A nursing intelligent teaching auxiliary system and method

Abstract

The application relates to the technical field of teaching, and particularly discloses a nursing intelligent teaching auxiliary system and method, which comprises a cloud server, a plurality of teacher terminals and a plurality of student terminals; the cloud server comprises a knowledge base construction module, which is used for acquiring teaching documents of adult nursing science, and establishing structured course knowledge graph data based on the teaching documents; a model construction module, which is used for acquiring a general large language model, and performing field adaptability training on the general model by using the course knowledge graph data, and outputting a special model fused with adult nursing knowledge; a teaching assistant module, which is used for accepting a lesson preparation request of the teacher terminal, inputting the special model, and generating lesson preparation information; and a student assistant module, which is used for receiving an auxiliary request of the student terminal, calling the special model according to the content input by the student, and generating auxiliary teaching content. The technical scheme of the application can meet the learning needs of different students and effectively improve the teaching quality.

Description

Technical Field

[0001] This invention relates to the field of teaching technology, and in particular to a nursing intelligent teaching aid system and method. Background Technology

[0002] Through long-term dedication to nursing education, especially adult nursing practical training, we have found through extensive classroom observation and teaching debriefing that the existing blended learning model is inadequate in addressing the core issue of dynamic cognitive bias.

[0003] We have observed that nursing education is not merely the transmission of knowledge, but also a high-intensity, hands-on exercise of clinical reasoning. In actual teaching scenarios, students often face a dual dilemma: firstly, the difficulty in acquiring knowledge. Traditional teaching resources are mostly static documents or general courseware, lacking in-depth understanding and real-time interactive capabilities specific to the nursing profession (such as complex clinical cases and the latest nursing guidelines). This makes it difficult for students to obtain precise and professional in-depth guidance when preparing lessons or studying independently, hindering the transformation of teaching content into students' clinical decision-making abilities.

[0004] Secondly, there are cognitive blind spots under high-pressure classroom conditions. When we immediately correct students' operational errors or frequently question them during practical training sessions, students often experience significant psychological pressure due to tension. This state of psychological stress can cause students to experience auditory filtering or fragmented memory in an instant. Even if the teacher provides crucial guidance, students may miss the most important logical chain because their attention is overly focused on the error itself. At this time, existing teaching management methods are often static and isolated, making it difficult for teachers to quantify the psychological load and information reception quality of students at a specific moment.

[0005] Therefore, there is an urgent need for a nursing intelligent teaching support system that can meet the learning needs of different students and effectively improve the quality of teaching. Summary of the Invention

[0006] One of the objectives of this invention is to provide a nursing intelligent teaching assistance system that can meet the learning needs of different students and effectively improve teaching quality.

[0007] To solve the above-mentioned technical problems, this application provides the following technical solution: A nursing intelligent teaching assistance system includes a cloud server, several teacher terminals, and several student terminals; The cloud server includes a knowledge base building module, which is used to obtain teaching documents for adult nursing and build structured course knowledge graph data based on the teaching documents; The model building module is used to obtain a general large language model, use course knowledge graph data to perform domain-adaptive training on the general model, and output a custom model that integrates adult care knowledge. The teaching assistant module is used to accept lesson preparation requests from teachers, input a unique model, and generate lesson preparation information. The learning assistance module is used to receive assistance requests from students, and based on the content input by the students, it calls a dedicated model to generate supplementary teaching content.

[0008] Furthermore, the cloud server also includes a teaching management module, which is used to obtain class information, bind several student accounts that have joined the same teaching class to the teacher account that is responsible for the class, and grant the teacher account management permissions to the bound student accounts. The learning support module is also used to collect interaction data between students and the dedicated model, generate student learning analysis reports, and push the learning analysis reports to the teachers of the corresponding classes.

[0009] Furthermore, the cloud server also includes a classroom situation analysis module, which is used to respond to a wake-up request sent by any student terminal, send an activation command to the corresponding teacher terminal and the target student terminal, and mark the student terminal as an auxiliary student terminal; Among them, the auxiliary student terminal is another student terminal that is in the same teaching class as the target student terminal; the wake-up request contains information about the target student terminal. The teacher's end is used to collect the first audio data and upload it to the classroom situation analysis module after receiving the activation command; The target student terminal is used to collect second audio data and upload it to the classroom situation analysis module after receiving the activation command; The classroom situation analysis module is also used to extract decibels from the first audio data, perform emotion intensity analysis based on acoustic features, and process speech to text to obtain the teacher's text, and to extract decibels from the second audio data and process speech to text to obtain the student's text. The classroom situation analysis module is also used to calculate the decibel difference between the first audio data and the second audio data to determine the distance of the sound source, and to calculate the text similarity between the teacher's text and the student's received text to determine the transmission clarity.

