Self-adaptive education system and method based on artificial intelligence

Abstract

The invention discloses a self-adaptive education system and method based on artificial intelligence, relates to the technical field of self-adaptive education, and solves the technical problem that the self-adaptive education system is difficult to dynamically adjust teaching duration according to learning habits and learning time of learners in teaching task setting. According to the method, the learning amount, the non-learning time, the predicted learning time and the learning rate of a learner are acquired; inputting the non-learning time and the predicted learning time into a time planning module to obtain preliminary planning time of the learner, and adjusting the preliminary planning time through the learning rate to obtain second planning time; acquiring the learned time of the learner in the week in real time, and dynamically adjusting the remaining second planning time based on the learned time; the learning duration can be flexibly adjusted according to different conditions of learners, and individual requirements can be better met.

Description

Technical Field

[0001] The present invention belongs to the field of smart education and relates to adaptive education technology, specifically an adaptive education system and method based on artificial intelligence. Background Art

[0002] Traditional education systems usually adopt a unified teaching schedule, which is difficult to meet the learning needs of different students. Each student's learning ability, self-control and study time vary greatly, and standardized teaching may cause some students to fall behind. For example, online learners similar to adult education usually have work and family responsibilities, and generally learn online through the Internet. This will cause learners with poor self-control to be unable to complete learning tasks on time and effectively. Therefore, the adaptive system can flexibly adjust the class time according to their schedule and provide them with opportunities for fragmented learning.

[0003] At present, most AI-based adaptive education systems have difficulty dynamically adjusting the teaching time according to learners' learning habits and learning time when setting teaching tasks. They only set fixed learning time or let learners arrange their own time for autonomous learning. Setting fixed learning time may conflict with learners' available learning time, which will lead to learners not learning or reduced learning efficiency; letting learners arrange their own time for autonomous learning may cause learners with poor deterrence to fail to learn effectively, reducing learning results.

[0004] Therefore, the present invention discloses an adaptive education system and method based on artificial intelligence to solve the above technical problems. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes an adaptive education system and method based on artificial intelligence, which is used to solve the technical problem that it is difficult to dynamically adjust the teaching time according to the learner's learning habits and learning time when setting teaching tasks. The present invention obtains the learner's preliminary planning time by inputting non-learning time and expected learning time into a time planning module, and adjusts the preliminary planning time by the learning rate to obtain the second planning time; the learner's learning time this week is obtained in real time, and the remaining second planning time is dynamically adjusted based on the learned time to solve the above problem.

[0006] To achieve the above-mentioned object, the first aspect of the present invention provides an adaptive education system based on artificial intelligence, comprising: an intelligent analysis module, and a data collection module, a duration adjustment module and a database connected thereto;

[0007] The data collection module is used to obtain the learner's target data, wherein the target data includes learning amount, non-learning time, expected learning time and learning rate;

[0008] The intelligent analysis module is used to input the non-learning time and the estimated learning time into the time planning module to obtain the learner's preliminary planning time, and adjust the preliminary planning time by the learning rate to obtain the second planning time; wherein the planning time is the learning time set by the system for the learner, and the time planning module is obtained through the training of the artificial intelligence model;

[0009] The duration adjustment module is used to obtain the learner's learning time this week in real time, and dynamically adjust the remaining second planned time based on the learning time.

[0010] Preferably, the acquiring of the learner's target data includes:

[0011] Extract the study time required by the learner in a week, and mark the study time as the study amount; obtain the weekly non-study time filled in by the learner into the system from the database, and mark the time period of the non-study time as non-study time; obtain the weekly expected study time period filled in by the learner into the system from the database, and mark the expected study time period as the expected study time;

[0012] After the learner's authorization, the camera of the online learning device is used to obtain the learner's actual learning time within the planned time set by the system, and the learner's learning rate during the planned time is obtained by dividing the actual learning time by the duration of this planned time; among which, the planned time is the learning time set by the system for the learner, and the actual learning time is the time the learner appears in the camera within the planned time.

