Child situation learning watch system based on environmental language perception and safety knowledge graph

Abstract

The invention relates to the technical field of smart watches, and discloses a child situation learning watch system based on environmental language perception and a safety knowledge graph, and the system comprises a graph construction module which is configured to construct the safety knowledge graph according to safety knowledge; the sensing module is configured to trigger different modes according to the tone judgment result and the sensing judgment result; the sensing module is further configured to analyze the environment language information to obtain environment language content; the analysis interaction module is configured to generate recommendation content according to the retrieval result and in combination with different modes; the analysis interaction module is further configured to perform child answer scoring according to the child language information and the recommended content; and the feedback module is configured to determine the next learning content and the next learning interval duration according to the child answer score. According to the invention, the safety and interactivity of children learning are improved, and the learning efficiency and effect are greatly improved through a personalized feedback mechanism.

Description

Technical Field

[0001] The present invention relates to the technical field of smart watches, and in particular to a children's contextual learning watch system based on environmental language perception and safety knowledge graph. Background Art

[0002] With the widespread adoption of smart wearable devices, children's smartwatches, as a key component, have gradually become an indispensable tool in children's daily lives. However, existing children's smartwatches have many limitations in terms of functionality and performance, particularly in safety education and personalized learning. Existing children's smartwatches primarily focus on basic communication, positioning, and safety monitoring functions, lacking in-depth environmental awareness capabilities and personalized adaptation of educational content.

[0003] First, existing children's smartwatches typically only provide simple safety tips and pre-set educational content for safety education. This content often lacks specificity and interactivity, making it difficult to maintain children's interest and engagement. Furthermore, these watches typically lack the ability to dynamically adjust educational content based on changes in a child's environment, significantly reducing educational effectiveness. Second, existing children's smartwatches also have shortcomings in environmental perception. Most are limited to simple ambient sound detection and lack the ability to deeply understand and process environmental language information. This limits their application in complex environments and prevents them from providing a richer and more personalized learning experience. Furthermore, existing children's smartwatches also have flaws in their personalized learning feedback mechanisms. They often fail to provide effective feedback and adjust learning plans based on children's learning progress and understanding. This can lead to children experiencing over-learning or under-learning, impacting learning efficiency and effectiveness. Finally, existing children's smartwatches also have shortcomings in constructing safety knowledge graphs. They typically rely on pre-set, static knowledge bases and lack the ability to dynamically update and expand them. This limits the scope and depth of their application in safety education. In summary, children's smart watches in the existing technology have many problems in environmental language perception, safety knowledge education, personalized learning feedback, and safety knowledge graph construction, and cannot meet the needs of modern children in safety education and personalized learning.

[0004] Therefore, it is necessary to provide a children's contextual learning watch system based on environmental language perception and safety knowledge graph to solve the problem that smart watches in the existing technology cannot meet the needs of modern children in safety education and personalized learning. Summary of the Invention

[0005] In view of this, the present invention proposes a children's contextual learning watch system based on environmental language perception and safety knowledge graph, aiming to solve the problem that smart watches in the existing technology cannot meet the needs of modern children in safety education and personalized learning.

[0006] The present invention proposes a children's contextual learning watch system based on environmental language perception and safety knowledge graph, including: A graph construction module is configured to obtain security knowledge from multiple channels and construct a security knowledge graph based on the security knowledge; a perception module configured to obtain environmental language information and somatosensory interaction information, perform timbre judgment on the environmental language information, perform perceptual judgment on the somatosensory interaction information, and trigger different modes based on the timbre judgment results and the perceptual judgment results; wherein the different modes include a learning mode and an alarm mode; The perception module is further configured to parse the environmental language information to obtain environmental language content; an analysis and interaction module configured to search the environmental language content in the safety knowledge graph and generate recommended content based on the search results and in combination with the different modes; the analysis and interaction module is further configured to obtain child language information and score the child's answer based on the child language information and the recommended content; The feedback module is configured to determine the next learning content and the next learning interval length based on the child's answer score.

