Search

CN-122022529-A - AI-based learning process monitoring and intervention method and system

CN122022529ACN 122022529 ACN122022529 ACN 122022529ACN-122022529-A

Abstract

The invention discloses an AI-based learning process monitoring and intervention method and system, which belong to the technical field of education management and comprise the steps of collecting question making data, learning positive behavior data, learning load expression data, user age and learning duration of a user in the learning process, generating knowledge point mastery degree according to the question making data, the learning positive behavior data, the user age and the learning duration, generating learning cognitive load indexes according to the learning load expression data, generating intervention decision requirement degree according to the knowledge point mastery degree and the learning cognitive load indexes, and intervening the user.

Inventors

  • WU YAN

Assignees

  • 北京拾光书房文化传媒有限公司

Dates

Publication Date
20260512
Application Date
20260224

Claims (10)

  1. 1. The AI-based learning process monitoring and intervention method is characterized by comprising the following steps: step 10, collecting question making data, learning positive behavior data, learning load expression data, user age and learning duration of a user in a learning process, wherein the question making data comprises answer accuracy and answer time; step S20, establishing a knowledge point mastering evaluation model according to the question making data, the learning positive behavior data, the age and the learning duration of the user, and generating knowledge point mastering degree; step S30, generating a learning cognitive load index according to the learning load expression data; step S40, generating intervention decision-making demand according to knowledge point mastery degree and learning cognitive load index; and S50, interfering the learning of the user according to the interference decision demand.
  2. 2. The AI-based learning process monitoring and intervention method of claim 1, wherein the knowledge point mastery generation method specifically comprises: s21, generating a user capability value according to the answer accuracy; Step S22, generating a learning forgetting rate of the user according to the learning positive behavior data, the age of the user and the learning duration; and S23, establishing a knowledge point mastering evaluation model according to the answering time, the user capacity value and the user learning forgetting rate, and generating knowledge point mastering degree.
  3. 3. The AI-based learning process monitoring and intervention method of claim 2, wherein the generating means of the user capability value specifically comprises: By the formula: ; Generating user capability values ; In the formula, m represents the total number of random basic questions, and r represents the correct answer number of random m-channel basic questions.
  4. 4. The AI-based learning process monitoring and intervention method of claim 2, wherein the generating means of the user learning forgetting rate specifically comprises: By the formula: ; generating user learning forgetting rate ; In the formula (i) the formula (ii), The function is activated for Sigmoid, The maximum forgetting rate is indicated, Represents the standardized value of the kth learning positive behavior data, m is the learning positive behavior data quantity, w k is the weight coefficient of the learning positive behavior, An age weight coefficient, age, a user age, a ref , a reference base age, T h , a learning strength of an h period, The period sensitivity coefficient of the h period is shown, and n is the period number.
  5. 5. The AI-based learning process monitoring and intervention method of claim 4, wherein the generation mode of the learning intensity of the h period specifically comprises: By the formula: ; Generating a learning intensity T h of the h period; In the formula, t h represents the learning duration of the h period, For the learning concentration of the h period, t day represents the average total daily learning duration of the user, and the sum of the durations of all the periods is one day.
  6. 6. The AI-based learning process monitoring and intervention method of claim 2, wherein the knowledge point grasping and evaluating model expression is specifically: ; In the formula, M k represents knowledge point mastery, Indicated are user capability values for knowledge point k, The question difficulty of the knowledge point k is shown, Represented is a topic differentiation degree coefficient, Indicating that the user has learned the forgetting rate, The time interval from the last learning is represented, t ans represents the time for answering the question, t k represents the standard answering time of the knowledge point k, Penalty coefficients for the duration of the answer.
  7. 7. The AI-based learning process monitoring and intervention method of claim 1, wherein the generating means of the learning cognitive load index specifically comprises: By the formula: ; Generating a learning cognitive load index L; In the formula (i) the formula (ii), As a standard normal cumulative function, Z (d j ) represents a normalized value of the learning load expression data d j , r j is a weight coefficient of the learning load expression, q is the number of the learning load expression data, and j is a number index of the learning load expression data.
  8. 8. The AI-based learning process monitoring and intervention method of claim 1, wherein the generating means of the intervention decision demand level specifically comprises: By the formula: ; generating an intervention decision requirement degree I; Wherein M k represents knowledge point grasping degree, L represents learning cognitive load index, and F (t) represents time period sensitivity factor.
  9. 9. The AI-based learning process monitoring and intervention method of claim 6, wherein the time-period-sensitive factor is generated by a method specifically comprising: By the formula: ; generating a period sensitivity factor F (t); In the formula, F base represents a basic period sensitivity factor, G is an amplitude of a sine function, T represents a current time, T peak represents an optimal learning time, and T is a time length of a period.
  10. 10. AI-based learning process monitoring and intervention system for performing the AI-based learning process monitoring and intervention method of any of claims 1-9, comprising in particular: The system comprises a data acquisition unit, a data processing unit and a data processing unit, wherein the data acquisition unit is used for acquiring question making data, learning positive behavior data, learning load expression data, user age and learning duration of a user in a learning process, wherein the question making data comprises question answering accuracy and question answering time; The knowledge point mastering and evaluating unit is used for establishing a knowledge point mastering and evaluating model according to the question making data, the learning positive behavior data, the age of the user and the learning duration to generate knowledge point mastering degree; The learning cognitive load analysis unit is used for generating a learning cognitive load index according to the learning load expression data; The intervention decision-making demand analysis unit is used for generating an intervention decision-making demand according to the knowledge point grasping degree and the learning cognitive load index; And the intervention unit is used for intervening the learning of the user according to the intervention decision demand.