[0013] Preferably, the time planning module is obtained through artificial intelligence model training, including:

[0014] Extract non-learning time, expected learning time, learning amount and preliminary planning time from the reference data from the database; integrate the non-learning time, expected learning time, learning amount and preliminary planning time into a number of training data and test data, use the training data to train the artificial intelligence model, use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally obtain a time planning module with non-learning time, expected learning time and learning amount as input and preliminary planning time as output; wherein the reference data includes a number of non-learning time, expected learning time and learning amount, and the preliminary planning time manually set according to the non-learning time, expected learning time and learning amount; the artificial intelligence model includes a BP neural network and an RBF neural network.

[0015] Specifically, the trained artificial intelligence model is tested using the test data, and the specific steps for adjusting the artificial intelligence model according to the test results are as follows:

[0016] The non-learning time, expected learning time and learning amount in the test data are input into the trained artificial intelligence model to obtain the corresponding preliminary planning time, and the corresponding preliminary planning time is compared with the corresponding preliminary planning time in the test data. When the range overlap of the two is within the threshold, which is obtained based on experience, no adjustment is required, and the next set of test data is tested; if the range overlap of the two is not within the threshold, the corresponding parameters are adjusted until the range overlap of the two is within the threshold, and then the next set of test data is tested. When the number of test data with the overlap within the threshold for all test data accounts for 95% or more of the total number, a time planning module is obtained with the input of non-learning time, expected learning time and learning amount, and the output of preliminary planning time.

[0017] Preferably, the adjusting the preliminary planning time by the learning rate to obtain the second planning time includes:

[0018] Mark the non-adjacent time periods in the preliminary planning time as preliminary planning time periods BTi, extract several learning rates of the learners, analyze the learning rates to obtain the learning rates of each time period, and obtain the time period learning rates DLi corresponding to each preliminary planning time period BTi; where i is the serial number of the preliminary planning time period;

[0019] The duration of each preliminary planning period BTi is added to obtain the preliminary total duration BC, and the learning rate DLi of each period is added to obtain the preliminary total learning rate BL. Based on the formula BTCi=BC×BTi / BL, the preliminary adjustment duration BTCi of the preliminary planning period BTi corresponding to the learning rate DLi of each period is obtained;

[0020] The initial time of each preliminary planning period BTi is added with the preliminary adjustment time BTCi to obtain the second planning period RTi, and the second planning period RTi is subjected to anomaly detection to obtain the second planning time.

[0021] Preferably, analyzing the learning rate to obtain the learning rate for each time period includes:

[0022] A week is planned into several fixed time periods, and the learning rates are integrated into several learning rate groups according to the fixed time periods, and the learning rate groups are extracted from the beginning to the end according to the time, and the variance of the extracted learning rate groups is obtained; it is determined whether the variance of the learning rate group is less than the corresponding variance definition threshold; if yes, the learning rates in the learning rate group are averaged to obtain the average value; if no, the learning rate in the learning rate group that is the most different from the mode of the learning rate is removed, and the variance is re-determined until the variance of the learning rate group is less than the corresponding variance definition threshold, and the learning rates in the learning rate group are averaged to obtain the average value; wherein the variance definition threshold is obtained through experience;

[0023] The average value of each learning rate group is marked as the period learning rate of the corresponding fixed period.

[0024] Preferably, the step of obtaining the time period learning rate corresponding to each preliminary planning time period BTi includes:

[0025] Extract the preliminary planning time periods BTi in sequence, and determine whether the preliminary planning time period BTi is in a fixed time period; if yes, mark the fixed time period as a target time period, and mark the time period learning rate of the target time period as the time period learning rate corresponding to the current preliminary planning time period BTi; if no, mark the fixed time period with the highest duration among several fixed time periods of the preliminary planning time period BTi as a target time period, and mark the time period learning rate of the target time period as the time period learning rate corresponding to the current preliminary planning time period BTi.