[0010] Furthermore, the classroom situation analysis module is also used to calculate the stress index of the student corresponding to the target student based on the analysis results of sound source distance, transmission clarity, and emotional intensity. Among them, the stress index is negatively correlated with the distance from the sound source, negatively correlated with the transmission clarity, and positively correlated with the intensity of emotions; if the stress index exceeds the preset stress threshold, the classroom situation analysis module is also used to send an auxiliary marker request to the auxiliary student terminal and receive the first key timestamp sequence of feedback from the auxiliary student terminal. After the teacher stops collecting the first audio data, the classroom situation analysis module encapsulates the teacher's text into a review note data package and sends it to the target student's end, and obtains the second key timestamp sequence marked in the review note data package by the target student's end; The classroom situation analysis module is also used to compare the second key timestamp sequence with the first key timestamp sequence. If a missing time period is identified that exists in the first key timestamp sequence but is missing in the second key timestamp sequence, the content corresponding to the missing time period in the teacher's text is extracted, and a dedicated model is called in combination with preset prompt words to generate multi-dimensional analytical notes for the missing time period.

[0011] Furthermore, the classroom situation analysis module is also used to send multi-dimensional analysis notes for the missed time period to the auxiliary student terminal, receive selection information for any dimension from the auxiliary student terminal, mark the dimension, and send the marked multi-dimensional analysis notes for the missed time period to the target student terminal.

[0012] Furthermore, the target student terminal is also used to receive initial cognitive dimension selection information for the multi-dimensional analysis notes for the missed time period, mark the dimension, and send the multi-dimensional analysis notes, which have been compared by both the auxiliary student terminal and the target student terminal, to the teacher terminal.

[0013] Furthermore, the classroom situation analysis module is also used to send the teacher's text to the teacher's end. After receiving the approval information from the teacher's end, the teacher's text is packaged into a review notes data package and sent to the target student's end.

[0014] Furthermore, the classroom situation analysis module is also used to retrieve the usage records of the target student's terminal and, in conjunction with the analysis of the real-time teacher text converted from the first audio data, determine the teacher's current explanation content and judge whether the target student has previewed the explanation content in advance. If the judgment result is that the target student has not previewed, the classroom situation analysis module still sends an auxiliary tag request to the auxiliary student terminal. If the judgment result indicates that the target student has already previewed the material, the classroom situation analysis module is also used to filter out other students in the same teaching class who have not previewed the material, send auxiliary tag requests to these students, and mark the students who accept the request as assisting students.

[0015] Furthermore, the student terminal also includes a motion sensor for real-time acquisition of device posture data to assist the student. The classroom situation analysis module is also used to receive device posture data uploaded by the auxiliary student terminal or the assisting student terminal, and combine it with the synchronously collected first audio data to determine the focus status; When a student is determined to be in a focused state, the classroom situation analysis module is also used to perform dynamic expansion of the matching range of timestamps in the first key timestamp sequence generated in that state.

[0016] The second objective of this invention is to provide a nursing intelligent teaching aid method using the aforementioned system.

[0017] This solution provides precise lesson preparation and learning support for teachers and students by constructing a dedicated large language model that integrates a knowledge graph of the nursing field. Furthermore, through a classroom situation analysis module, it can perceive students' psychological stress and information reception quality under high-pressure teaching scenarios in real time. By mobilizing classmates to collaboratively mark key teaching content, it intelligently identifies cognitive gaps in target students and automatically generates multi-dimensional, personalized analytical notes. This effectively compensates for cognitive blind spots caused by tension during nursing practical training and also provides teachers with optimized teaching strategies, thereby improving the quality of nursing education. Attached Figure Description

[0018] Figure 1 This is a logic block diagram of an embodiment of a nursing intelligent teaching assistance system. Detailed Implementation

[0019] The following detailed description illustrates the specific implementation method: Example 1 like Figure 1 As shown in this embodiment, a nursing intelligent teaching assistance system is applied in a blended teaching scenario for adult nursing, including a cloud server, several teacher terminals, and several student terminals.

