A nursing training effect evaluation system and method based on multi-source information fusion

Abstract

This invention relates to the field of medical training technology and discloses a nursing training effectiveness evaluation system based on multi-source information fusion. It includes a question bank module and a personalized terminal module for trainees. Trainees learn various knowledge points according to recommended content through learning units. The system records their learning information and historical test information, which are analyzed by a theoretical test question bank generation unit. Combined with preset strategies, a personalized test question bank is generated to assess the trainees' mastery of each knowledge point. A personalized learning tutoring unit further determines the self-study effectiveness level of each knowledge point, providing targeted guidance to instructors. The system integrates the forgetting rate constant, effective learning index, difficulty level determined by the number of learned knowledge points linked to the questions, and self-study effectiveness judgment index for each knowledge point to optimize test paper generation and tutoring prompts, reducing repeated testing of mastered knowledge points, lowering the probability of missing weak knowledge points, and improving the cognitive load of the test.

Description

Technical Field

[0001] This invention relates to the field of medical training technology, specifically to a nursing training effectiveness evaluation system and method based on multi-source information fusion. Background Technology

[0002] With the steady development of my country's social economy and the continuous improvement of medical conditions, the aging process of the population is accelerating, and the elderly population is growing rapidly. At the same time, influenced by multiple factors such as lifestyle and environment, the incidence of chronic diseases is also showing a year-on-year upward trend. This series of changes has led to a sharp increase in the demand for nursing services, exhibiting a significant explosive growth, and placing higher demands on the quantity, quality, and professionalism of nursing services. Therefore, effective medical and nursing training for nursing staff is particularly urgent and of great significance.

[0003] In the prior art, a medical training system based on a large language model, disclosed in publication number "CN116824933A", uses an interactive Q&A module to activate the medical training large language model through a model invocation module, thereby achieving effective communication with trainees. This module can provide accurate answers to various medical questions raised by trainees. At the same time, it comprehensively monitors the trainees' learning progress and performance, and generates personalized evaluation reports and training plans based on the monitoring results. The plans and reports are then sent to the trainees' accounts, creating a customized training plan for each trainee.

[0004] However, the aforementioned technologies still have significant shortcomings. In the field of nursing training, existing technologies are insufficient to meet the growing demand for high-quality training. The medical training systems mentioned above do not record the learning process of trainees in a detailed and comprehensive manner, making it difficult to accurately track the learning progress of trainees on each knowledge point. As a result, when instructors give lectures, they usually design courses based on their own experience, rather than closely aligning with the trainees' actual level of knowledge mastery and learning weaknesses. Due to a lack of precise understanding of students' learning progress on each knowledge point, the course content designed by instructors may deviate from students' actual needs. This results in students having to repeat some content they have already mastered, wasting valuable learning time and teaching resources. Meanwhile, some key knowledge points that students have not fully understood may not receive in-depth explanations because the instructors have not noticed. This not only reduces students' learning efficiency, affects their learning enthusiasm and learning outcomes, but also makes it difficult for nursing training to reach high-quality standards, failing to effectively meet society's growing demand for professional nursing personnel, and thus hindering the overall development and progress of the nursing industry. Furthermore, existing technologies do not quantitatively model the forgetting dynamics of knowledge points over time, nor do they nonlinearly couple the cumulative effective learning time with the forgetting rate. At the same time, existing question bank generation methods usually do not link the difficulty of questions with the number of learned knowledge points covered by the questions, making it difficult to avoid repeatedly testing mastered knowledge points or omitting truly weak knowledge points during test paper compilation. It also makes it difficult to accurately identify students who have invested a lot in learning but whose test performance is consistently low and who need active tutoring. Summary of the Invention

[0005] The purpose of this invention is to provide a nursing training effectiveness evaluation system and method based on multi-source information fusion, and to solve the following technical problems: 1. How to quantify the non-linear coupling relationship between the forgetting dynamics of knowledge points and learning input, and avoid the following technical defects in the generated test question bank: either repeatedly testing knowledge points that students have already mastered, resulting in a waste of testing resources and an unreasonable allocation of cognitive load; or omitting real knowledge weaknesses, making the assessment results unable to truly reflect the students' knowledge mastery status; 2. How to automatically and accurately identify students who have high learning engagement but consistently low test performance and require proactive tutoring.

[0006] The objective of this invention can be achieved through the following technical solutions: A nursing training effectiveness evaluation system based on multi-source information fusion, the system including a question bank module and a personalized terminal module for trainees: The question bank module is used to store test questions and feature information data of each test question; The personalized terminal module for students includes: Learning units are used to learn knowledge information data for each knowledge point according to recommended content; The student learning record storage unit is used to store the learning information data and historical test information data of each student for each knowledge point. The theoretical test question bank generation unit is connected to the logged-in student learning record storage unit and question bank module. It is used to analyze the learning information data and historical test information data of each logged-in student for each knowledge point, and generate the theoretical test question bank in combination with the theoretical test question generation strategy. The personalized learning tutoring unit is used to analyze the learning information data and historical test information data of the logged-in student to obtain the student's self-study effectiveness level for each knowledge point; the tutor provides targeted tutoring based on the student's self-study effectiveness level for each knowledge point. The theoretical test question bank generation unit and personalized learning tutoring unit are configured to perform the following core processes: Calculate the forgetting rate constant and effective learning index based on the cumulative effective learning time, preset effective learning time, learning progress, the time from the last learning time to the current time, historical test accuracy, and the time from the test time to the current time; classify effective learning levels according to the effective learning index; correlate the difficulty of the questions with the number of learned knowledge points associated with each question, where easy questions are associated with one learned knowledge point, moderate questions with two learned knowledge points, and difficult questions with three or more learned knowledge points; and determine whether active tutoring is needed based on the self-study effect judgment index.

