A teaching process optimization method based on adaptive cognitive feedback

Abstract

The present application relates to the technical field of education service, in particular to a teaching process optimization method based on adaptive cognitive feedback, the present application divides experimental teaching content into experimental project level teaching knowledge unit and subordinate experimental step and micro-operation level sub-knowledge unit, associates experimental behavior data with sub-knowledge unit one by one, and aggregates operation completion, error and correction time, scale reading deviation, repeated demonstration times, question or help times and other multi-source behavior characteristics in the dimension of sub-knowledge unit, calculates individual baseline parameters for each student on each sub-knowledge unit and updates cognitive feedback threshold interval combined with historical behavior data, so that the teaching management system can finely depict the cognitive mastery degree and operation proficiency of students at the specific experimental step level, thereby improving the resolution ability of students' real experimental ability and weak micro-operation links.

Description

Technical Field

[0001] This invention relates to the field of educational service technology, and in particular to a method for optimizing the teaching process based on adaptive cognitive feedback. Background Technology

[0002] In experimental teaching of subjects such as physics, chemistry, and biology, schools typically deploy experimental teaching management platforms or teaching information systems. These systems provide students with pre-set experimental procedure instructions, key points, and precautions through paper-based experimental manuals, PowerPoint presentations, or electronic experimental guidance interfaces. Teachers organize classes on terminal devices according to pre-configured experimental teaching scripts, and students complete practical tasks such as wiring, solution preparation, titration, microscope focusing, and data recording following the instructions. Some existing digital experimental teaching management systems can record students' experimental scores, online quiz responses, and some interaction logs on servers or teaching terminals for statistical analysis of teaching effectiveness and evaluation of teaching quality after class. Some systems also attempt to perform simple grading of experimental difficulty based on student accuracy or time taken in the background to support teaching management decisions.

[0003] However, existing experimental teaching management systems still have shortcomings in data processing and intelligent decision-making for the experimental process. On the one hand, the system usually uses overall outcome indicators such as "whether the experiment was completed, the test score of this experimental class, and the average accuracy rate of a certain chapter" as the main basis for students' experimental abilities. It only counts students' status at the level of experimental project or chapter. The data processing and evaluation logic in the background is difficult to distinguish whether students do not understand the whole experimental principle and idea, or repeatedly make mistakes in a certain key micro-operation such as wiring sequence, titration endpoint judgment, microscope focusing, and scale reading. This leads to many limitations in teaching management strategies. Coarse-grained script adjustments are insufficient to support fine-grained interventions for specific steps. On the other hand, the existing system lacks a one-to-one correspondence between the learning behavior data collected and the specific experimental steps. The relevant data is mostly stored in the form of raw logs or videos, and is only used for post-class playback or quality inspection. The system has not built structured features on the server side or teaching terminal side that can be directly called and processed by the algorithm. As a result, the experimental teaching management system is unable to quantitatively evaluate details such as wiring standardization, reading accuracy, and operational proficiency in data processing, and it is also difficult to drive the automatic adjustment of experimental scripts and the intelligent optimization of the teaching process based on fine-grained cognitive states.

[0004] To address this, a teaching process optimization method based on adaptive cognitive feedback is proposed. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, this invention provides a teaching process optimization method based on adaptive cognitive feedback.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a teaching process optimization method based on adaptive cognitive feedback, comprising the following steps: S1, the teaching management system divides the course content into teaching knowledge units and sub-knowledge units based on a preset teaching syllabus, collects learning behavior data from each collection terminal according to sub-knowledge units, collects and structures the learning behavior data, calculates the individual baseline parameters of students in each sub-knowledge unit, and updates the corresponding cognitive feedback threshold interval according to the individual baseline parameters; S2, based on the learning behavior data and the cognitive feedback threshold interval, calculates the cognitive feedback index and cognitive uncertainty of students in each sub-knowledge unit, and determines the information gain and collection cost based on the cognitive uncertainty and a preset collection cost model. The ratio between the two values ​​is used to determine the cognitive feedback collection method according to preset rules when the ratio is greater than the collection trigger threshold. Otherwise, cognitive feedback collection is not triggered. S3. For each teaching knowledge unit, the cognitive dispersion is calculated based on the cognitive feedback index of the sub-knowledge unit. When the cognitive dispersion is higher than the first threshold, the teaching knowledge unit is split into multiple sub-knowledge unit teaching links in the teaching script. When the cognitive dispersion is lower than the second threshold and the correlation condition is met, the teaching knowledge unit and adjacent teaching knowledge units are merged into one teaching link in the teaching script. S4. During the execution of the teaching script, the distribution of students' cognitive feedback indexes relative to their respective cognitive feedback threshold intervals is statistically analyzed, and the subsequent teaching scripts are adjusted according to the distribution results.

