CN113191438B

CN113191438B - Learning style recognition model training and recognition method, device, equipment and medium

Info

Publication number: CN113191438B
Application number: CN202110500644.7A
Authority: CN
Inventors: 张冰雪; 柴成亮; 史洋; 侯龙锋
Original assignee: Ah Ah Shanghai Technology Co ltd
Current assignee: Ah Ah Shanghai Technology Co ltd
Priority date: 2021-05-08
Filing date: 2021-05-08
Publication date: 2023-08-15
Anticipated expiration: 2041-05-08
Also published as: CN113191438A

Abstract

The embodiment of the application discloses a learning style recognition model training and recognition method, a device, equipment and a medium. The method comprises the following steps: displaying a first preset number of test questions to a target object so as to test the target object; the test questions are Rayleigh advanced reasoning test questions; acquiring electroencephalogram characteristic data of a target object in a test period, and taking the electroencephalogram characteristic data as sample electroencephalogram characteristic data; determining training data according to the sample electroencephalogram characteristic data and the learning style label of the target object; and carrying out model training according to the training data to obtain a learning style recognition model. According to the method and the device for identifying the learning style, the problem that the identification result is inaccurate due to the fact that subjectivity of the target object in test data acquired in the current learning style identification is strong is solved, the learning style expression of the target object is accurately stimulated through test questions, the learning style characteristics of the target object are accurately reflected through electroencephalogram characteristic data, accuracy of a learning style identification model is improved, and accuracy of learning style identification is further improved.

Description

Learning style recognition model training and recognition method, device, equipment and medium

Technical Field

The embodiment of the application relates to the technical field of computers, in particular to a learning style recognition model training and recognition method, device, equipment and medium.

Background

Learning Style (Learning Style) is a habitual Learning manner and tendency of a learner in a Learning process, is a relatively stable Learning manner and preference gradually formed in a long-term Learning activity, and is an important factor reflecting individual differences of the learner. The best teaching should first judge the individual learning style and tailor the teaching based thereon. Many researchers have applied learning style models in adaptive learning systems. The learning style can enable the self-adaptive system to better realize individualization of learning resources or learning activities so as to meet different learning demands of learners.

Currently, the learning styles are mainly identified by explicit identification and implicit identification, wherein the explicit identification is to judge the learning style of a learner by utilizing a learning style scale to survey each item score in an ILS questionnaire through a scoring table (Felder-Silverman Index of Learning Styles, ILS). However, the learner has difficulty in understanding the learning style and its concept, so it may not be possible to accurately fill out the answer to the questionnaire, and the learner may have subjective bias on the test result when answering the questionnaire, thereby affecting the objectivity of the answer result.

Implicit recognition is to indirectly recognize the learning style of the learner by mining and analyzing interactive behavior data (including the number of times the learner clicks a specific button in the system, learning activity time, test results, the number of times the learner posts and reads the posts in a forum, etc.) in the online learning system of the learner without the learner actively participating in filling out a questionnaire. However, the implicit recognition mode has the problem of cold start, a learner needs to acquire a large amount of online learning behavior data to perform more accurate recognition, and the credibility of the data source can have a larger influence on the recognition result.

Disclosure of Invention

The embodiment of the application provides a learning style recognition model training and recognition method, device, equipment and medium, so as to improve the accuracy of learning style recognition.

In one embodiment, the embodiment of the application provides a learning style recognition model training method, which comprises the following steps:

displaying a first preset number of test questions to a target object so as to test the target object; the test questions are Rayleigh advanced reasoning test questions;

acquiring electroencephalogram characteristic data of the target object in a test period, and taking the electroencephalogram characteristic data as sample electroencephalogram characteristic data;

Determining training data according to the sample electroencephalogram characteristic data and the learning style label of the target object;

and training the model according to the training data to obtain a learning style recognition model.

In another embodiment, the present application provides a learning style recognition method, including:

displaying a second preset number of test questions to the object to be identified so as to test the object to be identified; the test questions are Rayleigh advanced reasoning test questions;

acquiring electroencephalogram characteristic data of the object to be identified in a test period, and taking the electroencephalogram characteristic data as target electroencephalogram characteristic data;

inputting the target brain electrical characteristic data into a learning style recognition model to obtain a learning style recognition result of the object to be recognized;

the learning style recognition model is obtained by training based on the learning style recognition model training method provided by any embodiment of the application.