[0007] As a further aspect of the present invention, the learning information data includes the cumulative effective learning time and the learning progress of knowledge points.

[0008] As a further aspect of this invention: the working process of the theoretical test question bank generation unit is as follows: S10: Obtain the student's cumulative effective learning time for each knowledge point, learning progress, and historical test information data by logging into the student's learning record storage unit; S20: Based on the cumulative effective learning time and learning progress of each knowledge point of the student, as well as the test information data within the past preset time period, an effective learning index for each knowledge point is obtained. S30: Analyze the effective learning index of each knowledge point to obtain the effective learning level of each knowledge point; S40: Generate this theoretical test question bank based on the effective learning level of each knowledge point and the theoretical test question generation strategy.

[0009] As a further aspect of the present invention: through formula one: ; Calculate the effective learning index P of the nth knowledge point. n ; Where, ω nLet n be the forgetting rate constant for the nth knowledge point; t ns T represents the duration from the last time the nth knowledge point was learned to the current time; ne T represents the cumulative effective learning time for the nth knowledge point. ne0 F is the preset effective learning time for the nth knowledge point; n The learning progress for the nth knowledge point; K represents the number of theoretical tests within a preset time period, k∈K; C nk Let n be the accuracy rate of the nth knowledge point in the kth test; t k U represents the duration from the time of the k-th test to the current time. n θ represents the number of historical tests for the nth knowledge point; k γ1 is the weight coefficient for the k-th test; γ2 is the first weight coefficient; γ3 is the second weight coefficient.

[0010] As a further aspect of the present invention: through formula two: ; Calculate the forgetting rate constant ω of the nth knowledge point. n ; Where ω0 is the preset basic forgetting rate constant; ρ n T represents the difficulty coefficient of the nth knowledge point; ne T represents the cumulative effective learning time for the nth knowledge point. ne0 μ1 is the preset effective learning time for the nth knowledge point; μ1 is the first adjustment coefficient.

[0011] As a further aspect of the present invention: through formula three: ; Calculate the accuracy C of the nth knowledge point in the kth test. nk ; Where M represents the number of difficulty levels of the questions, m∈M; E knm0 Let E be the number of questions at the m-th difficulty level in the k-th test that involve the n-th knowledge point. knm0 When the value is greater than 0, this difficulty level item is included in the calculation; ρ represents the number of correct answers for the logged-in student in the m-th difficulty level question type that involves the n-th knowledge point in the k-th test; m is the weighting coefficient for the m-th difficulty level.

[0012] As a further aspect of the present invention: in step S40, the theoretical test question generation strategy includes: The effective learning level of knowledge points is 25%-35%; the difficulty level of all knowledge points is set to easy. The effective learning level of knowledge points is 50%-60%; the difficulty level of each knowledge point is easy, moderate and hard, and generated according to a preset ratio; The effective learning level of knowledge points is 5%-25%; the difficulty level of all knowledge points is set to difficult. Furthermore, the sum of the knowledge points for each effective learning level is 100%.

[0013] As a further aspect of the present invention, the difficulty levels include: easy, normal, and difficult.

[0014] As a further aspect of the present invention: the theoretical questions of the difficulty level of "simple" have only one learned knowledge point; The theoretical questions of general difficulty level contain only two previously learned knowledge points; The theoretical questions, classified as difficult, involve three or more previously learned knowledge points.

[0015] A method for evaluating the effectiveness of nursing training based on multi-source information fusion, the method comprising the following steps: S1: First, the test questions with characteristic information data are stored through the question bank module; S2: Students who log in to the personalized terminal module learn knowledge information data of each knowledge point through the learning unit according to the recommended content; S3: The learning information data and historical test information data of each knowledge point of each logged-in student are stored in the student learning record storage unit; S4: The theoretical test question bank generation unit analyzes the learning information data and historical test information data of each logged-in student for each knowledge point, and generates the theoretical test question bank in combination with the theoretical test question generation strategy. S5: The personalized learning tutoring unit analyzes the learning information data and historical test information data of the logged-in student to obtain the self-study effect level of the logged-in student for each knowledge point; the tutor provides targeted tutoring based on the self-study effect level of the logged-in student for each knowledge point.

[0016] The beneficial effects of this invention are: (1) This invention stores test questions and feature information data of each test question through the question bank module; it learns knowledge information data of each knowledge point according to recommended content through the learning unit of the student personalized terminal module; it provides students with personalized learning paths and content, which helps to improve students' learning interest and enthusiasm, and can meet the learning needs of students with different learning levels; it stores learning information data and historical test information data of each knowledge point of each logged-in student through the student learning record storage unit; it provides rich and accurate data support for a comprehensive understanding of students' learning status; it analyzes the learning information data and historical test information data of each knowledge point of each logged-in student through the current theoretical test question bank generation unit, and generates the current theoretical test question bank in combination with the theoretical test question generation strategy; it can accurately grasp the students' mastery of each knowledge point and generate a test question bank that is highly consistent with the students' actual level; this not only This approach makes testing more targeted and effective, accurately reflecting students' learning outcomes and allowing them to experience a balance between challenge and achievement, further enhancing their learning motivation. Personalized learning tutoring units analyze each student's learning information and historical test data to determine their self-study effectiveness level for each knowledge point. Tutors then provide targeted guidance based on this level. For the entire teaching system, this technology allows tutors to better adjust teaching strategies and methods and optimize resource allocation by accurately grasping each student's learning situation, making teaching activities more efficient and orderly. Simultaneously, this personalized teaching model helps cultivate students' self-learning abilities and innovative thinking, laying a solid foundation for their future development and promoting education towards a more intelligent, personalized, and efficient direction. Attached Figure Description

[0017] The invention will now be further described with reference to the accompanying drawings.