[0007] Furthermore, the graph construction module is configured to acquire security knowledge from multiple channels and construct a security knowledge graph based on the security knowledge, including: A security knowledge collection unit is configured to collect multi-source heterogeneous security knowledge from multiple channels; a knowledge fusion unit configured to clean, deduplicate, and standardize the multi-source heterogeneous safety knowledge to extract key safety factors; wherein the key safety factors include dangerous scene features, dangerous speech features, dangerous behavior features, and geographic location risk tags; The graph generation unit is configured to construct a security knowledge graph based on the key security elements.

[0008] Furthermore, the perception module is configured to obtain environmental language information and somatosensory interaction information, perform timbre judgment on the environmental language information, and perform perception judgment on the somatosensory interaction information, including: Acquire the timbre information pre-recorded in the watch system, compare the ambient timbre of the ambient language information with the timbre information, and determine the timbre consistency between the ambient timbre and the timbre information; if the ambient timbre belongs to one of the timbre information, the timbre determination result is timbre consistency; if the ambient timbre does not belong to any of the timbre information, the timbre determination result is timbre inconsistency; Obtain the distance information of the environmental language information from the watch, compare the distance information with the preset distance, and perform perception judgment. If the distance information is less than the preset distance, the perception judgment result is a dangerous distance; if the distance information is greater than or equal to the preset distance, the perception judgment result is a safe distance.

[0009] Furthermore, when the perception module is configured to trigger different modes according to the timbre judgment result and the perception judgment result, it includes: If the timbre judgment result is consistent, and the perception judgment result is dangerous distance, the learning mode is triggered; If the timbre judgment result is consistent, and the perception judgment result is a safe distance, the learning mode is triggered; If the sound color judgment result is inconsistent, and the perception judgment result is dangerous distance, the alarm mode will be triggered, and a level 1 warning will be issued. The alarm number will be automatically dialed and the positioning information will be sent; If the timbre judgment result is inconsistent, and the perception judgment result is a safe distance, the alarm mode will be triggered and a second-level warning will be issued.

[0010] Furthermore, the perception module is further configured to parse the environmental language information to obtain the environmental language content, further comprising: converting the environmental language information into environmental language content in text form, and dividing the environmental language content into different words; Stop words and redundant modifiers in the vocabulary are removed, and the remaining vocabulary is formed into a vocabulary set.

[0011] Furthermore, the analysis and interaction module is configured to search for the environmental language content in the security knowledge graph, and generate recommended content based on the search results and in combination with the different modes, including: Searching the vocabulary set in the security knowledge graph to obtain security knowledge nodes associated with the vocabulary set and the association relationships between the nodes; The security knowledge nodes are sorted according to the closeness of the association relationship, and the security knowledge corresponding to a preset number of nodes that are ranked first are extracted.

[0012] Furthermore, the analysis and interaction module is configured to search for the environmental language content in the security knowledge graph, and generate recommended content based on the search results and in combination with the different modes, further comprising: If in learning mode, the extracted safety knowledge will be converted into learning content in the form of pictures, texts, animations or voice as recommended content; If in alarm mode, during the first-level warning, emergency response methods will be screened out from the extracted safety knowledge and reported quickly as recommended content; during the second-level warning, the extracted safety knowledge will be integrated into warning information, displayed synchronously on the watch screen and pushed to the guardian's terminal as recommended content.

[0013] Furthermore, the analysis and interaction module is further configured to obtain child language information, and to score the child's answer based on the child language information and the recommended content, including: Obtaining child language information generated by the child in response to the environmental language information, and comparing the child language information with the recommended content to calculate answer similarity; A similarity interval is set, and if the answer similarity is less than a minimum value of the similarity interval, the child's answer score is the first score; If the answer similarity is within the similarity range, the child's answer score is the second score; If the answer similarity is greater than the maximum value of the similarity interval, the child's answer score is the third score; The first score is smaller than the second score, and the second score is smaller than the third score.

[0014] Furthermore, the feedback module is configured to determine the next learning content and the next learning interval duration based on the child's answer score, including: If the score is the first, then obtaining the same type of questions with a similarity greater than a similarity threshold with the environment language information from the security knowledge graph, and determining the next learning content to be the same type of questions; If the score is the second score, similar questions whose similarity to the environment language information is less than the similarity threshold and greater than zero are obtained from the security knowledge graph, and the next learning content is determined to be the similar questions; If the score is the third score, irrelevant questions with zero similarity to the environmental language information are obtained from the security knowledge graph, and the next learning content is determined to be the irrelevant questions.