[0026] Preferably, the step of obtaining the second planned time by performing anomaly detection on the second planned time period RTi includes:

[0027] Extract the second planning time periods RTi in sequence, and determine whether there is a time point in the extracted second planning time period RTi that is in the non-learning time; if yes, mark the time point when the second planning time period RTi is not in the non-learning time as the target time point of the current second planning time period RTi, and mark the time range included in the target time point as the current second planning time period RTi; if no, do not perform anomaly detection on the extracted second planning time period RTi.

[0028] Preferably, the real-time acquisition of the learner's learning time this week includes:

[0029] The time that the learner has learned before the current time this week is added to obtain the learner's learning time this week.

[0030] Preferably, the dynamically adjusting the remaining second planned time based on the learned time includes:

[0031] A1: Obtain the total learning time of the learning time and the remaining total time of the remaining second planned time, and determine whether the remaining total time is less than the value of the learning amount minus the total learning time; if yes, jump to A2; if no, do not make any adjustment;

[0032] A2: Mark the value of the learning amount minus the total learning time as the real remaining time, and subtract the remaining total time from the real remaining time to obtain the time to be compensated; extract the second planning period RTi without abnormal detection in the remaining second planning time, and mark the second planning period RTi as the compensation period DTj; extract the time difference TCj between the end time of each compensation period DTj and the most recent non-learning time, and obtain the proportion TLj of the time difference TCj corresponding to each compensation period DTj based on the formula TLj=TCj / ∑(TCj); compensate the time to be compensated into each compensation period DTj according to the proportion TLj corresponding to each compensation period DTj; where j is the serial number of the compensation period DTj; ∑ is the summation symbol, and the summation range is [1,j];

[0033] A3: Determine whether there is a time point of the compensation period DTj in each compensation period DTj that is in the non-learning time; if yes, mark the compensation period DTj as a secondary adjustment period and jump to A4; if no, do nothing;

[0034] A4: Extract the secondary adjustment periods in sequence, mark the time points of the extracted secondary adjustment periods that are not in the non-learning time as expected time points, and mark the time range included in the expected time points as the compensation period DTj corresponding to the current secondary adjustment period;

[0035] A5: The duration of the secondary adjustment period in the non-learning time is added to obtain the estimated non-learning time, and the estimated non-learning time is arranged to the fault tolerance time set each week for learning; the fault tolerance time is the time manually set for arranging the estimated non-learning time.

[0036] A second aspect of the present invention provides an artificial intelligence-based adaptive education method, comprising the following steps:

[0037] S1: Obtain learners’ target data;

[0038] S2: inputting the non-learning time and the estimated learning time into the time planning module to obtain the learner's preliminary planning time, and adjusting the preliminary planning time by the learning rate to obtain the second planning time;

[0039] S3: Obtain the learner’s learning time this week in real time, and dynamically adjust the remaining second planned time based on the learning time.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. The present invention obtains the learner's preliminary planned time by inputting non-learning time and estimated learning time into a time planning module, and obtains the second planned time by adjusting the preliminary planned time through the learning rate; the learner's learning time this week is obtained in real time, and the remaining second planned time is dynamically adjusted based on the learned time, thereby solving the technical problem that it is difficult to dynamically adjust the teaching time according to the learner's learning habits and learning time during the setting of teaching tasks; the present invention can flexibly adjust the learning time according to the different situations of the learners to better meet the individual needs.

[0042] 2. The present invention incorporates the learning rate into the adjustment process of the initial planning time, achieving accurate adaptation to the learner's personalized needs. Through the careful optimization of the initial planning time, the generated second planning time is more in line with the learner's learning habits. This adjustment strategy can be intelligently adjusted according to the learner's learning efficiency in different time periods, ensuring that more learning content is arranged during the period when the learner is energetic and has strong absorption ability, thereby effectively improving learning efficiency and improving the overall quality of learning results. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 It is a schematic diagram of the operation steps of the present invention;

[0045] Figure 2 It is a schematic diagram of the system module of the present invention;

[0046] Figure 3 It is a schematic diagram of the operation steps of dynamically adjusting the second planning time according to the present invention. DETAILED DESCRIPTION