[0007] As a preferred technical solution of the present invention, the teaching knowledge unit corresponds to the content of physics, chemistry and biology experimental courses. Each teaching knowledge unit corresponds to an experimental project or a class of closely related experimental projects in the experimental teaching syllabus. Each sub-knowledge unit corresponds to a micro-skill or sub-ability under the experimental project. The operational behaviors collected by students in the experimental steps are associated one-to-one with the corresponding sub-knowledge units, and the behavioral features are aggregated at the sub-knowledge unit level to complete the collection of learning behavior data by sub-knowledge unit. Within a preset sliding time window, the completion records and interactive behaviors of students in the relevant experimental tasks of the sub-knowledge unit are statistically analyzed, and the historical average level and fluctuation index of students in the sub-knowledge unit are calculated to form the individual baseline parameter vector of students in the sub-knowledge unit. The lower limit of the cognitive feedback threshold interval is set as the historical average level of students in the sub-knowledge unit minus a preset multiple of the fluctuation range, and the upper limit is set as the historical average level plus a preset multiple of the fluctuation range. When new experimental behavior data is written, the adjustment index of the historical average level and fluctuation range is updated online incrementally, and the cognitive feedback threshold interval is adjusted accordingly.

[0008] As a preferred technical solution of the present invention, the experimental learning behavior features corresponding to the sub-knowledge unit are standardized and weighted to generate an original performance score. Based on the relative position of the original performance score within the historical individual baseline and the corresponding cognitive feedback threshold range, a signed cognitive feedback index is calculated. The variance of the student's at least one cognitive feedback index for the sub-knowledge unit is statistically analyzed, or a probability model is used to perform a posterior estimation of the mastery status, with variance, confidence interval width, or distribution entropy used as the cognitive uncertainty. The acquisition cost model uses the number of cognitive feedback acquisitions triggered by the student within the current time window and the distance from the last cognitive feedback... The data collection time interval, average collection time, and remaining time in the current experimental stage are used as inputs. The collection cost is calculated using a weighted function. The collection cost increases when the number of cognitive feedback collections increases, the average collection time increases, or the remaining time in the current experimental stage decreases. A ratio is defined to the information gain and collection cost. When the ratio is greater than a preset collection trigger threshold, the teaching management system triggers cognitive feedback collection for students at the current moment. When the ratio is less than or equal to the trigger threshold, no new collection is triggered. The cognitive feedback collection method is selected according to preset rules based on the sub-knowledge unit or cognitive dimension where the cognitive uncertainty is mainly concentrated.

[0009] As a preferred embodiment of the present invention, for a given experimental project, the teaching management system first determines the multiple sub-knowledge units included in the experimental project, and obtains the cognitive feedback indicators corresponding to each sub-knowledge unit at the current moment. The cognitive feedback indicator for each sub-knowledge unit can be the average cognitive feedback indicator for the entire class on that sub-knowledge unit, or an indicator obtained by aggregating the cognitive feedback indicators of different students according to preset weights. Based on the degree of dispersion among the cognitive feedback indicators of all sub-knowledge units under the experimental project, the teaching management system calculates the cognitive dispersion, which characterizes the degree of cognitive difference within the experimental project. The cognitive dispersion can be expressed as the variance and standard deviation of the cognitive feedback indicators of each sub-knowledge unit. The variance can be determined by weighting the variance based on the importance of each sub-knowledge unit; a first threshold is set, and when the cognitive dispersion is greater than the first threshold, the corresponding experimental project is split into multiple sub-knowledge unit teaching segments in the teaching script; a second threshold is set, and a correlation condition is introduced: the cognitive feedback vectors of the experimental project and adjacent experimental projects are tracked on multiple time slices, and the correlation coefficient between the corresponding sub-knowledge units is calculated, or the overall correlation coefficient of the cognitive feedback vectors of the experimental project and adjacent experimental projects is calculated on multiple stages. When the cognitive dispersion is less than the second threshold and the above correlation coefficient is greater than the preset threshold, the two experimental projects are merged into a comprehensive experimental segment in the teaching script.

[0010] As a preferred technical solution of the present invention, for the experimental stage currently being executed or completed, the cognitive feedback indicators and corresponding threshold ranges of all students in the current class on the experimental projects or sub-knowledge units associated with the experimental stage are obtained. The proportions of students falling below the lower threshold, between the lower and upper thresholds, and above the upper threshold are statistically recorded as the first proportion, the second proportion, and the third proportion, respectively. When the first proportion is greater than the preset difficulty proportion threshold, remedial script units are automatically inserted or enhanced for the knowledge units in the subsequent experimental scripts that have not yet been executed, and the progress speed of the relevant stages is reduced. When the third proportion is greater than the preset ease proportion threshold, repetitive basic operation exercises are deleted, and challenging comprehensive experimental tasks are added. When the first proportion and the third proportion both exceed their respective difference thresholds, and the second proportion is small, a tiered experimental teaching script unit is inserted in the subsequent scripts.

[0011] Compared with the prior art, the beneficial effects that this invention can achieve are:

[0012] 1. This invention divides experimental teaching content into experimental project-level teaching knowledge units and their subordinate experimental steps and micro-operation-level sub-knowledge units. It associates experimental behavioral data with sub-knowledge units one-to-one and aggregates multi-source behavioral characteristics such as operation completion status, error and correction time, scale reading deviation, number of repeated demonstrations, and number of questions or requests for help at the sub-knowledge unit level. Combined with historical behavioral data, it calculates individual baseline parameters for each student in each sub-knowledge unit and updates the cognitive feedback threshold range. This enables the teaching management system to finely characterize students' cognitive mastery and operational proficiency at the level of specific experimental steps, thereby improving the ability to distinguish students' real experimental abilities and weak micro-operation links.