[0020] The cloud server serves as the core computing node, enabling data interaction with several teacher and student terminals via the network. In this embodiment, both teacher and student terminals are smartphones running a pre-installed app, distinguished by their logged-in accounts.

[0021] The cloud server includes modules for knowledge base construction, model construction, teaching management, teaching assistants, learning assistance, and classroom situation analysis.

[0022] The knowledge base construction module is used to acquire teaching documents for adult nursing and build a structured course knowledge graph based on these documents. Teaching documents include textbooks, guidelines, case studies, and journal articles; for example, the "Adult Nursing" textbook, clinical operation guidelines, critical care case studies, and the latest core nursing journal articles. In this embodiment, a structured course knowledge graph is constructed using named entity recognition and relation extraction technologies.

[0023] The model building module acquires a pre-defined general-purpose language model and uses course knowledge graph data to incrementally pre-train and fine-tune this general-purpose model, enabling it to complete domain-adaptive training and output a customized model with profound nursing professional competence. This customized model can understand complex nursing terminology and possesses logically rigorous clinical diagnostic abilities.

[0024] The teaching management module is used to obtain class information, bind several student accounts that have joined the same teaching class to the teacher account that is responsible for the class, and grant the teacher account management permissions to the bound student accounts.

[0025] In the daily lesson preparation process, the teaching assistant module receives lesson preparation requests from teachers, inputs a customized model, and generates lesson preparation information, including a teaching syllabus, detailed case studies, and teaching quizzes. The teaching quizzes are automatically generated based on knowledge points. When teachers create teaching materials, the teaching assistant module also identifies keywords in the materials and inserts the generated cases or quizzes as interactive components into the materials to improve lesson preparation efficiency.

[0026] On the student learning side, the learning support module provides 24-hour intelligent learning companion services through the student's app. When students are previewing or reviewing adult nursing knowledge, the student app receives content input via voice or text, invokes a dedicated model, generates supplementary teaching content, and provides answers to questions. Supplementary teaching content includes: answers to nursing knowledge points, recommendations for nursing literature, and tips on key points of nursing skills operation. For example, when a student has questions about the postural care of acute heart failure, the dedicated model can not only provide the standard answer but also recommend relevant latest research literature or demonstration videos, achieving personalized learning support guidance.

[0027] The learning support module also collects interaction data between students and the dedicated model, generates student learning analysis reports, and pushes these reports to the teachers' terminals for the corresponding classes. In this embodiment, the learning support module records all interaction logs between students and the dedicated model in real time in the background, including the distribution of knowledge points asked, interaction frequency, and requests for analysis of specific nursing cases. Algorithms are then used to statistically categorize this behavioral data. The generated learning analysis reports include heatmaps of students' mastery of knowledge points in different nursing chapters, learning activity trends, frequently occurring mistakes, and targeted review strategy suggestions.

[0028] In practical training or case discussion classes, when an instructor corrects or questions a specific student's nursing procedure (such as an error in aseptic technique), other students in the same class can send a wake-up request through their student devices. The classroom situation analysis module responds to wake-up requests sent by any student device. At this time, the student device that sent the request is marked as an auxiliary student device by the system.

[0029] The classroom situation analysis module, upon receiving a wake-up request, immediately sends activation commands to both the teacher's and the target student's devices, initiating audio acquisition at both ends. At this time, the teacher's device collects the first audio data, and the target student's device collects the second audio data, both of which are uploaded synchronously.

[0030] The classroom situation analysis module is also used to compare the decibel difference between the first and second audio data, and to estimate the sound source distance between the teacher and the target student using a sound wave attenuation model. In this embodiment, the classroom situation analysis module performs time alignment on the audio data collected synchronously by the teacher and student, and extracts the real-time decibel values ​​of the audio from both ends. Since sound waves experience predictable intensity loss as they propagate through the air, and the teacher's microphone is close to the teacher (near the sound source), the decibel value it collects can be considered the initial intensity, while the student's microphone collects the intensity after spatial propagation. In this embodiment, by calculating the decibel difference between the two and combining it with a preset sound attenuation coefficient for the indoor environment (e.g., a correction value considering wall reflection or environmental absorption), this energy difference can be converted into a specific physical length, thereby deducing the sound source distance between the teacher and the target student.