[0047] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] See also Figure 1-Figure 2 , the first aspect of the present invention provides an adaptive education system based on artificial intelligence, including: an intelligent analysis module, and a data collection module, a duration adjustment module and a database connected thereto;

[0049] Data collection module: used to obtain learners' target data; the target data includes learning amount, non-learning time, expected learning time and learning rate;

[0050] Intelligent analysis module: used to input non-learning time and estimated learning time into the time planning module to obtain the learner's preliminary planned time, and adjust the preliminary planned time through the learning rate to obtain the second planned time; wherein the planned time is the learning time set by the system for the learner, and the time planning module is obtained through artificial intelligence model training;

[0051] Duration adjustment module: used to obtain the learner's learning time this week in real time, and dynamically adjust the remaining second planned time based on the learning time.

[0052] The target data of learners obtained in this application include:

[0053] Extract the study time required by the learner in a week, and mark the study time as the study amount; obtain the weekly non-study time filled in by the learner into the system from the database, and mark the time period of the non-study time as non-study time; obtain the weekly expected study time period filled in by the learner into the system from the database, and mark the expected study time period as the expected study time;

[0054] After the learner's authorization, the camera of the online learning device is used to obtain the learner's actual learning time within the planned time set by the system, and the learner's learning rate during the planned time is obtained by dividing the actual learning time by the duration of this planned time; among which, the planned time is the learning time set by the system for the learner, and the actual learning time is the time the learner appears in the camera within the planned time.

[0055] It should be noted that the time when you cannot study includes working time, rest time, and social activity time, etc.

[0056] It should be noted that the types of planned time in the actual learning time of the learner within the planned time set by the system obtained through the camera of the online learning device can be preliminary planned time, second planned time, preliminary planned time and second planned time.

[0057] The time planning module in this application is obtained through artificial intelligence model training, including:

[0058] Extract non-learning time, expected learning time, learning amount and preliminary planning time from the reference data from the database; integrate the non-learning time, expected learning time, learning amount and preliminary planning time into a number of training data and test data, use the training data to train the artificial intelligence model, use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally, a time planning module is obtained with non-learning time, expected learning time and learning amount as input and preliminary planning time as output; wherein the reference data includes a number of non-learning time, expected learning time and learning amount, and the preliminary planning time manually set according to the non-learning time, expected learning time and learning amount; the artificial intelligence model includes BP neural network and RBF neural network.

[0059] Specifically, the trained artificial intelligence model is tested using the test data, and the specific steps for adjusting the artificial intelligence model according to the test results are as follows:

[0060] The non-learning time, expected learning time and learning amount in the test data are input into the trained artificial intelligence model to obtain the corresponding preliminary planning time, and the corresponding preliminary planning time is compared with the corresponding preliminary planning time in the test data. When the range overlap of the two is within the threshold, which is obtained based on experience, no adjustment is required, and the next set of test data is tested; if the range overlap of the two is not within the threshold, the corresponding parameters are adjusted until the range overlap of the two is within the threshold, and then the next set of test data is tested. When the number of test data with the overlap within the threshold for all test data accounts for 95% or more of the total number, a time planning module is obtained with the input of non-learning time, expected learning time and learning amount, and the output of preliminary planning time.

[0061] In this application, the initial planning time is adjusted by the learning rate to obtain the second planning time, including:

[0062] Mark the non-adjacent time periods in the preliminary planning time as preliminary planning time periods BTi, extract several learning rates of the learners, analyze the learning rates to obtain the learning rates of each time period, and obtain the time period learning rates DLi corresponding to each preliminary planning time period BTi; where i is the serial number of the preliminary planning time period;

[0063] The duration of each preliminary planning period BTi is added to obtain the preliminary total duration BC, and the learning rate DLi of each period is added to obtain the preliminary total learning rate BL. Based on the formula BTCi=BC×BTi / BL, the preliminary adjustment duration BTCi of the preliminary planning period BTi corresponding to the learning rate DLi of each period is obtained;

[0064] The initial time of each preliminary planning period BTi is added with the preliminary adjustment time BTCi to obtain the second planning period RTi, and the second planning period RTi is subjected to anomaly detection to obtain the second planning time.