[0018] Figure 1 This is a system module framework diagram of one embodiment of the present invention; Figure 2 This is a flowchart of a method according to an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see Figure 1As shown, in one embodiment, a nursing training effectiveness evaluation system based on multi-source information fusion is provided. The system includes a question bank module and a personalized terminal module for trainees. The question bank module is used to store test questions and feature information data of each test question; Specifically, the feature information data includes the knowledge points involved in each test question; The personalized terminal module for students includes: Learning units are used to learn knowledge information data for each knowledge point according to recommended content; Specifically, knowledge information data can be videos, papers, audio, or images, etc. Recommended content can include learning content and review content. After each test, knowledge points with an accuracy rate lower than the preset accuracy rate will be set as review content for this round. Learning content consists of knowledge points that are unlocked sequentially according to the preset learning order. If the learning progress of the previous knowledge point has not reached 80%, the next knowledge point will not be unlocked, and the knowledge information data of the previous knowledge point will be set as learning content. When the learning progress of the previous knowledge point reaches 80%, the next knowledge point will be unlocked, and the unlearned knowledge information data of the previous knowledge point and the knowledge information data of the next knowledge point will be set as learning content. The student learning record storage unit is used to store the learning information data and historical test information data of each student for each knowledge point. The theoretical test question bank generation unit is connected to the logged-in student learning record storage unit and question bank module. It is used to analyze the learning information data and historical test information data of each logged-in student for each knowledge point, and generate the theoretical test question bank in combination with the theoretical test question generation strategy. The personalized learning tutoring unit is used to analyze the learning information data and historical test information data of the logged-in student to obtain the student's self-study effectiveness level for each knowledge point; the tutor provides targeted tutoring based on the student's self-study effectiveness level for each knowledge point. Through the above technical solution, this embodiment stores test questions and their feature information data through a question bank module; it teaches students the knowledge information data of each knowledge point according to recommended content through the learning unit of the student personalized terminal module; it provides students with personalized learning paths and content, which helps to improve students' learning interest and enthusiasm and can meet the learning needs of students at different learning levels; it stores the learning information data and historical test information data of each logged-in student for each knowledge point through the student learning record storage unit; it provides rich and accurate data support for a comprehensive understanding of students' learning status; and it analyzes the learning information data and historical test information data of each logged-in student for each knowledge point through the theoretical test question bank generation unit, and generates the theoretical test question bank in combination with the theoretical test question generation strategy; it can accurately grasp the students' mastery of each knowledge point and generate a test question bank that is highly consistent with the students' actual level. This not only makes testing more targeted and effective, accurately reflecting students' learning outcomes, but also allows students to experience a balance between challenge and achievement, further enhancing their learning motivation. Personalized learning tutoring units analyze the student's learning information and historical test data to obtain the student's self-study effectiveness level for each knowledge point. Tutors then provide targeted guidance based on the student's self-study effectiveness level for each knowledge point. For the entire teaching system, this technology, by accurately grasping each student's learning situation, allows tutors to better adjust teaching strategies and methods, optimize the allocation of teaching resources, and make teaching activities more efficient and orderly. At the same time, this personalized teaching model is also conducive to cultivating students' self-learning ability and innovative thinking, laying a solid foundation for their future development and promoting education towards a more intelligent, personalized, and efficient direction.

[0021] As one embodiment of the present invention, the learning information data includes cumulative effective learning time and knowledge point learning progress: knowledge information data of each knowledge point is learned through the learning unit recommended content in the student's personalized terminal module; Specifically, the cumulative effective learning time is the total time when the single learning time for a knowledge point is greater than the preset effective learning time. The preset effective learning time can be 5 minutes or 10 minutes, etc., which can be set by the staff themselves and is not restricted here. The knowledge point learning progress is the percentage of the learning materials for that knowledge point that have been learned out of the total amount of learning materials. As one embodiment of the present invention, the working process of the theoretical test question bank generation unit is as follows: S10: Obtain the student's cumulative effective learning time for each knowledge point, learning progress, and historical test information data by logging into the student's learning record storage unit; S20: Based on the cumulative effective learning time and learning progress of each knowledge point of the student, as well as the test information data within the past preset time period, an effective learning index for each knowledge point is obtained. S30: Analyze the effective learning index of each knowledge point to obtain the effective learning level of each knowledge point; S40: Generate the theoretical test question bank based on the effective learning level of each knowledge point and the theoretical test question generation strategy; Through the above technical solution, this embodiment first obtains the student's cumulative effective learning time, learning progress, and historical test information data for each knowledge point by logging into the student's learning record storage unit; then, it analyzes the student's cumulative effective learning time, learning progress, and test information data within a preset time period to obtain the effective learning index for each knowledge point; this objectively reflects the student's mastery of each knowledge point; finally, it analyzes the effective learning index for each knowledge point to obtain the effective learning level for each knowledge point; and finally, it generates the theoretical test question bank based on the effective learning level of each knowledge point and the theoretical test question generation strategy; this ensures that the generated test question bank highly matches the student's actual level, allowing the student to demonstrate their true abilities in the test, and enabling the instructor to clearly understand the student's learning status, thereby adjusting teaching strategies, optimizing resource allocation, improving teaching effectiveness, and powerfully promoting education towards intelligence, personalization, and efficiency.