[0013] 2. Based on the calculation of cognitive feedback indicators, this invention introduces a cognitive uncertainty index obtained from recent experimental records or probability models, and constructs a collection cost model with inputs such as the number of cognitive feedback collections, collection time, and remaining class time. The ratio of information gain to collection cost is calculated to control whether to trigger new cognitive feedback collection and the collection method at the current moment. When the cognitive uncertainty is high and the collection cost is low, targeted short tests, operation checks, or self-assessment feedback can be initiated first. When the state is relatively stable or class time is tight, redundant assessments can be reduced, thereby reducing interference with the experimental process and improving the efficiency and timeliness of cognitive feedback collection.

[0014] 3. This invention calculates the cognitive dispersion within experimental projects by summarizing the cognitive feedback indicators of sub-knowledge units. Combined with threshold and correlation conditions, it performs splitting or merging of experimental projects in the teaching script, achieving adaptive adjustment of the granularity of experimental projects and steps. Simultaneously, during the execution of the experimental teaching script, it statistically analyzes the distribution of students' cognitive feedback indicators relative to their respective threshold ranges in the current experimental stage. Based on distribution characteristics such as overall difficulty, overall ease, or significant stratification, it automatically inserts remedial script units, compresses basic exercises, or adds stratified experimental teaching scripts for subsequent unexecuted scripts. This allows more teaching time to be invested in key weak areas, reduces low-value repetition of already mastered operations, improves training relevance and classroom time utilization, and enhances the adaptability of the physics, chemistry, and biology practical teaching process to students' dynamic states. Attached Figure Description

[0015] Figure 1 This is a flowchart of the method of the present invention;

[0016] Figure 2 This is a flowchart of step S1 of the present invention;

[0017] Figure 3 This is a flowchart of step S2 of the present invention;

[0018] Figure 4 This is a flowchart of step S3 of the present invention;

[0019] Figure 5 This is a flowchart of step S4 of the present invention;

[0020] Figure 6 This is a schematic diagram of the structure of the experimental teaching process optimization device of the present invention.

[0021] Among them: 1. Smart all-in-one machine; 2. Global camera; 3. Detail camera; 4. Front camera. Detailed Implementation

[0022] To make the technical means, creative features, objectives, and effects of this invention easier to understand, the invention is further described below with reference to specific embodiments. However, the following embodiments are merely preferred embodiments of this invention and not all of them. Other embodiments obtained by those skilled in the art based on the embodiments described herein without creative effort are all within the protection scope of this invention.

[0023] Example: Figure 1 As shown, a teaching process optimization method based on adaptive cognitive feedback includes the following steps:

[0024] S1. The teaching management system divides the course content into teaching knowledge units and sub-knowledge units based on the preset teaching syllabus. It collects learning behavior data from each collection terminal according to the sub-knowledge units, collects and stores the learning behavior data in a structured manner, calculates the individual baseline parameters of students in each sub-knowledge unit, and updates the corresponding cognitive feedback threshold range according to the individual baseline parameters.

[0025] S2. Based on learning behavior data and cognitive feedback threshold range, calculate the cognitive feedback index and cognitive uncertainty of students in each sub-knowledge unit. Determine the ratio between information gain and collection cost according to the cognitive uncertainty and the preset collection cost model. When the ratio is greater than the collection trigger threshold, trigger cognitive feedback collection at the current moment and determine the cognitive feedback collection method according to the preset rules. Otherwise, do not trigger cognitive feedback collection.

[0026] S3. For each teaching knowledge unit, calculate the cognitive dispersion based on the cognitive feedback index of the sub-knowledge unit. When the cognitive dispersion is higher than the first threshold, the teaching knowledge unit is split into multiple sub-knowledge unit teaching segments in the teaching script. When the cognitive dispersion is lower than the second threshold and the relevance condition is met, the teaching knowledge unit and adjacent teaching knowledge units are merged into one teaching segment in the teaching script.

[0027] S4. During the execution of the teaching script, statistically analyze the distribution of students' cognitive feedback indicators relative to their respective cognitive feedback threshold ranges, and adjust subsequent teaching scripts based on the distribution results.