[0065] It is worth noting that the present invention introduces a learning rate when adjusting the initial planned time. The initial planned time is adjusted to suit the learner's individual needs through the learning rate. This enables the second planned time to adapt to the learner's learning habits and increase the amount of learning during the time period when the learner's learning efficiency is high. This enables the learner to learn the course better and enhance learning outcomes.

[0066] It should be noted that the second planning period RTi can be obtained by adding the initial adjustment time BTCi to the initial time of each preliminary planning period BTi, which can be illustrated as follows:

[0067] If the initial adjustment time BT2 is Monday 12:00-12:30, therefore, the initial time of the initial adjustment time BT2 is Monday 12:00, at this time the initial adjustment time BTC2 = 20min, then add the initial adjustment time BTC2 = 20min on the basis of Monday 12:00, and the second planning time period RT2 is Monday 12:00-12:20.

[0068] In this application, the learning rate is analyzed to obtain the learning rate of each period, including:

[0069] Plan a week into several fixed time periods, integrate the learning rates into several learning rate groups according to the fixed time periods, extract the learning rate groups from the beginning to the end according to the time, and obtain the variance of the extracted learning rate groups; determine whether the variance of the learning rate group is less than the corresponding variance definition threshold; if yes, calculate the average value of the learning rates in the learning rate group to obtain the average value; if no, remove the learning rate in the learning rate group that is the most different from the mode of the learning rate, and re-judge the variance until the variance of the learning rate group is less than the corresponding variance definition threshold, and then calculate the average value of the learning rates in the learning rate group to obtain the average value; wherein the variance definition threshold is obtained through experience;

[0070] The average value of each learning rate group is marked as the period learning rate of the corresponding fixed period.

[0071] It is worth noting that the learning rate for each time period calculated by the present invention is obtained by judging each learning rate group through variance. After variance judgment, the fluctuation value or abnormal value in each learning rate group can be removed. The average value calculated in this way can effectively represent the learning rate of the learner in the current time period, providing data support for subsequent analysis.

[0072] It should be noted that planning a week into several time periods is achieved through manual planning.

[0073] In this application, the period learning rate corresponding to each preliminary planning period BTi is obtained, including:

[0074] Extract the preliminary planning period BTi in sequence, and determine whether the preliminary planning period BTi is in a fixed period; if so, mark the fixed period as the target period, and mark the period learning rate of the target period as the period learning rate corresponding to the current preliminary planning period BTi; if not, mark the fixed period with the highest duration of the preliminary planning period BTi among several fixed periods as the target period, and mark the period learning rate of the target period as the period learning rate corresponding to the current preliminary planning period BTi.

[0075] It should be noted that marking the fixed period with the highest duration ratio among several fixed periods in the preliminary planning period BTi as the target period can be illustrated as follows:

[0076] If fixed time period one is Monday 14:00-18:00, and fixed time period two is Monday 18:00-20:00; the preliminary planned time period BT6 is Monday 17:30-19:30; at this time, the preliminary planned time period BT6 has duration in both fixed time period one and fixed time period two, and the duration of the preliminary planned time period BT6 in fixed time period one accounts for 30 / 120=0.25, and the duration of the preliminary planned time period BT6 in fixed time period two accounts for 90 / 120=0.75; therefore, the duration of the preliminary planned time period BT6 in fixed time period two accounts for the highest proportion, so fixed time period two is marked as the target time period.

[0077] In this application, the second planning period RTi is obtained by anomaly detection to obtain the second planning time, including:

[0078] Extract the second planning time periods RTi in sequence, and determine whether there is a time point in the extracted second planning time period RTi that is in the non-learning time; if yes, mark the time point when the second planning time period RTi is not in the non-learning time as the target time point of the current second planning time period RTi, and mark the time range included in the target time point as the current second planning time period RTi; if no, do not perform anomaly detection on the extracted second planning time period RTi.