[0022] As one embodiment of the present invention, using Formula 1: ; Calculate the effective learning index P of the nth knowledge point. n ; Where, ω n Let n be the forgetting rate constant for the nth knowledge point; t ns This represents the duration from the last time the nth knowledge point was studied to the current time. ne This represents the cumulative effective learning time for the nth knowledge point. ne0 F is the preset effective learning time for the nth knowledge point; n The learning progress for the nth knowledge point; K represents the number of theoretical tests within a preset time period, k∈K; C nk Let n be the accuracy rate of the nth knowledge point in the kth test; t k U represents the duration from the time of the k-th test to the current time. n θ represents the number of historical tests for the nth knowledge point; k γ1 represents the weighting coefficient for the k-th test; γ2 represents the first weighting coefficient; γ3 represents the second weighting coefficient. Formula 1 Explanation: When U n =0 and F n When the value is ≥0.8, it indicates that the learning progress of the nth knowledge point has exceeded 80%, and no test has been conducted; therefore, the content of the nth knowledge point has not been effectively tested, and thus the effective learning index P of the nth knowledge point is low. n =0; when U n >0 and F n A value ≥ 0.8 indicates that the nth knowledge point has been tested more than once and can pass. Obtain the effective learning index of the nth knowledge point; the effective learning index reflects the degree of effective learning of the nth knowledge point. It represents the first forgetting decay coefficient of the nth knowledge point based on the Ebbinghaus forgetting curve, which is the time elapsed between the last time the nth knowledge point was learned and the current time. It reflects the degree of mastery of the nth knowledge point between the last time the nth knowledge point was learned and the current time. This indicates the completion rate of effective learning time; This represents the learning outcome after considering forgetting; the accuracy rate of the nth knowledge point in the kth test. This reflects the level of mastery of the nth knowledge point in the kth test. This represents the second forgetting decay coefficient of the test content based on the time from the k-th test to the current time, according to the Ebbinghaus forgetting curve. This represents the level of mastery of the nth knowledge point in the kth test up to the current time; This represents the overall mastery level of the nth knowledge point across all tests conducted within a preset time period past the current time; when F... n If the value is less than 0.8, it means that the learning progress of the nth knowledge point is less than 80%, and the content of that knowledge point will not be tested. It should be noted that the weight coefficient θ for the k-th test... k satisfy ,and The test that is furthest from the current time is the first test; the test that is closest to the current time is the kth test. It should be noted that the first weighting coefficient γ1 and the second weighting coefficient γ2 are used to represent the influence of the degree of mastery of the nth knowledge point between the last time of learning and the current time on the effective learning assessment, and the degree of comprehensive mastery of the nth knowledge point in all tests within the past preset time period. The degree of comprehensive mastery of the nth knowledge point in all tests within the past preset time period more intuitively reflects the degree of effective understanding of the nth knowledge point. Therefore, γ2 > γ1 and γ1 + γ2 = 1. The weight coefficient θ for the k-th testk The values ​​of the first weighting coefficient γ1 and the second weighting coefficient γ2 are not arbitrarily set, but are derived based on cognitive patterns in the field of nursing education and empirical data fitting. The specific determination process is as follows: The inventors selected 200 nursing students as experimental subjects and conducted a follow-up experiment in a real nursing training environment for six months. During the experiment, the initial assumption was that γ1+γ2=1. Given the conditions, a grid search method is used to traverse γ1 in the interval [0.1, 0.9] with a step size of 0.05, while simultaneously performing a search on each θ. k Constraint optimization is performed in the interval [0, 1]; the "accuracy rate of the question bank in detecting students' knowledge weaknesses" (defined as the consistency ratio between the weak knowledge points identified by the test and the actual weak knowledge points manually assessed by the tutor) is used as the optimization objective function; When K=3, the optimal weight configuration is θ1=0.20, θ2=0.35, θ3=0.45. Under this configuration, the weight of the most recent test is the highest, which is in line with the cognitive rule that "the most recent test can better reflect the current mastery status". When K=2, the optimal weight configuration is θ1=0.35 and θ2=0.65.

[0023] The above value scheme was statistically significant (p < 0.05) after cross-validation, and can effectively balance the contribution of "recent learning effect" and "historical stable performance" to the assessment of current mastery.

[0024] Through the above technical solution, this embodiment introduces a forgetting decay coefficient based on the Ebbinghaus forgetting curve, combined with the forgetting rate constant, learning duration, learning progress, and historical test accuracy, to accurately reflect the learner's effective learning of knowledge points and the impact on their mastery of those points. Simultaneously, by setting weighting coefficients based on the number of tests and their proximity to the current time, the influence of historical test results on the current assessment is made more realistic. Furthermore, it differentiates between different learning progress levels, avoiding unnecessary testing. This formula provides an accurate basis for the nursing training effectiveness evaluation system, helping to generate a more reasonable test question bank, achieve personalized learning guidance, and improve training quality.