[0015] Furthermore, when the feedback module is configured to determine the next learning content and the next learning interval duration based on the child's answer score, it also includes: Setting a standard interval length, and adjusting the standard interval length using an adjustment coefficient based on the child's answer score to obtain the next learning interval length; The child's answer score is proportional to the adjustment coefficient, and the adjustment coefficient has a value range of (0, 1], and the next learning interval duration is the product of the standard interval duration and the adjustment coefficient.

[0016] Compared to existing technologies, the present invention offers the following advantages: By integrating environmental language perception with a safety knowledge graph, it provides a highly interactive and safe learning environment for children. First, the graph construction module integrates safety knowledge from multiple sources to construct a comprehensive safety knowledge graph. This not only ensures the richness and accuracy of the information, but also provides strong knowledge support for the system, enabling it to better understand and respond to the child's environment. The perception module allows the watch system to acquire environmental language information and somatosensory interaction information in real time. Based on timbre and perceptual judgment, it triggers different modes, such as learning mode and alarm mode, allowing the watch to flexibly adjust its response based on environmental changes and child interaction, thereby providing a more personalized learning experience. Furthermore, the analysis of environmental language information enables the watch to understand and respond to the child's surrounding language environment, further enhancing its interactivity. The analysis and interaction module retrieves environmental language content from the safety knowledge graph and generates recommendations based on different modes. This not only improves the relevance of learning content, but also assesses the child's learning progress and understanding by acquiring language information and scoring responses, thereby providing more precise and personalized learning recommendations. Finally, the feedback module determines the next learning content and the duration of the learning interval based on the child's response score. This adaptive learning mechanism ensures that children learn in an optimal state, avoiding over-learning or under-learning. Overall, the system's highly integrated modular design not only improves the safety and interactivity of children's learning, but also greatly enhances learning efficiency and effectiveness through personalized feedback mechanisms. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings: Figure 1 Functional block diagram of the children's contextual learning watch system based on environmental language perception and safety knowledge graph provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, unless there is a conflict, the embodiments of the present disclosure and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0019] In some embodiments of the present application, see Figure 1 As shown, this embodiment provides a children's contextual learning watch system based on environmental language perception and safety knowledge graph, including: A graph construction module is configured to obtain security knowledge from multiple channels and construct a security knowledge graph based on the security knowledge; a perception module configured to obtain environmental language information and somatosensory interaction information, perform timbre judgment on the environmental language information, perform perceptual judgment on the somatosensory interaction information, and trigger different modes based on the timbre judgment results and the perceptual judgment results; wherein the different modes include a learning mode and an alarm mode; The perception module is further configured to parse the environmental language information to obtain environmental language content; an analysis and interaction module configured to search the environmental language content in the safety knowledge graph and generate recommended content based on the search results and in combination with the different modes; the analysis and interaction module is further configured to obtain child language information and score the child's answer based on the child language information and the recommended content; The feedback module is configured to determine the next learning content and the next learning interval length based on the child's answer score.

[0020] As can be understood, the present invention, by integrating environmental language perception with a safety knowledge graph, provides a highly interactive and safe learning environment for children. First, the graph construction module constructs a comprehensive safety knowledge graph by integrating safety knowledge from multiple sources. This not only ensures the richness and accuracy of the information, but also provides strong knowledge support for the system, enabling it to better understand and respond to the child's environment. The perception module allows the watch system to acquire environmental language information and somatosensory interaction information in real time. Based on timbre and perceptual judgment, it triggers different modes, such as learning mode and alarm mode, allowing the watch to flexibly adjust its response based on environmental changes and the child's interaction, thereby providing a more personalized learning experience. Furthermore, the analysis function of environmental language information enables the watch to understand and respond to the child's surrounding language environment, further enhancing its interactivity. The analysis and interaction module retrieves environmental language content from the safety knowledge graph and generates recommended content based on different modes. This not only improves the relevance of learning content, but also assesses the child's learning progress and understanding by acquiring language information and scoring responses, thereby providing more accurate and personalized learning recommendations. Finally, the feedback module determines the next learning content and the duration of the learning interval based on the child's response score. This adaptive learning mechanism ensures that children learn in an optimal state, avoiding over-learning or under-learning. Overall, the system's highly integrated modular design not only improves the safety and interactivity of children's learning, but also greatly enhances learning efficiency and effectiveness through personalized feedback mechanisms.