[0079] It should be noted that the second planning period RT20 is 21:00-22:30 on Thursday, and 22:00-24:00 on Thursday is non-learning time. Because 22:00-22:30 on Thursday is non-learning time, the time points included in 21:00-22:00 on Thursday are marked as target time points, and the time range included in the target time points is marked as the current second planning period RT20, and the second planning period RT20 is 21:00-22:00 on Thursday.

[0080] It should be noted that the time point can be understood as the time point per second.

[0081] This application obtains the learner's learning time this week in real time, including:

[0082] The time that the learner has learned before the current time this week is added to obtain the learner's learning time this week.

[0083] See also Figure 3 In this application, the remaining second planning time is dynamically adjusted based on the learned time, including:

[0084] A1: Get the total learning time of the learning time and the remaining total time of the remaining second planned time, and determine whether the remaining total time is less than the value of the learning amount minus the total learning time; if yes, jump to A2; if no, do not make any adjustment;

[0085] A2: Mark the value of the learning amount minus the total learning time as the real remaining time, and subtract the remaining total time from the real remaining time to obtain the time to be compensated; extract the second planning period RTi without abnormal detection in the remaining second planning time, and mark the second planning period RTi as the compensation period DTj; extract the time difference TCj between the end time of each compensation period DTj and the most recent non-learning time, and obtain the proportion TLj of the time difference TCj corresponding to each compensation period DTj based on the formula TLj=TCj / ∑(TCj); compensate the time to be compensated into each compensation period DTj according to the proportion TLj corresponding to each compensation period DTj; where j is the serial number of the compensation period DTj; ∑ is the summation symbol, and the summation range is [1,j];

[0086] A3: Determine whether there is a time point of the compensation period DTj in each compensation period DTj that is in the non-learning time; if yes, mark the compensation period DTj as a secondary adjustment period and jump to A4; if no, do nothing;

[0087] A4: Extract the secondary adjustment periods in sequence, mark the time points of the extracted secondary adjustment periods that are not in the non-learning time as expected time points, and mark the time range included in the expected time points as the compensation period DTj corresponding to the current secondary adjustment period;

[0088] A5: The duration of the secondary adjustment period in the non-learning time is added to obtain the estimated non-learning time, and the estimated non-learning time is arranged to the fault tolerance time set each week for learning; the fault tolerance time is the time manually set for arranging the estimated non-learning time.

[0089] It is worth noting that the present invention dynamically adjusts the remaining second planned time according to the learned time, and is used to adjust each time period in the remaining second planned time when the learned time is less, thereby avoiding the situation where the total duration of the second planned time is less than the learning amount of this week, thereby improving the stability and adaptability of the system.

[0090] It should be noted that, in the time difference TCj extracted between the end time of each compensation period DTj and the latest non-learning time, the latest non-learning time is the non-learning time that is closest to and after the end time of each compensation period DTj.

[0091] It should be noted that the time to be compensated is compensated into each compensation period DTj according to the ratio TLj corresponding to each compensation period DTj. An example is as follows: if the compensation period DT6 is 21:00-22:00 on Tuesday, the time to be compensated is 10 minutes calculated according to the ratio TL6=0.06 corresponding to each compensation period DT6, then 10 minutes is added to 21:00-22:00 on Tuesday, and the compensation period DT6 is 21:00-22:10 on Tuesday.

[0092] The second aspect of the present invention provides an artificial intelligence-based adaptive education method, comprising the following steps:

[0093] S1: Obtain learners’ target data;

[0094] S2: inputting the non-learning time and the estimated learning time into the time planning module to obtain the learner's preliminary planning time, and adjusting the preliminary planning time by the learning rate to obtain the second planning time;

[0095] S3: Obtain the learner’s learning time this week in real time, and dynamically adjust the remaining second planned time based on the learning time.