[0025] As one embodiment of the present invention, formula two is used: ; Calculate the forgetting rate constant ω of the nth knowledge point. n ; Where ω0 is the preset basic forgetting rate constant; ρ n T represents the difficulty coefficient of the nth knowledge point; ne T represents the cumulative effective learning time for the nth knowledge point. ne0μ1 is the preset effective learning time for the nth knowledge point; μ1 is the first adjustment coefficient. Explanation of Formula 2: The ratio of the cumulative effective learning time for the nth knowledge point to the preset effective learning time for the nth knowledge point; when When the value is greater than 0, it means that the cumulative effective learning time for the nth knowledge point exceeds the preset effective learning time for the nth knowledge point. This indicates that the student has repeatedly watched and studied the nth knowledge point, resulting in a deeper impression and a slower rate of forgetting. Therefore, the forgetting rate constant ω for the nth knowledge point is... n Get smaller; when When the value is less than 0, it indicates that the cumulative effective learning time for the nth knowledge point has not exceeded the preset effective learning time for the nth knowledge point. This means that the student has not systematically and completely learned the nth knowledge point, and therefore will forget it faster. Therefore, the forgetting rate constant ω for the nth knowledge point is... n The larger the difficulty coefficient of the nth knowledge point, the more difficult it is to understand; therefore, the forgetting rate constant ω of the nth knowledge point... n The larger the value, the smaller the difficulty coefficient of the nth knowledge point, indicating that the nth knowledge point is more difficult to understand. Therefore, the forgetting rate constant ω of the nth knowledge point... n The smaller; It should be noted that the preset effective learning time for each knowledge point is based on the tutor's judgment of knowledge information data and experience for each knowledge point, and is a preset value with a range of [0.5h, 3h], which is not limited here; It should be noted that the values ​​of the preset basic forgetting rate constant ω0 and the first adjustment coefficient μ1 are determined based on the following: The determination of ω0 is based on the classic research findings of the Ebbinghaus forgetting curve. Experimental psychology research shows that the 24-hour forgetting rate of rote memorized materials without review is approximately 66%, corresponding to a forgetting rate constant of approximately 0.01 (in hours). Considering that nursing training content mainly consists of procedural and conceptual knowledge, its forgetting rate is usually lower than that of rote memorized materials. The inventors conducted a small-sample preliminary experiment (N=30) to track and test nursing knowledge points, and fitted the optimal value range of ω0 to [0.005, 0.008]. In this embodiment, ω0 = 0.0065 is taken as the median of this interval. μ1 was determined using a data-driven approach: the "learning input-forgetting inhibition effect function" was defined as... Among them, by performing nonlinear least squares fitting on the individual learning trajectory data of 200 students, the optimal estimate of μ1 was obtained as 1.2 (95% confidence interval: [1.05, 1.38]). This value indicates that when the cumulative effective learning time exceeds twice the preset effective learning time ( The forgetting rate constant decays to its original value. ≈0.30 times, meaning that learning input has a significant inhibitory effect on forgetting, which is consistent with the basic principle of "overlearning" in cognitive psychology that strengthens memory; Through the above technical solution, this embodiment can more accurately reflect the forgetting status of trainees on different knowledge points by preset basic forgetting rate constant, knowledge point difficulty coefficient, and the ratio of cumulative effective learning time to preset effective learning time deviation. This helps to more scientifically assess the trainees' mastery of each knowledge point, providing a reliable basis for subsequent theoretical test question bank generation, personalized learning tutoring, etc., thereby improving the accuracy and effectiveness of the nursing training effect evaluation system.

[0026] As one embodiment of the present invention, formula three is used: ; Calculate the accuracy C of the nth knowledge point in the kth test. nk ; Where M represents the number of difficulty levels of the questions, m∈M; E knm0 Let E be the number of questions at the m-th difficulty level in the k-th test that involve the n-th knowledge point. knm0 When the value is greater than 0, this difficulty level item is included in the calculation; ρ represents the number of correct answers for the logged-in student in the m-th difficulty level question type that involves the n-th knowledge point in the k-th test; m This represents the weighting coefficient for the m-th difficulty level. Formula 3 explains, The accuracy rate at the m-th difficulty level in the k-th test; the weighting coefficient of the m-th difficulty level. This represents the influence of the accuracy rate of the question type involving the nth knowledge point in the mth difficulty level in the kth test on the accuracy rate of the nth knowledge point in the kth test; the weight coefficient of the mth difficulty level is a preset value. ρ1 is the lowest weighting coefficient for the difficulty level; ρ M The highest weighting coefficient represents the difficulty level. The higher the difficulty level of a question, the more flexible and profound the understanding of the learned knowledge points is required. Therefore, the higher the difficulty level, the higher the accuracy rate, and the better it reflects the comprehensive mastery of the nth knowledge point.

[0027] Difficulty level weighting coefficient The Delphi method, combined with empirical calibration, was used to determine the method. The specific steps are as follows: First, 15 experts with more than 5 years of nursing teaching experience were invited to independently score the importance of the three difficulty levels of "easy", "moderate" and "difficult" (1-5 points). After three rounds of feedback and iteration, the experts' opinions tended to converge, and the basic weight ratio was ρ1 (easy): ρ2 (moderate): ρ3 (difficult) = 0.6:1.0:1.4.

[0028] Subsequently, the data was validated using test responses from 500 real students. Using the "correlation between the difficulty level weight coefficient and the assessment of knowledge point mastery" as an indicator, the weight ratios were fine-tuned using the gradient descent method. The optimization results showed that when ρ1=0.58, ρ2=1.00, and ρ3=1.42, the assessment results had the highest predictive correlation with students' subsequent performance (such as the accuracy rate of subsequent tests) (Pearson correlation coefficient r=0.73).

[0029] In this embodiment, ρ1=0.6, ρ2=1.0, and ρ3=1.4 are chosen. These values ​​are consistent with expert experience and have been verified by empirical data. They can reasonably reflect the higher contribution of correct answers to high-difficulty questions to the assessment of knowledge mastery.