[0028] like Figure 2 As shown, in practice, the teaching management system first breaks down the content of practical physics, chemistry, and biology courses into knowledge units. Each experimental course is divided into several teaching knowledge units according to teaching organization habits. Each teaching knowledge unit corresponds to an experimental project or a closely related type of experimental project in the experimental teaching syllabus, such as Ohm's Law experiment, connection and measurement of series and parallel circuits, determination of solution concentration by acid-base titration, use of an optical microscope, and determination of enzyme catalytic reaction rate. Based on this, each teaching knowledge unit is further refined into several sub-knowledge units. Each sub-knowledge unit corresponds to a micro-skill or sub-ability under an experimental project, such as correctly identifying electrical... The process involves several steps: connecting the flow meter terminals, properly connecting the sliding rheostat, correctly washing and rinsing the burette before titration, keeping the line of sight horizontal to the liquid surface when reading the scale, adjusting the coarse and fine focal wheels of the microscope according to specifications until the image is clear, and recording experimental data in a standardized manner while retaining significant figures. This process forms a hierarchical knowledge structure, from the course to the experimental project (teaching knowledge unit) and then to the experimental micro-operation (sub-knowledge unit). Through this breakdown, on the one hand, the knowledge granularity is ensured to be sufficiently fine, which can be mapped to specific experimental steps, operational points, or safety checks; on the other hand, the logical order of the original experimental course system is not disrupted, making it easy to connect with existing experimental guides and experimental teaching scripts.

[0029] Based on the aforementioned hierarchical knowledge structure, this embodiment associates and collects learning behavior data generated in the experimental teaching platform with sub-knowledge units in a one-to-one correspondence. Specifically, when a student generates a behavioral event in an experimental step corresponding to a certain sub-knowledge unit, such as inserting a wire in the circuit connection step, looking down to read the meter in the meter reading step, observing the color change and stopping the titration in the titration endpoint determination step, or rotating the coarse and fine adjustment wheels in the microscope focusing step, the collected action segments will be marked as operational behaviors under the corresponding sub-knowledge unit by the teaching management system. At the same time, student feedback operations, such as clicking "completed step," "need demonstration," or "does not understand," will also be recorded. Buttons, submission of quizzes for corresponding steps, etc., are also recorded and associated with the same sub-knowledge unit number; preferably, the teaching management system can aggregate one or more of the following behavioral characteristics according to the sub-knowledge unit dimension, such as whether the operation is completed in one go, whether there are incorrect wiring or incorrect operation, the time taken to correct the error, whether the reading scale deviates significantly from the actual value, the number of times the instruction steps are repeatedly demonstrated or read repeatedly, the number of times questions are asked or requests for help are made for the steps, etc.; by collecting learning behavior data according to sub-knowledge units, the subsequent estimation of students' cognitive mastery and operational proficiency in the practical stage can be accurate to the specific micro-operation, avoiding the mutual cancellation of performance between different steps.

[0030] Furthermore, in this embodiment, in order to characterize the long-term stable level of students at the sub-knowledge unit granularity for subsequent judgment of cognitive feedback, the teaching management system calculates individual baseline parameters for each student at each sub-knowledge unit based on the learning behavior data collected by sub-knowledge unit; specifically, for students In sub-knowledge units Within a preset sliding time window, the teaching management system statistically analyzes the completion records and interactive behaviors of recent sub-knowledge unit-related experimental tasks. Preferably, this includes at least one of the following: the accuracy rate of step-related test questions, the average time spent completing the operation steps, the number of times errors were made and corrected, the number of times the step demonstration video was watched repeatedly, and the number of questions or requests for help regarding the steps. The teaching management system then calculates the students'... In sub-knowledge units Historical average accuracy Historical average operation time The historical average level, along with corresponding fluctuation indicators such as standard deviation or quantiles, are used to compile the above statistical results into a student... In sub-knowledge units Individual baseline parameter vectors on.

[0031] Based on this, the teaching management system updates the corresponding cognitive feedback threshold range according to individual baseline parameters; preferably, for students In sub-knowledge units A comprehensive behavioral indicator For example, the comprehensive performance score is obtained by weighting normalized factors such as accuracy, time taken, and number of errors. The teaching management system uses a formula... , Determine the lower threshold With upper limit threshold ,in For the corresponding volatility indicator, The preset coefficients form the cognitive feedback threshold range. , As the experimental teaching process progresses and new experimental data is continuously written in, the teaching management system uses exponentially weighted moving averages or other online update strategies to update the data. , An online incremental update method is adopted, in which the original parameter values ​​are corrected only by a small update step size in each update, and the update is synchronized accordingly. , This allows each student's cognitive feedback threshold range for each sub-knowledge unit to reflect both their long-term performance and recent changes.

[0032] Through the above design, on the one hand, by adopting a hierarchical knowledge structure from experimental projects to experimental steps or micro-operations, and collecting experimental behavior data according to sub-knowledge units, the teaching management system can characterize students' practical abilities at the micro-operation granularity. Compared with the approach of only calculating scores for the entire experiment or the entire chapter, it can more accurately distinguish whether students do not understand the whole, are not familiar with individual operations, or have only made systemic errors in a key step. On the other hand, by calculating individual baseline parameters based on historical behavior data, and setting cognitive feedback threshold ranges for each student and each sub-knowledge unit accordingly, the bias caused by using a uniform threshold or class average threshold is avoided. It can fully consider the individual differences of different students in terms of operation speed, level of caution (whether they frequently repeat confirmation), experimental experience, etc., thereby reducing the probability of misjudging operation as too slow or too fast, insufficient mastery, or wasteful repetitive training, and providing highly reliable basic data for subsequent cognitive feedback collection strategies and script optimization.