[0021] In some embodiments of the present application, the graph construction module is configured to acquire security knowledge from multiple channels and construct a security knowledge graph based on the security knowledge, including: A security knowledge collection unit is configured to collect multi-source heterogeneous security knowledge from multiple channels; a knowledge fusion unit configured to clean, deduplicate, and standardize the multi-source heterogeneous safety knowledge to extract key safety factors; wherein the key safety factors include dangerous scene features, dangerous speech features, dangerous behavior features, and geographic location risk tags; The graph generation unit is configured to construct a security knowledge graph based on the key security elements.

[0022] It can be understood that the present invention can effectively integrate security information from different sources to ensure the comprehensiveness and accuracy of the information. For example, the security knowledge acquisition unit can collect security incident information from multiple channels such as online news, social media, and security reports. The knowledge fusion unit is responsible for cleaning and standardizing this information to ensure the consistency and comparability of the information, such as removing duplicate information and unifying terminology from different sources. The extraction of key security elements helps to identify and classify potential security threats, such as predicting possible dangerous activities by analyzing specific dangerous scene characteristics, or identifying high-risk areas through geographic location risk labels. Finally, the graph generation unit integrates these key elements into a structured knowledge graph to provide an intuitive and easy-to-understand view for security analysis and decision-making. For example, a security knowledge graph can reveal the pattern of dangerous activities in a specific area within a specific time period, helping law enforcement agencies optimize resource allocation and prevention strategies.

[0023] In some embodiments of the present application, the perception module is configured to obtain environmental language information and somatosensory interaction information, perform timbre judgment on the environmental language information, and perform perception judgment on the somatosensory interaction information, including: Acquire the timbre information pre-recorded in the watch system, compare the ambient timbre of the ambient language information with the timbre information, and determine the timbre consistency between the ambient timbre and the timbre information; if the ambient timbre belongs to one of the timbre information, the timbre determination result is timbre consistency; if the ambient timbre does not belong to any of the timbre information, the timbre determination result is timbre inconsistency; Obtain the distance information of the environmental language information from the watch, compare the distance information with the preset distance, and perform perception judgment. If the distance information is less than the preset distance, the perception judgment result is a dangerous distance; if the distance information is greater than or equal to the preset distance, the perception judgment result is a safe distance.

[0024] In some embodiments of the present application, when the perception module is configured to trigger different modes according to the timbre judgment result and the perception judgment result, the following steps are included: If the timbre judgment result is consistent, and the perception judgment result is dangerous distance, the learning mode is triggered; If the timbre judgment result is consistent, and the perception judgment result is a safe distance, the learning mode is triggered; If the sound color judgment result is inconsistent, and the perception judgment result is dangerous distance, the alarm mode will be triggered, and a level 1 warning will be issued. The alarm number will be automatically dialed and the positioning information will be sent; If the timbre judgment result is inconsistent, and the perception judgment result is a safe distance, the alarm mode will be triggered and a second-level warning will be issued.

[0025] In some embodiments of the present application, the perception module is further configured to parse the environmental language information to obtain environmental language content, further comprising: converting the environmental language information into environmental language content in text form, and dividing the environmental language content into different words; Stop words and redundant modifiers in the vocabulary are removed, and the remaining vocabulary is formed into a vocabulary set.

[0026] Specifically, the perception module first performs a timbre judgment on the ambient language information, comparing the ambient timbre with the timbre information pre-stored in the watch system to determine whether it is consistent. For example, if the speaker in the environment is a trusted person pre-set by the watch owner, the timbre judgment result is consistent, which helps the watch provide personalized services or responses in specific situations. At the same time, the perception module also evaluates the distance information of the ambient language information to determine whether the distance from the watch constitutes a danger. In addition, after the ambient language information is parsed into text content, the perception module further processes this information, removing stop words and redundant modifiers, extracting key words, and forming a vocabulary set. This helps the watch more accurately understand the ambient language content. For example, in specific scenarios, such as meetings or classrooms, the watch can filter out irrelevant words and focus on recording important information. In summary, the present invention enables the perception module to intelligently identify and respond to language information in the environment by integrating timbre judgment, distance perception, and language content analysis, thereby providing users with safer, more convenient, and personalized services.