[0031] The classroom situation analysis module is also used to analyze the acoustic characteristics of the teacher's tone and speed in the first audio data to assess the teacher's emotional intensity. In this embodiment, the classroom situation analysis module converts the audio into a text stream and calculates the teacher's real-time speaking speed based on the number of characters or syllables identified per unit time. The classroom situation analysis module is also used to extract the tone characteristics of the voice through a preset fundamental frequency tracking algorithm, that is, to analyze the frequency fluctuations of vocal cord vibration. Usually, when emotions are agitated, the mean and dynamic range of the fundamental frequency will increase significantly. The classroom situation analysis module pre-stores the acoustic benchmark model of the teacher under normal teaching conditions as a comparison reference. The classroom situation analysis module calculates the difference between the real-time collected speaking speed, tone fluctuations, and sound loudness (decibels) and the benchmark model. For example, when a sudden increase in speaking speed and a significant upward shift in the fundamental frequency center point are detected, the system will map the deviation of these physical characteristics to a continuous numerical range between 0 and 1 according to a preset weight allocation (e.g., tone fluctuation accounts for 40%, speaking speed change accounts for 30%, and energy change accounts for 30%). A high tone and fast speaking speed correspond to a high emotional intensity value. Ultimately, this weighted and aggregated value becomes the output emotional intensity index, reflecting the stress level of the teacher's current verbal expression.

[0032] The classroom situation analysis module also performs speech-to-text processing on the first audio data to obtain the teacher's text, and performs speech-to-text processing on the second audio data to obtain the student's received text. The similarity between the two texts is used to measure the clarity of information transmission received at the target student's location. In this embodiment, the string similarity between the two texts is calculated to obtain a value between 0 and 1 as a representation of transmission clarity. When this value is close to 1, it indicates that the audio content received by the student is highly consistent with the language content spoken by the teacher; conversely, if the recognized text has a large number of missing or ambiguous parts due to a noisy training environment, obstruction, or excessive distance from the sound source, the text similarity will decrease significantly.

[0033] Based on the above parameters, the classroom situation analysis module is also used to calculate the stress index of the target students in real time according to the distance of the sound source, the intensity of the teacher's emotions, and the clarity of the transmission.

[0034] Specifically, the classroom situation analysis module is used to normalize the raw data of sound source distance, emotional intensity, and transmission clarity. It uses an inverse proportional function or a piecewise function to map the sound source distance to a value between 0 and 1, that is, the smaller the sound source distance, the closer the value is to 1.

[0035] Clarity stress factor , where C represents transmission resolution.

[0036] The classroom situation analysis module uses a weighted summation algorithm to calculate the final stress index. The formula is:

[0037] in, Distance from the sound source Intensity of emotion; These are preset weighting coefficients for sound source distance, emotional intensity, and transmission clarity, respectively. The weighting coefficients can be preset according to the characteristics of nursing practical training courses. For example, considering that the teacher's strict explanation (emotion) and close guidance (distance) are the main stressors in practical training courses, the following weighting coefficients can be set: .

[0038] Real-time calculation If the value exceeds the preset pressure threshold, such as 0.75, it is determined that the target student is in a high-pressure state, which triggers the subsequent process of sending an auxiliary tag request to the auxiliary student terminal.

[0039] In this embodiment, the stress index will significantly increase when the teacher and students are too close, the teacher is emotionally agitated, or the environment is noisy and reduces clarity. Once the stress index exceeds a preset stress threshold, the system determines that the target student may experience cognitive biases or memory omissions due to excessive tension. At this time, the classroom situation analysis module sends an auxiliary marking request to the auxiliary student's device. As an observer, the auxiliary student is more composed and can calmly mark the timestamps of key knowledge points on their own device based on the teacher's criticism or guidance, obtaining a first-priority timestamp sequence, which can effectively compensate for the target student's cognitive deficiencies under pressure. In this embodiment, after the auxiliary student agrees to the marking, the timestamp can be marked by pressing the volume button on a smartphone; the content within 3 seconds before and after the timestamp is considered key knowledge points.

[0040] After the classroom correction or guidance process concludes, i.e., after the teacher stops collecting the first audio data, the classroom situation analysis module is also used to send the automatically summarized teacher text to the teacher's end. After receiving the approval information from the teacher's end, the teacher's text is packaged into a review note data package and sent to the target student's end. In this embodiment, the process can be automatically terminated by the teacher's end or terminated by the student's end sending a request.