[0096] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0097] Working principle of the present invention:

[0098] The learner's learning amount, non-learning time, expected learning time and learning rate are obtained; the non-learning time and expected learning time are input into the time planning module to obtain the learner's preliminary planning time; the preliminary planning time is adjusted by the learning rate to obtain the second planning time. In this step, the preliminary planning time is adapted to the learner's personalized adjustment, so that the obtained second planning time can adapt to the learner's learning habits; the learner's learning time this week is obtained in real time, and the remaining second planning time is dynamically adjusted based on the learned time. In this step, the remaining second planning time is dynamically adjusted by the learned time, which is used to adjust the various time periods in the remaining second planning time when the learned time is less, so as to avoid the situation where the total duration of the second planning time is less than the learning amount of this week, thereby improving the stability and adaptability of the system.

[0099] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An adaptive education system based on artificial intelligence, characterized in that: include: Intelligent analysis module, and the data collection module, duration adjustment module and database connected thereto; The data collection module is used to obtain the learner's target data, wherein the target data includes learning amount, non-learning time, expected learning time and learning rate; The intelligent analysis module is used to input the non-learning time and the estimated learning time into the time planning module to obtain the learner's preliminary planning time, and adjust the preliminary planning time by the learning rate to obtain the second planning time; wherein the planning time is the learning time set by the system for the learner, and the time planning module is obtained through the training of the artificial intelligence model; The duration adjustment module is used to obtain the learner's learning time this week in real time, and dynamically adjust the remaining second planned time based on the learning time.

2. The adaptive education system based on artificial intelligence according to claim 1, characterized in that: The acquiring of the learner's target data includes: Extract the study time required by the learner in a week, and mark the study time as the study amount; obtain the weekly non-study time filled in by the learner into the system from the database, and mark the time period of the non-study time as non-study time; obtain the weekly expected study time period filled in by the learner into the system from the database, and mark the expected study time period as the expected study time; After the learner's authorization, the camera of the online learning device is used to obtain the learner's actual learning time within the planned time set by the system, and the learner's learning rate during the planned time is obtained by dividing the actual learning time by the duration of this planned time; among which, the planned time is the learning time set by the system for the learner, and the actual learning time is the time the learner appears in the camera within the planned time.

3. The adaptive education system based on artificial intelligence according to claim 1, characterized in that: The time planning module is obtained through artificial intelligence model training, including: Extract non-learning time, expected learning time, learning amount and preliminary planning time from the reference data from the database; integrate the non-learning time, expected learning time, learning amount and preliminary planning time into a number of training data and test data, use the training data to train the artificial intelligence model, use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model according to the test results; finally obtain a time planning module with non-learning time, expected learning time and learning amount as input and preliminary planning time as output; wherein the reference data includes a number of non-learning time, expected learning time and learning amount, and the preliminary planning time manually set according to the non-learning time, expected learning time and learning amount; the artificial intelligence model includes a BP neural network and an RBF neural network.

4. The adaptive education system based on artificial intelligence according to claim 1, characterized in that: The step of adjusting the initial planning time by the learning rate to obtain the second planning time includes: Mark the non-adjacent time periods in the preliminary planning time as preliminary planning time periods BTi, extract several learning rates of the learners, analyze the learning rates to obtain the learning rates of each time period, and obtain the time period learning rates DLi corresponding to each preliminary planning time period BTi; where i is the serial number of the preliminary planning time period; The duration of each preliminary planning period BTi is added to obtain the preliminary total duration BC, and the learning rate DLi of each period is added to obtain the preliminary total learning rate BL. Based on the formula BTCi=BC×BTi / BL, the preliminary adjustment duration BTCi of the preliminary planning period BTi corresponding to the learning rate DLi of each period is obtained; The initial time of each preliminary planning period BTi is added with the preliminary adjustment time BTCi to obtain the second planning period RTi, and the second planning period RTi is subjected to anomaly detection to obtain the second planning time.