[0033] like Figure 3 As shown, in one embodiment of the present invention, based on establishing individual baseline parameters for each student in each sub-knowledge unit and obtaining the corresponding cognitive feedback threshold range accordingly, it is determined whether to initiate new cognitive feedback collection at the current moment and what collection method to use; specifically, this embodiment first combines the experimental behavior data collected by sub-knowledge unit with the cognitive feedback threshold range to quantify the student's current performance in each sub-knowledge unit and obtain cognitive feedback indicators that reflect the current mastery and operation status.

[0034] The teaching management system can standardize multiple experimental behavioral characteristics, such as quiz correctness, experimental operation completion status (whether it was successful on the first try), operation time, number of times the demonstration video was watched repeatedly, number of incorrect operations detected by the camera, and number of times the teacher was consulted or the "can't do it" button was clicked, to generate a raw performance score. Based on the relative position of the raw performance score within the historical baseline and corresponding threshold range, a signed cognitive feedback index is calculated. When the current performance is significantly better than the historical average and the upper limit threshold, a signified cognitive feedback index is generated. When the cognitive feedback index is positive and relatively large, it indicates that students perform the corresponding experimental steps with significant ease and stability; when it is significantly worse than the historical average and lower threshold, it indicates that students perform the corresponding experimental steps with significant ease and stability. When the cognitive feedback index is negative and the value is large, it indicates that the corresponding experimental steps are too difficult or unfamiliar to students; within the normal fluctuation range, it is close to zero.

[0035] In this embodiment, preferably, the cognitive uncertainty and the collection triggering rules can be implemented as follows: The teaching management system introduces cognitive uncertainty based on the above-mentioned cognitive feedback indicators. To characterize the reliability of cognitive feedback indicators; preferably, the teaching management system collects student statistics. In sub-knowledge units The variance of the most recent N cognitive feedback indicators can be used, or a probabilistic model can be used to make a posterior estimate of the mastery state, and the variance, confidence interval width, or distribution entropy can be used as the cognitive uncertainty. When the number of recent experiments is low, the performance fluctuates greatly each time, or the camera's recognition of the operation status of the corresponding experimental step is very unstable, Take the larger value; when students show consistent performance multiple times in the corresponding experimental steps. Take the smaller value.

[0036] To avoid disrupting the experimental process by inserting too many additional quizzes or subjective questionnaires, this embodiment establishes a data collection cost model; preferably, the data collection cost model is based on student... Number of cognitive feedback collections triggered within the current time window The time interval since the last cognitive feedback collection Average time taken for the most recent data collection Remaining time for the current experimental stage As well as the location of experimental safety and critical steps, etc., are taken as inputs and weighted by a function. Calculate the cost of data collection ,when Larger longer or In shorter times, It increases accordingly.

[0037] Based on this, the teaching management system defines the ratio of information benefits to collection costs. ,in This represents the measure of cognitive uncertainty that can be reduced by collecting an additional cognitive feedback, preferably in conjunction with... Proportional; when Greater than the preset data collection trigger threshold At that moment, the teaching management system triggers a response targeting students. Cognitive feedback collection; when Less than or equal to If this happens, no new data collection will be triggered.

[0038] When triggering data collection, the teaching management system also selects the cognitive feedback collection method according to preset rules based on the sub-knowledge unit or cognitive dimension where the cognitive uncertainty is mainly concentrated. For example, when the uncertainty is mainly concentrated on conceptual understanding and phenomenon identification steps such as titration endpoint judgment, a small number of short multiple-choice or true / false questions on endpoint judgment are pushed first. When the uncertainty is mainly concentrated on operational execution dimensions such as whether the wiring is standardized or whether the steps are completed in sequence, simple self-assessment scale questions are popped up first, such as how do you feel about this step: very proficient, average, or completely unable, and prompting students to take photos or short videos for self-checking. When the uncertainty is more dispersed, short tests or comprehensive operation checklists containing multiple key steps can be used for data collection. Through the above-mentioned triggering mechanism that links uncertainty with collection cost, this embodiment can prioritize the collection of cognitive feedback that is most valuable to subsequent experimental arrangements without interrupting the experimental process as much as possible.

[0039] like Figure 4 As shown, in one embodiment of the present invention, after obtaining the cognitive feedback indicators of students on each sub-knowledge unit, the granularity of the experimental project corresponding to the teaching knowledge unit is dynamically adjusted based on these indicators; specifically, for each experimental project, this embodiment summarizes the cognitive feedback indicators corresponding to each experimental step or micro-operation, and measures the consistency between the cognitive state and operational proficiency within the experimental project.

[0040] The teaching management system arranges the cognitive feedback indicators of all sub-knowledge units under the corresponding experimental project into a vector, and calculates the cognitive dispersion based on the vector. The cognitive dispersion can be represented by the variance, range, weighted standard deviation, or weighted variance based on preset weights of the cognitive feedback indicators of the sub-knowledge units.