[0030] Through the above technical solution, this embodiment comprehensively reflects the actual mastery of knowledge points at each difficulty level by considering the number of knowledge points involved in question types of different difficulty levels and the number of correct answers by the trainees; and by presetting weight coefficients for different difficulty levels, it reasonably reflects the influence of each difficulty level on the overall knowledge point mastery evaluation, making the accuracy calculation more in line with the actual learning situation. This helps to more scientifically and accurately assess the trainees' mastery level of knowledge points, providing a reliable basis for the subsequent generation of personalized test question banks and the provision of targeted learning guidance, thereby improving the overall effectiveness of the nursing training effect evaluation system.

[0031] In one embodiment of the present invention, in step S30, the effective learning level of the nth knowledge point is obtained as follows: The effective learning index P of the nth knowledge point n Compare with the preset comparison value R1; When P n When = 0, the effective learning level of the nth knowledge point is level one; When 0 < P n When R1 ≤ R1, the effective learning level of the nth knowledge point is level two; When R1 < P n At that time, the effective learning level of the nth knowledge point is level three; The preset comparison value R1 is determined based on the effective learning index P from the experimental data of the aforementioned 200 students. n The distribution characteristics were analyzed using the tertile method; statistical results showed that P n The distribution of exhibits a right-skewed characteristic, wherein: The first tertiary (33.3 percentile) corresponds to P n The value is 0.32; The Pn value corresponding to the second tertiary (66.7th percentile) is 0.68.

[0032] To maintain a relatively balanced number of students across the three learning levels (Level 1, Level 2, and Level 3) and to avoid statistical instability caused by overly detailed divisions, this embodiment selects R1=0.5. This value falls between two tertiles, allowing approximately 60% of students to be classified as Level 2 (intermediate mastery), and approximately 20% each as Level 1 (low mastery) and Level 3 (high mastery), respectively. This aligns with the "larger middle and smaller ends" distribution pattern observed in teaching practice.

[0033] As one embodiment of the present invention, the difficulty levels include: easy, normal, and hard; The difficulty level is defined as simple theoretical questions with only one previously learned knowledge point; The theoretical questions of general difficulty level contain only two previously learned knowledge points; The theoretical questions with a difficulty level of "hard" contain three or more previously learned knowledge points. As one embodiment of the present invention, in step S40, the theoretical test question generation strategy may include: The effective learning level of knowledge points is 25%-35%; the difficulty level of all knowledge points is set to easy. The effective learning level of knowledge points is 50%-60%; the difficulty level of each knowledge point is easy, moderate and hard, and generated according to a preset ratio; Specifically, the preset ratio can be set by the staff themselves, such as 1:2:1, and there are no restrictions here; The effective learning level of knowledge points is 5%-25%; the difficulty level of all knowledge points is set to difficult. Furthermore, the sum of the knowledge points at each effective learning level is 100%; When P n When 0 < P, it indicates that the learning progress for this knowledge point exceeds 80% but has not been tested, thus failing to effectively assess learning progress. In this case, the test question bank should appropriately include basic and comprehensive simple questions to help learners gradually build confidence and a solid foundation. nWhen R1 is ≤, Level 2 effective learning indicates that the learner has some learning but a general level of mastery. For this type of knowledge point, the number of questions should be increased, with the difficulty level mainly at a moderate level. The questions should cover common test points and common mistakes of this knowledge point, and comprehensively test the learner's understanding and application of knowledge points through different question types, helping the learner consolidate knowledge and identify their weaknesses. For knowledge points with an effective learning level of Level 3, the learner has a relatively high level of mastery. The number of questions for this type of knowledge point in the test question bank can be relatively small, and the difficulty level should be set to difficult. The difficulty of the questions should be appropriately increased, and comprehensive and extended questions should be added to test the learner's in-depth understanding and flexible application of knowledge points, tap the learner's potential, and at the same time allow the learner to further improve themselves through challenges. This can further challenge learners and improve their ability to apply knowledge and solve complex problems.

[0034] The percentage range of questions for each effective learning level (25%-35%, 50%-60%, 5%-25%) and the preset ratio are determined as easy: moderate: difficult = 1:2:1, based on cognitive load theory and teaching experiments in the field of nursing training.

[0035] The experimental design was as follows: 200 participants were randomly divided into 4 groups, each with a different question ratio configuration, and a comparative experiment was conducted for 3 months (tested once a month). The evaluation indicators were "the participants' self-assessment score of cognitive load after the test" (out of 10 points, with lower scores indicating more reasonable cognitive load) and "the recall rate of knowledge gap detection".

[0036] The experimental results are shown in the table below: Experimental results show that: (1) When the A group configuration is adopted (25% of the first-level knowledge points, 50% of the second-level knowledge points, and 25% of the third-level knowledge points, and the second-level knowledge points are generated in the ratio of easy: normal: difficult = 1:2:1), the average cognitive load of the students is 5.6 points (out of 10), the recall rate of the knowledge weakness points reaches 91%, the average test duration is 27 minutes, the student satisfaction is 93%, and the overall performance is the best. (2) When the proportion of Level 3 knowledge points decreased from 25% to 15% (Group B), the recall rate of weak points dropped to 84%, indicating that the lack of difficult questions will significantly affect the detection effect of high-level knowledge points. (3) When the proportion of first-level knowledge points increased from 25% to 35% (Group D), the cognitive load increased to 6.8 and the recall rate of weak points decreased to 79%, indicating that too many simple questions not only failed to reduce the cognitive load, but also reduced the detection accuracy by squeezing out the effective question space. (4) When the proportion of secondary knowledge points increased from 50% to 60% (Group B), the test duration was extended to 29 minutes, which is close to the upper limit of the reasonable range, and the recall rate decreased, indicating that an excessive increase in medium-difficulty questions would lead to a decrease in test efficiency.