[0027] In some embodiments of the present application, the analysis and interaction module is configured to search the security knowledge graph for the environmental language content, and generate recommended content based on the search results and in combination with the different modes, including: Searching the vocabulary set in the security knowledge graph to obtain security knowledge nodes associated with the vocabulary set and the association relationships between the nodes; The security knowledge nodes are sorted according to the closeness of the association relationship, and the security knowledge corresponding to a preset number of nodes that are ranked first are extracted.

[0028] In some embodiments of the present application, the analysis and interaction module is configured to search for the environmental language content in the security knowledge graph, and generate recommended content based on the search results and in combination with the different modes, further comprising: If in learning mode, the extracted safety knowledge will be converted into learning content in the form of pictures, texts, animations or voice as recommended content; If in alarm mode, during the first-level warning, emergency response methods will be screened out from the extracted safety knowledge and reported quickly as recommended content; during the second-level warning, the extracted safety knowledge will be integrated into warning information, displayed synchronously on the watch screen and pushed to the guardian's terminal as recommended content.

[0029] In some embodiments of the present application, the analysis and interaction module is further configured to obtain child language information, and score the child's answer based on the child language information and the recommended content, including: Obtaining child language information generated by the child in response to the environmental language information, and comparing the child language information with the recommended content to calculate answer similarity; A similarity interval is set, and if the answer similarity is less than a minimum value of the similarity interval, the child's answer score is the first score; If the answer similarity is within the similarity range, the child's answer score is the second score; If the answer similarity is greater than the maximum value of the similarity interval, the child's answer score is the third score; The first score is smaller than the second score, and the second score is smaller than the third score.

[0030] It is understandable that the analysis and interaction module can provide personalized information and guidance by retrieving environmental language content in the safety knowledge graph and generating recommended content in combination with different modes. Specifically, the module first retrieves the safety knowledge nodes related to the environmental language content and their associations, and then sorts the nodes according to the closeness of these relationships to extract the most relevant safety knowledge. For example, when asking a safety question about "fire", the system will find safety nodes related to "fire" in the knowledge graph, such as "Do not play with fire", "Fire alarm phone number", etc., and sort them according to the association between them, giving priority to extracting the safety knowledge most relevant to the question.

[0031] Specifically, in learning mode, this safety knowledge can be transformed into learning content in the form of graphics, text, animation, or audio, helping children better understand and retain it. For example, the system can transform the safety knowledge "Don't play with fire" into an animated video, allowing children to learn through watching the animation. In alarm mode, the system provides emergency response measures or warnings based on the warning level. For example, in a level one warning, the system can quickly announce emergency response measures such as "If you see a fire, tell an adult immediately." In a level two warning, the system integrates safety knowledge into a warning message, displays it on the watch screen, and pushes it to the guardian's terminal to ensure timely transmission and processing of the information. In addition, the analysis and interaction module can also obtain children's language information and assign a score based on the similarity between the child's answer and the recommended content. For example, if a child answers "Don't play with matches," the system will compare this answer with the recommended content "Don't play with fire" and assign a score based on the similarity range. This scoring mechanism helps assess children's understanding of safety knowledge and provides feedback for subsequent education.

[0032] In summary, by retrieving and analyzing environmental language content in the safety knowledge graph, combining different modes to generate recommended content, and scoring children's answers, the present invention can provide more personalized, timely and effective safety education and emergency response, thereby better protecting children's safety.

[0033] In some embodiments of the present application, the feedback module is configured to determine the next learning content and the next learning interval duration based on the child's answer score, including: If the score is the first, then obtaining the same type of questions with a similarity greater than a similarity threshold with the environment language information from the security knowledge graph, and determining the next learning content to be the same type of questions; If the score is the second score, similar questions whose similarity to the environment language information is less than the similarity threshold and greater than zero are obtained from the security knowledge graph, and the next learning content is determined to be the similar questions; If the score is the third score, irrelevant questions with zero similarity to the environmental language information are obtained from the security knowledge graph, and the next learning content is determined to be the irrelevant questions.