[0041] In this embodiment, to achieve dynamic adjustment of the granularity of experimental projects, preferably, the teaching management system calculates the cognitive dispersion within the experimental project based on the cognitive feedback index of the sub-knowledge unit, and determines the splitting or merging operation in combination with relevance conditions; specifically, for a certain experimental project For example, in the Ohm's Law experiment, the set of subordinate knowledge units is denoted as... For example, identifying component symbols, standardizing wiring, correctly reading voltage and ammeter readings, organizing experimental data and plotting graphs, etc., the teaching management system constructs a cognitive feedback vector for the experimental project at the current moment. ,in It can be used for the whole class in sub-knowledge units Average cognitive feedback metrics or weighted aggregated metrics; teaching management system with The variance or standard deviation is used as cognitive dispersion. Alternatively, a weighted variance approach can be used to account for safety-related and error-sensitive critical steps. Their contribution is even greater.

[0042] To identify experimental items suitable for splitting, this embodiment sets a first threshold. When cognitive dispersion Greater than When this occurs, it indicates a significant difference in cognitive feedback between different experimental steps within the experimental project. For example, the wiring and safety check steps are generally performed well, while the reading and data processing steps are generally difficult. In this case, the teaching management system will adjust the corresponding experimental project... The teaching script is divided into multiple sub-knowledge units. In the subsequent experimental arrangements, the Ohm's Law experiment is no longer simply repeated as a whole, but is broken down into sub-units such as wiring and safety inspection training and reading and error control training. Each unit is organized with demonstrations, exercises and reviews that match the corresponding sub-knowledge unit.

[0043] To identify experimental items suitable for merging, this embodiment sets a second threshold. ( The teaching management system tracks experimental projects across multiple time slices, and introduces relevant conditions. and adjacent experimental projects The cognitive feedback vector is used to calculate the correlation coefficient between corresponding sub-knowledge units, or to calculate the correlation coefficient between multiple stages. and The overall correlation coefficient when Less than Furthermore, the aforementioned correlation coefficient is greater than the preset threshold. At that time, it was considered that the experimental project and adjacent experimental projects Students exhibit high consistency in cognitive performance and operational proficiency with no significant internal differentiation. For example, students demonstrate similar levels of mastery of each operational step in the two projects: basic titration operation training and acid-base titration for determining the concentration of an unknown solution. In this case, the teaching management system can merge these two experimental projects into a comprehensive experimental segment in the teaching script, retaining necessary explanations of key steps and a small number of representative operational exercises, while deleting repetitive basic operational training, thereby reducing teaching time and improving classroom time utilization.

[0044] By splitting and merging the above-mentioned conditions based on cognitive dispersion and relevance, this embodiment enables the granularity of the items and steps used in the experimental teaching script to change dynamically with cognitive feedback, ensuring that key abilities are fully trained while avoiding low-value repetitive drills.

[0045] like Figure 5 As shown, in one embodiment of the present invention, during the execution of a specific experimental classroom, the experimental teaching script that has not yet been executed is adjusted online based on the distribution of student cognitive feedback at the class level, so that the teaching process can still be optimized based on the latest cognitive feedback during the execution stage.

[0046] When the experimental teaching script executes each experimental step in a preset order, the teaching management system calculates the distribution of cognitive feedback indicators of all students in the current class on the corresponding experimental items or sub-knowledge units of the experimental step relative to their respective cognitive feedback threshold ranges during or after the execution of each experimental step.

[0047] Furthermore, in this embodiment, in order to adjust subsequent experimental scripts based on the latest cognitive feedback during the classroom execution phase, preferably, the teaching management system statistically analyzes the distribution of students' cognitive feedback indicators relative to their respective cognitive feedback threshold ranges in real time according to the experimental stage, and triggers different script adjustment strategies based on the distribution characteristics; specifically, for the experimental stage that is currently being executed or has just ended... For example, the acid-base titration practice in this lesson, the experimental section Related experimental projects or sub-knowledge units are denoted as The teaching management system retrieves information on all students in the current class. Cognitive feedback metrics and the corresponding threshold range [ , ], statistics fall into The following, between and Between and above The student proportions are respectively denoted as the first proportion. Second proportion Third proportion .

[0048] Preferably, when Greater than the preset difficulty ratio threshold At that time, the teaching management system determined that the current experimental step was too difficult for most students or that they did not fully understand the key operations. Therefore, in subsequent experimental scripts that had not yet been executed, adjustments were made to the knowledge units... Automatically insert or enhance remedial script units, including adding more basic operation demonstration videos, breaking down operation instructions into more detailed steps (e.g., breaking down titration into several smaller steps such as reagent loading, venting, zero-point calibration, slow addition—endpoint judgment), arranging demonstration experiments with lower difficulty and more obvious phenomena, and appropriately slowing down the progress of related steps; when Greater than the preset simple ratio threshold When the teaching management system determines that the current experimental step is too easy or too repetitive for most students, it can remove some repetitive basic operation exercises in subsequent scripts and add more challenging comprehensive experimental tasks, such as requiring students to design a circuit or prepare a solution of a target concentration under given constraints, or introducing subsequent experimental projects in advance to avoid wasting class time on operations that most students are already familiar with; and At the same time, they exceeded their respective difference thresholds. , ,and The smaller the number of students, the more obvious the stratification of experimental projects within the class. In this embodiment, a stratified experimental teaching script unit is inserted into the subsequent script to divide students into different subgroups according to their current cognitive feedback level. For example, students with lower mastery levels are given intensive basic operation training, while students with higher mastery levels are given exploratory or extended experimental tasks.