[0041] After calming down, the target student independently marks the key points in their review notes, generating a second key timestamp sequence, which is received by the target student. The classroom situation analysis module also compares the first key timestamp sequence marked by the assistant student with the second key timestamp sequence marked by the target student. In this embodiment, timestamps falling within 3 seconds of each other are considered the same. If it is found that the target student has missed some key guidance (i.e., content present in the first key timestamp sequence but missing in the second key timestamp sequence), the classroom situation analysis module also uses preset prompts to drive a dedicated model to generate multi-dimensional analytical notes for the missed content from multiple dimensions, including theoretical principles, clinical cases, and comparisons of common mistakes.

[0042] For example, in a nursing practice class, the teacher noticed that student A (the target student) forgot to double-check the patient's information before performing an intravenous infusion on a simulated patient. The teacher immediately corrected him, speaking in a rather stern tone. Student A, due to nervousness, only remembered that "the teacher said I checked it wrong," but failed to fully understand and remember the teacher's explanations regarding "why double-checking is necessary" and "how to perform standard procedures."

[0043] After the system was activated, it helped students mark the key points in the teacher's explanation regarding the importance of secondary verification and the standard verification process. After class, student A only marked the parts of the verification that they remembered making mistakes in their review notes. After comparison by the system, it was found that student A had missed the crucial explanation of the connection between secondary verification and medication safety laws.

[0044] For this missed time period (i.e., the part where the teacher explains the legal relevance), the classroom situation analysis module calls upon a dedicated model, combined with preset prompts (such as "generate an analysis of the legal requirements for secondary verification in nursing procedures, covering theory, cases, and risk comparisons"), to generate a multi-dimensional analytical note: From a theoretical perspective, this can be explained from the angles of nursing ethics and laws and regulations. Double verification (i.e., verification before, during, and after the procedure) is not only an operational standard but also a mandatory requirement of the "Regulations on the Handling of Medical Accidents" and core hospital systems. It constitutes a crucial link in the "three lines of defense" for medical safety, and its legal significance lies in elevating nursing actions from "technical operations" to "legally responsible actions." Ignoring this step could directly constitute "negligent tort" should an error occur.

[0045] From a clinical case perspective, let's cite a publicly reported adverse nursing event: A nurse at a hospital, failing to strictly adhere to the double-check procedure, mistakenly administered antibiotics intended for patient in bed 8 to patient in bed 9. This resulted in a severe allergic reaction in patient 9, who survived after emergency treatment. The incident was ultimately classified as a level three medical malpractice incident. The nurse involved was suspended from practice and given a hospital-wide notice of punishment. The department and the hospital also incurred corresponding financial compensation and reputational damage.

[0046] Operational Steps: First Check (Before Operation): When preparing medications in the treatment room, verify the doctor's orders, medication labels, and patient information sheet. Second Check (Before Bedside Operation): Use at least two verification methods, such as scanning the patient's wristband barcode and asking open-ended questions like, "Auntie, could you please tell me your full name and date of birth?". Third Check (During / After Operation): Double-check that the IV bag / bottle matches the patient's information.

[0047] In this embodiment, in order to ensure that the generated analytical notes can be understood by the target students, the classroom situation analysis module is also used to send the generated multi-dimensional analytical notes to the auxiliary student terminal first.

[0048] The system assists students in selecting the dimension they understand from multiple dimensions of analysis based on their real-time observations of the classroom situation. The student's device receives the selection information for any dimension and marks that dimension. The classroom situation analysis module is also used to send the marked multi-dimensional analysis notes for the missed time period to the target student's device.

[0049] The target student's end also receives initial cognitive dimension selection information for the multi-dimensional analytical notes on the missed time periods, marks these dimensions, and sends the multi-dimensional analytical notes, compared between the auxiliary student's end and the target student's end, to the teacher's end. This helps teachers intuitively understand the information reception blind spots of students under pressure, thereby optimizing subsequent teaching methods and communication strategies.

[0050] This embodiment also provides a nursing intelligent teaching aid method, using the above-mentioned system.

