number of questions (nt) are collected. The accuracy rate (Acc) is then calculated based on this data. 00%; Next, obtain the time it takes for the child to complete the learning content, i.e., the completion time t; Set accuracy threshold and completion time threshold And the corresponding acceptable threshold range, which is compared with the accuracy and completion time: When a child's accuracy rate is higher than the set accuracy rate threshold and the completion time is less than the set completion time threshold, it means that the child is learning relatively easily at the current level. At this time, the difficulty of the learning content should be increased quickly. If a child's accuracy rate is lower than the set accuracy rate threshold and the completion time is greater than the set completion time threshold, it indicates that the child is having difficulty learning at the current level. In this case, it is necessary to appropriately reduce the difficulty of the learning content and provide additional pre-set tutoring resources. If a child's accuracy rate is within the acceptable threshold range and their completion time is within the acceptable threshold range, then the current difficulty level will be maintained.

9. The interactive learning system for preschool education according to claim 1, characterized in that, It also includes a historical attention management module and an attention storage module; The attention storage module is used to record and store the attention-related parameters corresponding to the time range from the child's first learning moment to the current moment as historical records; The historical attention management module is used to analyze and process attention-related parameters in the historical records. Specifically, it obtains the time of each learning session of the child, calculates the difference between adjacent learning times to obtain the learning interval, and numbers the learning intervals in the historical records. Construct a learning attention line graph by inputting the historical numbers and learning attention scores into the line graph, and marking the position of the learning attention score in the line graph as the construction point; connect adjacent construction points to obtain construction lines; and form the learning attention line graph by combining all the construction lines. Identify any construction line, take the midpoint of the construction line as the starting point, draw a perpendicular line to the starting point with the learning interval as the height, and connect the vertex of the perpendicular line to the two construction points of the corresponding construction line to obtain the learning attention change pattern. Computational learning focuses on changing graphics to obtain the surface value of the inter-learning influence; Inter-learning impact facets are used to determine the influence between children's adjacent learning intervals and their learning attention scores; Then, the inter-learning influence value is obtained by using the standard deviation formula to calculate the inter-learning influence surface value in the learning attention line graph.

Images