[0049] The aforementioned online adjustment process of experimental scripts based on cognitive feedback distribution can form a closed-loop optimization system together with cognitive feedback collection strategies and experimental item granularity adjustment strategies: first, it determines when and what kind of feedback to collect; then, it determines the granularity at which experimental items are organized and combined; and finally, it determines how to adjust subsequent scripts in real time during actual execution. Compared with the method of adjusting the experimental teaching plan as a whole based on grades only at the beginning or end of the semester, this invention introduces a mechanism for online script adjustment based on cognitive feedback distribution within a single experimental class. This makes the experimental teaching process not only optimizable on a long-term scale, but also adaptable to the dynamic changes in students' operational level and comprehension within each experimental class, thereby further improving the timeliness and precision of the optimization of the physics, chemistry, and biology practical teaching process.

[0050] like Figure 6As shown, in one embodiment of the present invention, the experimental teaching process optimization device for performing the above method includes: an intelligent all-in-one machine 1, a global camera 2 for capturing the experimental operation process, a detail camera 3 for capturing subtle movements, and a front-facing camera 4 for identifying the examinee's identity; the intelligent all-in-one machine 1 integrates a computing unit, a storage unit, and a display screen, and is equipped with a teaching management system that carries the method and is responsible for course configuration and data processing. The intelligent all-in-one machine 1 is used to run the experimental teaching platform and cognitive feedback analysis program, load teaching scripts, display experimental step guidance, instant prompts, and a cognitive feedback collection interface, and structurally store experimental behavior data and analysis results; the global camera 2 is installed above the intelligent all-in-one machine 1, and its shooting range covers the entire area of ​​the experimental platform, used to collect panoramic video of the student's experimental operation process; the detail camera 3 is installed on one side of the intelligent all-in-one machine 1, aimed at the key operation area, used to collect video of students' fine-grained actions such as reading scales, adjusting knobs, connecting wires, and operating microscopes, to identify whether the actions are standardized and whether the readings are accurate, etc. Detailed behavior: The front-facing camera 4 is positioned above the display screen of the smart all-in-one machine 1 to capture student facial images. Before the experiment begins, the student's identity is bound via facial recognition. Preferably, the global camera 2, the detailed camera 3, and the front-facing camera 4 are all connected to the smart all-in-one machine 1 via wired or wireless means. The captured video streams and snapshot images are sent to the smart all-in-one machine 1 in real time for processing. The experimental operation behavior features extracted by algorithms such as action recognition, trajectory analysis, and reading recognition are associated with the corresponding sub-knowledge unit identifier and timestamp and written into the learning behavior data record. In specific implementation, each step of the above-mentioned teaching process optimization method can be executed by the smart all-in-one machine 1. The learning behavior data can include experimental behavior data collected and processed by the global camera 2, the detailed camera 3, and the front-facing camera 4, or it can include regular behavior data generated by online classroom systems, learning platforms, or classroom feedback systems. This invention does not limit the specific source of the learning behavior data. As long as it can reflect the student's learning and experimental operation status at the sub-knowledge unit granularity, it can be used as input.

[0051] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A method for optimizing a teaching process based on adaptive cognitive feedback, characterized in that, Includes the following steps: S1. The teaching management system divides the course content into teaching knowledge units and sub-knowledge units based on the preset teaching syllabus. It collects learning behavior data from each collection terminal according to the sub-knowledge units, collects and stores the learning behavior data in a structured manner, calculates the individual baseline parameters of students in each sub-knowledge unit, and updates the corresponding cognitive feedback threshold range according to the individual baseline parameters. S2. Based on learning behavior data and cognitive feedback threshold range, calculate the cognitive feedback index and cognitive uncertainty of students in each sub-knowledge unit. Determine the ratio between information gain and collection cost according to the cognitive uncertainty and the preset collection cost model. When the ratio is greater than the collection trigger threshold, trigger cognitive feedback collection at the current moment and determine the cognitive feedback collection method according to the preset rules. Otherwise, do not trigger cognitive feedback collection. S3. For each teaching knowledge unit, calculate the cognitive dispersion based on the cognitive feedback index of the sub-knowledge unit. When the cognitive dispersion is higher than the first threshold, the teaching knowledge unit is split into multiple sub-knowledge unit teaching segments in the teaching script. When the cognitive dispersion is lower than the second threshold and the relevance condition is met, the teaching knowledge unit and adjacent teaching knowledge units are merged into one teaching segment in the teaching script. S4. During the execution of the teaching script, statistically analyze the distribution of students' cognitive feedback indicators relative to their respective cognitive feedback threshold ranges, and adjust the subsequent teaching scripts based on the distribution results.

2. The method of claim 1, wherein the method further comprises: The teaching knowledge units correspond to the content of physics, chemistry and biology experimental courses. Each teaching knowledge unit corresponds to an experimental project or a closely related type of experimental project in the experimental teaching syllabus. Each sub-knowledge unit corresponds to a micro-skill or sub-ability under the experimental project.