Description

AI-based learning process monitoring and intervention method and system Technical Field The invention belongs to the technical field of education management, and particularly relates to an AI-based learning process monitoring and intervention method and system. Background With the popularization of online education platforms and the development of intelligent technologies, how to accurately monitor learning processes and to implement effective interventions becomes a core challenge in the education field. The traditional learning evaluation method mainly depends on static indexes such as examination results or completion rates, and is difficult to reflect knowledge mastering states and cognitive load levels of users in real time. Most systems judge the learning effect only through the answering accuracy, and ignore dynamic factors such as answering time, forgetting curves and the like, so that the evaluation result is one-sided. For example, users may have insufficient actual mastery due to guessing answers to questions, and conventional approaches cannot recognize such situations, and in addition, conventional interventions rely on fixed thresholds or rules to dynamically adapt to individual differences (e.g., age, learning period preference) and real-time state changes of users. For example, nighttime learning efficiency is generally reduced, but the system still pushes high intensity exercises on an daytime basis, resulting in a reduced user experience. These problems lead to the accuracy of the prior art in monitoring the learning process and the intervention of the learning process, thereby affecting the learning efficiency of the user. The prior art lacks comprehensive analysis capability of multidimensional data in the learning process, cannot accurately evaluate the dynamic grasping degree of knowledge points, and is difficult to perform personalized intervention according to the real-time cognitive load state of a user, so that the ineffective learning time of the user is prolonged. Disclosure of Invention Aiming at the defects of the prior art, the invention provides an AI-based learning process monitoring and intervention method and system, which solve the problems. The invention is realized by the following technical proposal that the AI-based learning process monitoring and intervention method comprises the following steps: collecting question making data, learning positive behavior data, learning load expression data, user age and learning duration of a user in a learning process, wherein the question making data comprises question answering accuracy and question answering time; establishing a knowledge point mastering evaluation model according to the question making data, the learning positive behavior data, the user age and the learning duration, and generating knowledge point mastering degree; Generating a learning cognitive load index according to the learning load expression data; Generating intervention decision-making demand according to knowledge point mastery degree and learning cognitive load index; And according to the interference decision demand, performing interference on the learning of the user. Based on the technical scheme, the invention also provides the following optional technical schemes: The further technical scheme is that the knowledge point mastery degree generation mode specifically comprises the following steps: generating a user capability value according to the answer accuracy; Generating a learning forgetting rate of the user according to the learning positive behavior data, the age of the user and the learning duration; And establishing a knowledge point mastering evaluation model according to the answering time, the user capability value and the user learning forgetting rate, and generating knowledge point mastering degree. The generation mode of the user capacity value specifically comprises the following steps: By the formula: ; Generating user capability values ; In the formula, m represents the total number of random basic questions, and r represents the correct answer number of random m-channel basic questions. The generation mode of the user learning forgetting rate specifically comprises the following steps: By the formula: ; generating user learning forgetting rate ; In the formula (i) the formula (ii),The function is activated for Sigmoid,The maximum forgetting rate is indicated,Represents the standardized value of the kth learning positive behavior data, m is the learning positive behavior data quantity, w k is the weight coefficient of the learning positive behavior,An age weight coefficient, age, a user age, a ref, a reference base age, T h, a learning strength of an h period,The period sensitivity coefficient of the h period is shown, and n is the period number. The generation mode of the learning intensity of the h period specifically comprises the following steps: By the formula: ; Generating a learning intensity T h of the h period; In the form