[0051] In nursing training or case discussions, when students receive one-on-one guidance or are frequently questioned by instructors due to improper procedures, the proximity, the instructor's stern tone, or the rapid pace of speech often cause significant stress for the students. This tension prevents them from fully absorbing the instructor's guidance, and they may also struggle to recall it completely during post-instruction review. This solution introduces a secondary student from the same class as a calm observer. When neither the student nor the instructor has time to collect information, the secondary student's and the student's devices are activated to record the instructor's guidance via speech-to-text, facilitating a more comprehensive review for the student later.

[0052] This solution also utilizes a classroom situation analysis module to comprehensively calculate decibel differences, emotional intensity, and text similarity to quantitatively assess the classroom teaching atmosphere, thereby characterizing the current stress level of the target students. If the stress level is high and they are tense, they are prone to cognitive blind spots due to psychological stress, making it difficult to fully understand what the teacher is saying.

[0053] By inviting assisted students to mark the teacher's key teaching points with a first-key timestamp sequence, the objective perspective of the assisted students compensates for the subjective cognitive deficiencies of the target students. This ensures that key teaching knowledge points and operational corrections are fully and accurately captured and recorded under the high-pressure environment of the classroom, avoiding the loss of teaching information due to the target students' anxiety. Assisted students only mark the key content they hear from the teacher with timestamps, minimizing disruption to their normal classroom learning.

[0054] The system compares the first set of key timestamps marked by the student with the second set of key timestamps marked by the student during subsequent review. This allows for the precise identification of content that an observer might perceive as important, but the student themselves might not perceive as such. For these specific cognitive omissions, such as those covered by a customized model, the system generates analytical notes encompassing theoretical principles, clinical cases, and other dimensions. This provides the student with more detailed and logically rigorous supplementary knowledge than the original teacher explanation, significantly improving their grasp of weak areas and enhancing their understanding of the knowledge points.

[0055] After generating multi-dimensional analytical notes for the missed time periods, the system first filters the notes on the student's end and then confirms them on the target student's end. Finally, the analytical notes that have been doubly confirmed are fed back to the teacher's end. This allows the teacher to intuitively see the blind spots in students' thinking under pressure and the explanation dimensions that students are most likely to accept. It helps the teacher to objectively evaluate the effectiveness of their expression in high-pressure teaching scenarios, thereby optimizing the tone of subsequent lectures, communication strategies and knowledge presentation methods in a targeted manner, and realizing two-way dynamic optimization of teaching and learning.

[0056] Example 2 The difference between this embodiment and Embodiment 1 is that the classroom situation analysis module in this embodiment has different filtering logic when selecting which student should perform the auxiliary marking request.

[0057] Specifically, the classroom situation analysis module also retrieves the usage records of the target student's device and, combined with the analysis of the real-time teacher text converted from the first audio data, determines the teacher's current explanation content and whether the target student has previewed the content. If the result indicates that the target student has not previewed the content, the classroom situation analysis module still sends an auxiliary tag request to the auxiliary student's device while avoiding interference with the learning of other students.

[0058] If the assessment result indicates that the target student has previewed the material but still receives corrections or one-on-one explanations from the teacher during practical training due to non-compliance with operational requirements, it is determined that the target student may have a deep comprehension bias or cognitive rigidity. In this case, the classroom situation analysis module is also used to filter out other student terminals in the same teaching class that have not previewed the content being explained, send auxiliary tagging requests to these student terminals, and mark the student terminals that accept the requests as assisting student terminals.

[0059] The corresponding assisting students mark the key knowledge points with timestamps on their respective assisting student terminals according to the teacher's explanation. The assisting student terminals then generate an initial first key point timestamp sequence and upload it.

[0060] The classroom situation analysis module is also used to receive initial first-key timestamp sequences from multiple assisting student terminals, and to merge multiple initial third-key timestamp sequences to obtain the final first-key timestamp sequence, which is used for subsequent comparison with the second-key timestamp sequence marked by the target student.

[0061] During the note-taking generation phase, students are assisted in selecting the dimension they understand from multiple dimensions of analysis based on their real-time observations of the classroom situation. The student's device receives the selection information for any given dimension and marks that dimension, thereby providing the target student with an intuitive cognitive reference based on other students' perspectives.

[0062] When target students still experience misunderstandings or make operational errors despite completing pre-reading, this is usually not due to a lack of basic knowledge, but rather to potentially deep-seated cognitive rigidity or misunderstandings. By selecting students in the same class who did not pre-read as assistants, an objective perspective with zero baseline is introduced. These students, unaffected by the prior logic of the pre-reading material, can more purely record key points based on the teacher's on-site explanation. This complementary perspective based on different cognitive backgrounds breaks down the target students' fixed mindsets and provides them with a completely different dimension of understanding than their original pre-reading path.