[0034] In some embodiments of the present application, when the feedback module is configured to determine the next learning content and the next learning interval duration based on the child's answer score, it further includes: Setting a standard interval length, and adjusting the standard interval length using an adjustment coefficient based on the child's answer score to obtain the next learning interval length; The child's answer score is proportional to the adjustment coefficient, and the adjustment coefficient has a value range of (0, 1], and the next learning interval duration is the product of the standard interval duration and the adjustment coefficient.

[0035] It's understood that the feedback module uses the scores of children's answers to optimize learning content and the duration of learning intervals, thereby improving learning efficiency and providing a personalized experience. Specifically, if a child's answer receives a first score, indicating an incorrect answer and lack of understanding, the system will select similar questions from the safety knowledge graph that are highly similar to the surrounding language information as the next learning content to consolidate and deepen understanding. For example, if a child incorrectly answers a question about traffic lights while learning about traffic safety, the system may provide another question about traffic signs. If a child's answer receives a second score, indicating a correct answer but average understanding, the system will select similar questions with lower similarity to the surrounding language information as the next learning content to promote further understanding and expansion of knowledge. For example, if a child correctly answers a question about home safety, the system may provide a question about how to properly use household appliances. If a child's answer receives a third score, indicating a correct answer and deep understanding, the system will select an unrelated question completely unrelated to the surrounding language information as the next learning content to avoid duplication of learning. For example, if a child correctly answers a question about traffic safety, the system may provide a question about healthy eating.

[0036] In addition, the feedback module adjusts the interval between learning sessions based on the child's response score. The system sets a standard interval length and adjusts it based on the score using an adjustment factor. The lower the score, the smaller the adjustment factor, and the shorter the interval, ensuring that children can move on to the next stage of learning promptly after mastering the knowledge. For example, if a child receives high scores several times in a row, the system may extend the interval between learning sessions. In this way, the learning system can dynamically adjust the learning content and pace based on the child's learning progress, ensuring consistency and depth of learning while avoiding repetition and boredom, thereby improving learning efficiency and interest.

[0037] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0038] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0039] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0040] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0041] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A children's contextual learning watch system based on environmental language perception and safety knowledge graph, characterized by: include: A graph construction module is configured to obtain security knowledge from multiple channels and construct a security knowledge graph based on the security knowledge; a perception module configured to obtain environmental language information and somatosensory interaction information, perform timbre judgment on the environmental language information, perform perceptual judgment on the somatosensory interaction information, and trigger different modes based on the timbre judgment results and the perceptual judgment results; wherein the different modes include a learning mode and an alarm mode; The perception module is further configured to parse the environmental language information to obtain environmental language content; an analysis and interaction module configured to search the environmental language content in the safety knowledge graph and generate recommended content based on the search results and in combination with the different modes; the analysis and interaction module is further configured to obtain child language information and score the child's answer based on the child language information and the recommended content; The feedback module is configured to determine the next learning content and the next learning interval length based on the child's answer score.

2. The children's contextual learning watch system based on environmental language perception and safety knowledge graph according to claim 1 is characterized in that: The graph construction module is configured to acquire security knowledge from multiple channels and construct a security knowledge graph based on the security knowledge, including: A security knowledge collection unit is configured to collect multi-source heterogeneous security knowledge from multiple channels; a knowledge fusion unit configured to clean, deduplicate, and standardize the multi-source heterogeneous safety knowledge to extract key safety factors; wherein the key safety factors include dangerous scene features, dangerous speech features, dangerous behavior features, and geographic location risk tags; The graph generation unit is configured to construct a security knowledge graph based on the key security elements.

3. The children's contextual learning watch system based on environmental language perception and safety knowledge graph according to claim 1 is characterized in that: The perception module is configured to obtain environmental language information and somatosensory interaction information, perform timbre judgment on the environmental language information, and perform perception judgment on the somatosensory interaction information, including: Acquire the timbre information pre-recorded in the watch system, compare the ambient timbre of the ambient language information with the timbre information, and determine the timbre consistency between the ambient timbre and the timbre information; if the ambient timbre belongs to one of the timbre information, the timbre determination result is timbre consistency; if the ambient timbre does not belong to any of the timbre information, the timbre determination result is timbre inconsistency; Obtain the distance information of the environmental language information from the watch, compare the distance information with the preset distance, and perform perception judgment. If the distance information is less than the preset distance, the perception judgment result is a dangerous distance; if the distance information is greater than or equal to the preset distance, the perception judgment result is a safe distance.