[0037] Therefore, the percentage range (Level 1 25%-35%, Level 2 50%-60%, Level 3 5%-25%) and preset ratio (1:2:1) adopted in this application can achieve the optimal balance between cognitive load and detection accuracy. Among them, Level 1 25%, Level 2 50%, and Level 3 25% is the optimal configuration, which has significant technical effects.

[0038] As one embodiment of the present invention, the process for determining the self-learning effectiveness level of the nth knowledge point is as follows: First, let's use formula four: ; Calculate the judgment index Y n ; Where g(X) is the judgment function, g(X)=1 when X>0; g(X)=0 when X≤0; σ0 is the preset ratio; C0 is the preset accuracy rate; C nmax Q represents the highest accuracy rate in historical tests for the nth knowledge point. n Q0 represents the historical test count for the nth knowledge point; Q0 is the preset test count. It should be noted that the preset ratio σ0 satisfies 1.8≤σ0≤3; the preset accuracy C0 satisfies C0≤60%; and the preset number of attempts Q0 satisfies Q0≥3. When Y=1, the self-learning effectiveness level is unsatisfactory; personalized guidance from a tutor is required. When Y=0, the self-learning effectiveness level is qualified; no personalized guidance from the tutor is required. Formula 4 Explanation: When >0 and C0>C nmax And Q n >Q 0， This indicates that although the student spent a lot of time studying this knowledge point, the test results were not ideal, possibly due to weaknesses in understanding and applying nursing concepts. Based on this assessment, the teacher can provide targeted guidance to the student. Through the above technical solution, this embodiment can quickly distinguish whether students need proactive personalized guidance from tutors by judging the index, making teaching more targeted. The detailed explanation of Formula 4 allows teachers to clearly understand the students' problems. For example, if students spend a lot of time but their results are not ideal, it can be determined that they are weak in terms of concept understanding and application, and then targeted tutoring can be carried out to effectively improve students' learning outcomes, optimize the allocation of teaching resources, and improve the overall teaching efficiency and quality.

[0039] It should be noted that the preset parameters involved in the mathematical models of this invention (including but not limited to ω0, μ1, γ1, γ2, θ) k ρ m The parameters (R1 and the proportions in each generation strategy) are determined using a dual-track method of "expert experience initialization + data-driven optimization": First, the initial range of parameter values ​​is set based on cognitive patterns and expert experience in the field of nursing education. Then, grid search or gradient optimization is performed using large-scale real student learning trajectory data (N≥200), with multi-objective comprehensive indicators such as "assessment accuracy," "recall," and "cognitive load" as optimization targets, to finally determine the optimal parameter configuration. The above parameter determination method ensures the reproducibility and optimality of the technical solution. R1 can use the effective learning index P. n The sample distribution is determined by the ternary method. σ0, C0, and Q0 can be constrained and optimized based on the inefficient self-study samples manually evaluated by the tutor. The above thresholds are not set arbitrarily, but are used to stably trigger tutor intervention in samples with "high input but low effect", thereby improving the reproducibility of self-study effect identification.

[0040] Taking the knowledge point of "indwelling urinary catheter care" as an example, let F n =0.90, T ne =2.4h, T ne0 =2.0h, t ns =24h, ρ n =1.1, ω0=0.0065h 1, μ1=1.2, γ1=0.45, γ2=0.55; the weights for the past three tests were θ1=0.20, θ2=0.35, and θ3=0.45, corresponding to C. nk The values ​​are 0.52, 0.66, and 0.74, corresponding to t. k The values ​​are 168h, 72h, and 24h. From formula two, we get ω. n ≈0.00562; Substituting into Formula 1, we get P n ≈0.621. If R1=0.50, then this knowledge point is at level three of the effective learning level. Difficult questions will be prioritized in this test to assess its comprehensive application ability. If another knowledge point satisfies T... ne / T ne0 =2.25, C nmax =0.55, Q n =4, and σ0=2.0, C0=0.60, Q0=3, then Y n =1, the system outputs a prompt for proactive tutoring.

[0041] Please see Figure 2 As shown, a method for evaluating the effectiveness of nursing training based on multi-source information fusion is described, the method comprising the following steps: S1: First, the test questions with characteristic information data are stored through the question bank module; S2: Students who log in to the personalized terminal module learn knowledge information data of each knowledge point through the learning unit according to the recommended content; S3: The learning information data and historical test information data of each knowledge point of each logged-in student are stored in the student learning record storage unit; S4: The theoretical test question bank generation unit analyzes the learning information data and historical test information data of each logged-in student for each knowledge point, and generates the theoretical test question bank in combination with the theoretical test question generation strategy. S5: The personalized learning tutoring unit analyzes the learning information data and historical test information data of the logged-in student to obtain the self-study effect level of the logged-in student for each knowledge point; the tutor provides targeted tutoring based on the self-study effect level of the logged-in student for each knowledge point.