[0063] Since individual students inevitably experience fluctuations in attention or comprehension biases in real-time classrooms, the labeling by a single assistant may be incomplete. However, by merging multiple initial sequences, outliers can be eliminated using collective intelligence, and a final, high-confidence first key timestamp sequence with commonalities can be extracted.

[0064] This solution also allows students to choose analytical dimensions based on the real-time classroom situation and their own understanding. This results in analytical notes that are no longer dry theoretical compilations, but rather more easily absorbed knowledge slices transformed by peer understanding. For students who still struggle to understand the target material after previewing, this multi-dimensional analysis from a non-preview perspective, sharing common understanding with peers, provides a more persuasive and inspiring cognitive reference, significantly improving the quality of their post-lesson review.

[0065] Example 3 In actual teaching scenarios, when the content explained by the teacher is extremely important, the assistant or student can easily enter a deep listening state, that is, become engrossed in listening. At this time, their thinking closely follows the teacher's logic. Often, by the time they realize they need to mark the point, the teacher has already explained that knowledge point, causing the timestamp generated by the assistant to lag behind the actual time when the content occurs.

[0066] To address this issue, the student terminal in this embodiment also includes motion sensors (such as accelerometers and gyroscopes) to collect real-time device posture data to assist the student terminal.

[0067] The classroom situation analysis module is also used to receive device posture data uploaded by the auxiliary student terminal or the assisting student terminal, and combine it with the synchronously collected first audio data to determine the focus status; If the device posture data shows that the displacement or shaking amplitude of the auxiliary student terminal within a preset time period is lower than the preset static threshold, it indicates that the device is placed or held stably and there is no interference from moving around; and when the classroom situation analysis module analyzes the first audio data and confirms that the environmental signal-to-noise ratio is higher than the preset clarity threshold and there is no additional background noise interference, it is determined that the student corresponding to the auxiliary student terminal or the assisting student terminal is in a focused state, that is, an immersive listening state.

[0068] When the assisted student is determined to be in a focused state, the classroom situation analysis module also dynamically expands the matching range of the timestamps in the first key timestamp sequence generated in that state. When comparing the first key timestamp sequence with the second key timestamp sequence marked by the target student, the fixed 3-second matching standard from Implementation Example 1 is no longer used; instead, an asymmetric expansion of the matching interval is adopted. Specifically, the classroom situation analysis module extends the effective matching range of the timestamp forward, for example, adjusting it to a range of 5 seconds before and 3 seconds after the timestamp. If the second key timestamp marked by the target student falls within this expanded range, it is considered the same mark and not treated as omitted content; if it still does not fall within, it is considered an omitted time period, and the content is extracted to generate multi-dimensional analytical notes.

[0069] By introducing joint analysis of motion sensor data and environmental sound data, it is possible to identify scenarios where students are highly focused, thus solving the problem of non-aligned physical timestamps at both ends caused by the delayed response of the assisting or helping students, thereby significantly reducing the false negative rate.

[0070] The above are merely embodiments of the present invention. The invention is not limited to the fields covered by these embodiments. Commonly known structures and characteristics in the solutions are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are able to access all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A nursing intelligent teaching assistance system, comprising a cloud server, several teacher terminals, and several student terminals; The cloud server includes a knowledge base building module, which is used to obtain teaching documents for adult nursing and build structured course knowledge graph data based on the teaching documents; The model building module is used to obtain a general large language model, use course knowledge graph data to perform domain-adaptive training on the general model, and output a custom model that integrates adult care knowledge. The teaching assistant module is used to accept lesson preparation requests from teachers, input a unique model, and generate lesson preparation information. The learning assistance module is used to receive assistance requests from students, and based on the content input by the students, it calls a dedicated model to generate supplementary teaching content.

2. The nursing intelligent teaching assistance system according to claim 1, characterized in that: The cloud server also includes a teaching management module, which is used to obtain class information, bind several student accounts that have joined the same teaching class to the teacher account that is responsible for the class, and grant the teacher account management permissions to the bound student accounts. The learning support module is also used to collect interaction data between students and the dedicated model, generate student learning analysis reports, and push the learning analysis reports to the teachers of the corresponding classes.