4. The children's contextual learning watch system based on environmental language perception and safety knowledge graph according to claim 3 is characterized in that: The perception module is configured to trigger different modes according to the timbre judgment result and the perception judgment result, including: If the timbre judgment result is consistent, and the perception judgment result is dangerous distance, the learning mode is triggered; If the timbre judgment result is consistent, and the perception judgment result is a safe distance, the learning mode is triggered; If the sound color judgment result is inconsistent, and the perception judgment result is dangerous distance, the alarm mode will be triggered, and a level 1 warning will be issued. The alarm number will be automatically dialed and the positioning information will be sent; If the timbre judgment result is inconsistent, and the perception judgment result is a safe distance, the alarm mode will be triggered and a second-level warning will be issued.

5. The children's contextual learning watch system based on environmental language perception and safety knowledge graph according to claim 4 is characterized in that: The perception module is further configured to parse the environmental language information to obtain environmental language content, further comprising: converting the environmental language information into environmental language content in text form, and dividing the environmental language content into different words; Stop words and redundant modifiers in the vocabulary are removed, and the remaining vocabulary is formed into a vocabulary set.

6. The children's contextual learning watch system based on environmental language perception and safety knowledge graph according to claim 5 is characterized in that: The analysis and interaction module is configured to search the environmental language content in the security knowledge graph, and generate recommended content based on the search results and in combination with the different modes, including: Searching the vocabulary set in the security knowledge graph to obtain security knowledge nodes associated with the vocabulary set and the association relationships between the nodes; The security knowledge nodes are sorted according to the closeness of the association relationship, and the security knowledge corresponding to a preset number of nodes that are ranked first are extracted.

7. The children's contextual learning watch system based on environmental language perception and safety knowledge graph according to claim 6 is characterized in that: The analysis and interaction module is configured to search for the environmental language content in the security knowledge graph, and generate recommended content based on the search results and in combination with the different modes, further comprising: If in learning mode, the extracted safety knowledge will be converted into learning content in the form of pictures, texts, animations or voice as recommended content; If in alarm mode, during the first-level warning, emergency response methods will be screened out from the extracted safety knowledge and reported quickly as recommended content; during the second-level warning, the extracted safety knowledge will be integrated into warning information, displayed synchronously on the watch screen and pushed to the guardian's terminal as recommended content.

8. The children's contextual learning watch system based on environmental language perception and safety knowledge graph according to claim 7 is characterized in that: The analysis and interaction module is further configured to obtain child language information, and score the child's answer based on the child language information and the recommended content, including: Obtaining child language information generated by the child in response to the environmental language information, and comparing the child language information with the recommended content to calculate answer similarity; A similarity interval is set, and if the answer similarity is less than a minimum value of the similarity interval, the child's answer score is the first score; If the answer similarity is within the similarity range, the child's answer score is the second score; If the answer similarity is greater than the maximum value of the similarity interval, the child's answer score is the third score; The first score is smaller than the second score, and the second score is smaller than the third score.

9. The children's contextual learning watch system based on environmental language perception and safety knowledge graph according to claim 8 is characterized in that: The feedback module is configured to determine the next learning content and the next learning interval duration according to the child's answer score, including: If the score is the first, then obtaining the same type of questions with a similarity greater than a similarity threshold with the environment language information from the security knowledge graph, and determining the next learning content to be the same type of questions; If the score is the second score, similar questions whose similarity to the environment language information is less than the similarity threshold and greater than zero are obtained from the security knowledge graph, and the next learning content is determined to be the similar questions; If the score is the third score, irrelevant questions with zero similarity to the environmental language information are obtained from the security knowledge graph, and the next learning content is determined to be the irrelevant questions.

10. The children's contextual learning watch system based on environmental language perception and safety knowledge graph according to claim 9 is characterized in that: The feedback module is configured to determine the next learning content and the next learning interval duration according to the child's answer score, and further includes: Setting a standard interval length, and adjusting the standard interval length using an adjustment coefficient based on the child's answer score to obtain the next learning interval length; The child's answer score is proportional to the adjustment coefficient, and the adjustment coefficient has a value range of (0, 1], and the next learning interval duration is the product of the standard interval duration and the adjustment coefficient.

Images