3. The method of claim 2, wherein the method further comprises: The student's operational behaviors collected during the experimental steps are correlated one-to-one with the corresponding sub-knowledge units, and behavioral features are aggregated at the sub-knowledge unit level to complete the collection of learning behavior data by sub-knowledge unit.

4. The method of claim 3, wherein the method further comprises: Within a preset sliding time window, the completion records and interaction behaviors of students in the relevant experimental tasks of the sub-knowledge unit are statistically analyzed. The historical average level and fluctuation index of students in the sub-knowledge unit are calculated respectively. The above historical average level and fluctuation index are combined to form the individual baseline parameter vector of students in the sub-knowledge unit.

5. The teaching process optimization method based on adaptive cognitive feedback according to claim 4, characterized in that, The lower limit of the cognitive feedback threshold range is set as the fluctuation range of the student's historical average level in the sub-knowledge unit minus a preset multiple, and the upper limit is set as the fluctuation range of the historical average level plus a preset multiple. When new experimental behavioral data is written, the indicators for historical average levels and fluctuation ranges are updated online incrementally, and the cognitive feedback threshold range is adjusted accordingly.

6. The method of claim 5, wherein the method further comprises: The experimental learning behavior features corresponding to the sub-knowledge units are standardized and weighted to generate an original performance score. Based on the relative position of the original performance score in the historical individual baseline and the corresponding cognitive feedback threshold range, a signed cognitive feedback index is calculated.

7. The method of claim 6, wherein the method further comprises: The variance of students’ cognitive feedback indicators at least once in sub-knowledge units is statistically analyzed, or a probabilistic model is used to make a posterior estimate of mastery status, and one of variance, confidence interval width or distribution entropy is used as cognitive uncertainty.

8. The method of claim 7, wherein the method further comprises: The data collection cost model takes the number of cognitive feedback collections triggered by the student in the current time window, the time interval since the last cognitive feedback collection, the average collection time, and the remaining time in the current experimental stage as inputs. The data collection cost is calculated through a weighted function. When the number of cognitive feedback collections increases, the average collection time increases, or the remaining time in the current experimental stage decreases, the data collection cost increases accordingly. The ratio of information gain to collection cost is defined. When the ratio is greater than the preset collection trigger threshold, the teaching management system triggers cognitive feedback collection for students at the current moment; when the ratio is less than or equal to the trigger threshold, no new collection is triggered. Based on the sub-knowledge unit or cognitive dimension where the cognitive uncertainty set is located, select the cognitive feedback collection method according to preset rules.

9. The method of claim 8, wherein the method further comprises: For a certain experimental project, the teaching management system first determines the multiple sub-knowledge units contained in the experimental project, and obtains the cognitive feedback indicators corresponding to each sub-knowledge unit at the current moment. The cognitive feedback indicator of each sub-knowledge unit is the average cognitive feedback indicator of the whole class on the sub-knowledge unit, or the indicator obtained by aggregating the cognitive feedback indicators of different students according to preset weights. The teaching management system calculates the cognitive dispersion, which characterizes the degree of cognitive difference within the experimental project, based on the degree of dispersion among the cognitive feedback indicators of all sub-knowledge units under the experimental project. The cognitive dispersion is determined by the variance and standard deviation of the cognitive feedback indicators of each sub-knowledge unit, or by the weighted variance with the importance of each sub-knowledge unit as the weight. A first threshold is set. When the cognitive dispersion is greater than the first threshold, the corresponding experimental project is broken down into multiple sub-knowledge unit teaching links in the teaching script. A second threshold is set, and a correlation condition is introduced: the cognitive feedback vectors of the experimental project and adjacent experimental projects are tracked on multiple time slices, and the correlation coefficient between the corresponding sub-knowledge units is calculated, or the overall correlation coefficient of the cognitive feedback vectors of the experimental project and adjacent experimental projects are calculated on multiple stages. When the cognitive dispersion is less than the second threshold and the above correlation coefficient is greater than the preset threshold, the two experimental projects are merged into a comprehensive experimental segment in the teaching script.

10. The method of claim 9, wherein the method further comprises: For the experimental phase that is currently being executed or has ended, obtain the cognitive feedback indicators and corresponding threshold ranges of all students in the current class on the experimental projects or sub-knowledge units associated with the experimental phase, and count the proportions of students that fall below the lower threshold, are between the lower and upper thresholds, and are above the upper threshold, respectively, and record them as the first proportion, the second proportion, and the third proportion. When the first proportion exceeds the preset difficulty proportion threshold, remedial script units are automatically inserted or enhanced for the knowledge units in the subsequent unexecuted experimental scripts, and the progress speed of the relevant links is reduced; when the third proportion exceeds the preset ease proportion threshold, repetitive basic operation exercises are deleted and challenging comprehensive experimental tasks are added; when the first proportion and the third proportion both exceed their respective difference thresholds, and the second proportion is small, tiered experimental teaching script units are inserted in the subsequent scripts.

Images