5. The adaptive education system based on artificial intelligence according to claim 4, characterized in that: The learning rate is analyzed to obtain the learning rate of each period, including: Plan a week into several fixed time periods, integrate the learning rates into several learning rate groups according to the fixed time periods, extract the learning rate groups from the beginning to the end according to the time, and obtain the variance of the extracted learning rate groups; determine whether the variance of the learning rate group is less than the corresponding variance definition threshold; if yes, calculate the average value of the learning rates in the learning rate group to obtain the average value; if no, remove the learning rate in the learning rate group that is the most different from the mode of the learning rate, and re-determine the variance until the variance of the learning rate group is less than the corresponding variance definition threshold, and then calculate the average value of the learning rates in the learning rate group to obtain the average value; The average value of each learning rate group is marked as the period learning rate of the corresponding fixed period.

6. The adaptive education system based on artificial intelligence according to claim 5, characterized in that: The step of obtaining the time period learning rate corresponding to each preliminary planning time period BTi includes: Extract the preliminary planning time periods BTi in sequence, and determine whether the preliminary planning time period BTi is in a fixed time period; if yes, mark the fixed time period as a target time period, and mark the time period learning rate of the target time period as the time period learning rate corresponding to the current preliminary planning time period BTi; if no, mark the fixed time period with the highest duration among several fixed time periods of the preliminary planning time period BTi as a target time period, and mark the time period learning rate of the target time period as the time period learning rate corresponding to the current preliminary planning time period BTi.

7. The adaptive education system based on artificial intelligence according to claim 4, characterized in that: The step of obtaining a second planned time by performing anomaly detection on the second planned time period RTi includes: Extract the second planning time periods RTi in sequence, and determine whether there is a time point in the extracted second planning time period RTi that is in the non-learning time; if yes, mark the time point when the second planning time period RTi is not in the non-learning time as the target time point of the current second planning time period RTi, and mark the time range included in the target time point as the current second planning time period RTi; if no, do not perform anomaly detection on the extracted second planning time period RTi.

8. The adaptive education system based on artificial intelligence according to claim 1, characterized in that: The real-time acquisition of the learner's learning time this week includes: The time that the learner has learned before the current time this week is added to obtain the learner's learning time this week.

9. The adaptive education system based on artificial intelligence according to claim 4, characterized in that: The dynamically adjusting the remaining second planned time based on the learned time includes: A1: Obtain the total learning time of the learning time and the remaining total time of the remaining second planned time, and determine whether the remaining total time is less than the value of the learning amount minus the total learning time; if yes, jump to A2; if no, do not make any adjustment; A2: Mark the value of the learning amount minus the total learning time as the real remaining time, and subtract the remaining total time from the real remaining time to obtain the time to be compensated; extract the second planning period RTi without abnormal detection in the remaining second planning time, and mark the second planning period RTi as the compensation period DTj; extract the time difference TCj between the end time of each compensation period DTj and the most recent non-learning time, and obtain the proportion TLj of the time difference TCj corresponding to each compensation period DTj based on the formula TLj=TCj / ∑(TCj); compensate the time to be compensated into each compensation period DTj according to the proportion TLj corresponding to each compensation period DTj; where j is the serial number of the compensation period DTj; ∑ is the summation symbol, and the summation range is [1,j]; A3: Determine whether there is a time point of the compensation period DTj in each compensation period DTj that is in the non-learning time; if yes, mark the compensation period DTj as a secondary adjustment period and jump to A4; if no, do nothing; A4: Extract the secondary adjustment periods in sequence, mark the time points of the extracted secondary adjustment periods that are not in the non-learning time as expected time points, and mark the time range included in the expected time points as the compensation period DTj corresponding to the current secondary adjustment period; A5: The duration of the secondary adjustment period in the non-learning time is added to obtain the estimated non-learning time, and the estimated non-learning time is arranged to the fault tolerance time set each week for learning; the fault tolerance time is the time manually set for arranging the estimated non-learning time.

10. An artificial intelligence-based adaptive education method, based on the artificial intelligence-based adaptive education system according to any one of claims 1 to 9, characterized in that: S1: Obtain learners’ target data; S2: inputting the non-learning time and the estimated learning time into the time planning module to obtain the learner's preliminary planning time, and adjusting the preliminary planning time by the learning rate to obtain the second planning time; S3: Obtain the learner’s learning time this week in real time, and dynamically adjust the remaining second planned time based on the learning time.

Images