3. The nursing intelligent teaching assistance system according to claim 2, characterized in that: The cloud server also includes a classroom situation analysis module, which is used to respond to a wake-up request sent by any student terminal, send an activation command to the corresponding teacher terminal and the target student terminal, and mark the student terminal as an auxiliary student terminal; Among them, the auxiliary student terminal is another student terminal that is in the same teaching class as the target student terminal; the wake-up request contains information about the target student terminal. The teacher's end is used to collect the first audio data and upload it to the classroom situation analysis module after receiving the activation command; The target student terminal is used to collect second audio data and upload it to the classroom situation analysis module after receiving the activation command; The classroom situation analysis module is also used to extract decibels from the first audio data, perform emotion intensity analysis based on acoustic features, and process speech to text to obtain the teacher's text, and to extract decibels from the second audio data and process speech to text to obtain the student's text. The classroom situation analysis module is also used to calculate the decibel difference between the first audio data and the second audio data to determine the distance of the sound source, and to calculate the text similarity between the teacher's text and the student's received text to determine the transmission clarity.

4. The nursing intelligent teaching assistance system according to claim 3, characterized in that: The classroom situation analysis module is also used to calculate the stress index of the target student based on the analysis results of sound source distance, transmission clarity and emotional intensity. Among them, the stress index is negatively correlated with the distance from the sound source, negatively correlated with the transmission clarity, and positively correlated with the intensity of emotions; if the stress index exceeds the preset stress threshold, the classroom situation analysis module is also used to send an auxiliary marker request to the auxiliary student terminal and receive the first key timestamp sequence of feedback from the auxiliary student terminal. After the teacher stops collecting the first audio data, the classroom situation analysis module encapsulates the teacher's text into a review note data package and sends it to the target student's end, and obtains the second key timestamp sequence marked in the review note data package by the target student's end; The classroom situation analysis module is also used to compare the second key timestamp sequence with the first key timestamp sequence. If a missing time period is identified that exists in the first key timestamp sequence but is missing in the second key timestamp sequence, the content corresponding to the missing time period in the teacher's text is extracted, and a dedicated model is called in combination with preset prompt words to generate multi-dimensional analytical notes for the missing time period.

5. The nursing intelligent teaching assistance system according to claim 4, characterized in that: The classroom situation analysis module is also used to send multi-dimensional analysis notes for the missed time period to the auxiliary student terminal, receive selection information for any dimension from the auxiliary student terminal, mark the dimension, and send the marked multi-dimensional analysis notes for the missed time period to the target student terminal.

6. The nursing intelligent teaching assistance system according to claim 5, characterized in that: The target student terminal is also used to receive initial cognitive dimension selection information for the multi-dimensional analysis notes for the missed time period, mark the dimension, and send the multi-dimensional analysis notes, which are compared by both the auxiliary student terminal and the target student terminal, to the teacher terminal.

7. The nursing intelligent teaching assistance system according to claim 6, characterized in that: The classroom situation analysis module is also used to send the teacher's text to the teacher's end. After receiving the approval information from the teacher's end, the teacher's text is packaged into a review notes data package and sent to the target student's end.

8. The nursing intelligent teaching assistance system according to claim 7, characterized in that: The classroom situation analysis module is also used to retrieve the usage records of the target student's terminal and, in conjunction with the analysis of the real-time teacher text converted from the first audio data, determine the teacher's current explanation content and judge whether the target student has previewed the explanation content in advance. If the judgment result is that the target student has not previewed, the classroom situation analysis module still sends an auxiliary tag request to the auxiliary student terminal. If the judgment result indicates that the target student has already previewed the material, the classroom situation analysis module is also used to filter out other students in the same teaching class who have not previewed the material, send auxiliary tag requests to these students, and mark the students who accept the request as assisting students.

9. The nursing intelligent teaching assistance system according to claim 8, characterized in that: The student terminal also includes a motion sensor for real-time acquisition of device posture data. The classroom situation analysis module is also used to receive device posture data uploaded by the auxiliary student terminal or the assisting student terminal, and combine it with the synchronously collected first audio data to determine the focus status; When a student is determined to be in a focused state, the classroom situation analysis module is also used to perform dynamic expansion of the matching range of timestamps in the first key timestamp sequence generated in that state.

10. A nursing intelligent teaching aid method, characterized in that, Use the system according to any one of claims 1-9.

Images