[0042] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A nursing training effectiveness evaluation system based on multi-source information fusion, characterized in that, The system includes a question bank module and a personalized student terminal module; The question bank module is used to store test questions and feature information data of each test question; The personalized terminal module for students includes: Learning units are used to learn knowledge information data for each knowledge point according to recommended content; The student learning record storage unit is used to store the learning information data and historical test information data of each student for each knowledge point. The theoretical test question bank generation unit is connected to the logged-in student learning record storage unit and question bank module. It is used to analyze the learning information data and historical test information data of each logged-in student for each knowledge point, and generate the theoretical test question bank in combination with the theoretical test question generation strategy. The personalized learning tutoring unit is used to analyze the learning information data and historical test information data of the logged-in student to obtain the student's self-study effectiveness level for each knowledge point; the tutor provides targeted tutoring based on the student's self-study effectiveness level for each knowledge point. The theoretical test question bank generation unit and personalized learning tutoring unit are configured to perform the following core processes: Calculate the forgetting rate constant and effective learning index based on the cumulative effective learning time, preset effective learning time, learning progress, the time from the last learning time to the current time, historical test accuracy, and the time from the test time to the current time; classify effective learning levels according to the effective learning index; correlate the difficulty of the questions with the number of learned knowledge points associated with each question, where easy questions are associated with one learned knowledge point, moderate questions with two learned knowledge points, and difficult questions with three or more learned knowledge points; and determine whether active tutoring is needed based on the self-study effect judgment index.

2. The nursing training effectiveness evaluation system based on multi-source information fusion according to claim 1, characterized in that, The learning information data includes the cumulative effective learning time and the progress of learning knowledge points.

3. The nursing training effectiveness evaluation system based on multi-source information fusion according to claim 2, characterized in that, The process of generating the question bank for this theoretical test is as follows: S10: Obtain the student's cumulative effective learning time for each knowledge point, learning progress, and historical test information data by logging into the student's learning record storage unit; S20: Based on the cumulative effective learning time and learning progress of each knowledge point of the student, as well as the test information data within the past preset time period, an effective learning index for each knowledge point is obtained. S30: Analyze the effective learning index of each knowledge point to obtain the effective learning level of each knowledge point; S40: Generate this theoretical test question bank based on the effective learning level of each knowledge point and the theoretical test question generation strategy.

4. The nursing training effectiveness evaluation system based on multi-source information fusion according to claim 3, characterized in that, Using Formula 1: ； Calculate the effective learning index P of the nth knowledge point. n ; Where, ω n Let n be the forgetting rate constant for the nth knowledge point; t ns This represents the duration from the last time the nth knowledge point was studied to the current time. ne This represents the cumulative effective learning time for the nth knowledge point. ne0 F is the preset effective learning time for the nth knowledge point; n The learning progress for the nth knowledge point; K represents the number of theoretical tests within a preset time period, k∈K; C nk Let n be the accuracy rate of the nth knowledge point in the kth test; t k U represents the duration from the time of the k-th test to the current time. n θ represents the number of historical tests for the nth knowledge point; k γ1 is the weight coefficient for the k-th test; γ2 is the first weight coefficient; γ3 is the second weight coefficient.

5. A nursing training effectiveness evaluation system based on multi-source information fusion according to claim 4, characterized in that, Through formula two: ； Calculate the forgetting rate constant ω of the nth knowledge point. n ; Where ω0 is the preset basic forgetting rate constant; ρ n T represents the difficulty coefficient of the nth knowledge point; ne T represents the cumulative effective learning time for the nth knowledge point. ne0 μ1 is the preset effective learning time for the nth knowledge point; μ1 is the first adjustment coefficient.

6. The nursing training effectiveness evaluation system based on multi-source information fusion according to claim 5, characterized in that, Through formula three: ； Calculate the accuracy C of the nth knowledge point in the kth test. nk ; Where M represents the number of difficulty levels of the questions, m∈M; E knm0 Let E be the number of questions at the m-th difficulty level in the k-th test that involve the n-th knowledge point. knm0 When the value is greater than 0, this difficulty level item is included in the calculation; ρ represents the number of correct answers for the logged-in student in the m-th difficulty level question type that involves the n-th knowledge point in the k-th test; m is the weighting coefficient for the m-th difficulty level.

7. A nursing training effectiveness evaluation system based on multi-source information fusion according to claim 6, characterized in that, In step S40, the strategy for generating theoretical test questions includes: The effective learning level of knowledge points is 25%-35%; the difficulty level of all knowledge points is set to easy. The effective learning level of knowledge points is 50%-60%; the difficulty level of each knowledge point is easy, moderate and hard, and generated according to a preset ratio; The effective learning level of knowledge points is 5%-25%; the difficulty level of all knowledge points is set to difficult. Furthermore, the sum of the knowledge points for each effective learning level is 100%.

8. A nursing training effectiveness evaluation system based on multi-source information fusion according to claim 7, characterized in that, The difficulty levels are: easy, normal, and hard.

9. A nursing training effectiveness evaluation system based on multi-source information fusion according to claim 8, characterized in that, The difficulty level is defined as simple theoretical questions with only one previously learned knowledge point; The theoretical questions of general difficulty level contain only two previously learned knowledge points; The theoretical questions, classified as difficult, involve three or more previously learned knowledge points.

10. A method for evaluating the effectiveness of nursing training based on multi-source information fusion, applicable to the nursing training effectiveness evaluation system based on multi-source information fusion as described in any one of claims 1-9, characterized in that, The method includes the following steps: S1: First, the test questions with characteristic information data are stored through the question bank module; S2: Students who log in to the personalized terminal module learn knowledge information data of each knowledge point through the learning unit according to the recommended content; S3: The learning information data and historical test information data of each knowledge point of each logged-in student are stored in the student learning record storage unit; S4: The theoretical test question bank generation unit analyzes the learning information data and historical test information data of each logged-in student for each knowledge point, and generates the theoretical test question bank in combination with the theoretical test question generation strategy. S5: The personalized learning tutoring unit analyzes the learning information data and historical test information data of the logged-in student to obtain the self-study effect level of the logged-in student for each knowledge point; the tutor provides targeted tutoring based on the self-study effect level of the logged-in student for each knowledge point.

Images