In one embodiment, the embodiment of the application also provides a learning style recognition model training device, which comprises:

the test question display module is used for displaying a first preset number of test questions to a target object so as to test the target object; the test questions are Rayleigh advanced reasoning test questions;

The electroencephalogram characteristic acquisition module is used for acquiring electroencephalogram characteristic data of the target object in the test period and taking the electroencephalogram characteristic data as sample electroencephalogram characteristic data;

the training data determining module is used for determining training data according to the sample electroencephalogram characteristic data and the learning style label of the target object;

and the recognition model training module is used for carrying out model training according to the training data to obtain a learning style recognition model.

In another embodiment, the embodiment of the application further provides a learning style recognition device, which comprises:

the test module is used for displaying a second preset number of test questions to the object to be identified so as to test the object to be identified; the test questions are Rayleigh advanced reasoning test questions;

the data acquisition module is used for acquiring the brain electrical characteristic data of the object to be identified in the test period and taking the brain electrical characteristic data as target brain electrical characteristic data;

the recognition result determining module is used for inputting the target electroencephalogram characteristic data into a learning style recognition model to obtain a learning style recognition result of the object to be recognized;

In yet another embodiment, an embodiment of the present application further provides an electronic device, including: one or more processors;

a memory for storing one or more programs;

when the one or more programs are executed by the one or more processors, the one or more processors implement the learning style recognition model training method provided by any embodiment of the present application, or implement the learning style recognition method provided by any embodiment of the present application.

In yet another embodiment, the present application further provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the learning style recognition model training method provided by any of the embodiments of the present application, or implements the learning style recognition method provided by any of the embodiments of the present application.

One of the above technical solutions has the following technical effects: the method comprises the steps of testing a target object by displaying a first preset number of test questions to the target object; the test questions are Rayleigh advanced reasoning test questions; acquiring electroencephalogram characteristic data of a target object in a test period, and taking the electroencephalogram characteristic data as sample electroencephalogram characteristic data; determining training data according to the sample electroencephalogram characteristic data and the learning style label of the target object; the learning style recognition model is obtained by training the model according to the training data, the problem that the recognition result is inaccurate due to strong subjectivity of the target object in the test data acquired in the current learning style recognition is solved, the learning style expression of the target object is accurately stimulated through the test questions, the learning style characteristics of the target object are accurately reflected through the electroencephalogram characteristic data, the accuracy of the learning style recognition model is improved, and the accuracy of the learning style recognition is further improved.

Drawings

FIG. 1 is a flowchart of a learning style recognition model training method according to an embodiment of the present application;

FIG. 2 is a schematic illustration of a stimulation sequence provided in one embodiment of the present application;

FIG. 3 is a flowchart of a learning style recognition model training method according to another embodiment of the present application;

fig. 4 is a schematic diagram of sample electroencephalogram feature data extraction according to another embodiment of the present application;

fig. 5 is a schematic diagram of sample electroencephalogram feature data segmentation according to another embodiment of the present application;

FIG. 6 is a flowchart of a learning style recognition method according to an embodiment of the present application;

FIG. 7 is an overall flow chart provided by one embodiment of the present application;

FIG. 8 is a schematic diagram of learning style recognition according to an embodiment of the present application;

FIG. 9 is a schematic structural diagram of a learning style recognition model training device according to an embodiment of the present application;

fig. 10 is a schematic structural diagram of a learning style recognition device according to an embodiment of the present application;

fig. 11 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

The application is described in further detail below with reference to the drawings and examples.

Fig. 1 is a flowchart of a learning style recognition model training method according to an embodiment of the present application. The learning style recognition model training method provided by the embodiment of the application can be suitable for training the learning style recognition model for recognizing the learning style. The method may be performed in particular by a learning style recognition model training device, which may be implemented in software and/or hardware, which may be integrated in an electronic device capable of implementing the learning style recognition model training method. Referring to fig. 1, the method in the embodiment of the present application specifically includes:

S110, displaying a first preset number of test questions to a target object so as to test the target object; the test questions are Rayleigh advanced reasoning test questions.

The target object may be a person pre-selected for performing a test to obtain electroencephalogram characteristic data thereof, the selected target object includes target objects having different learning style labels, and a difference in number of target objects corresponding to the different learning style labels is smaller than a preset threshold. By way of example, learning styles may include a sinkage type and an active type, where explicit behavior of a sinkage type target object tends to be deliberate when it appears to face a problem, the problem is considered and examined with sufficient time, the methods of solving various problems are weighed, and then an optimal solution satisfying various conditions is selected therefrom, so that there is less error. The apparent behavior of the active target object is shown to be a quick test hypothesis, and a prompt decision is made according to partial information of the problem or without thorough analysis of the problem, so that the response speed is high, but errors are easy to occur. When the target object is selected, the desizing target object and the active target object are required to be selected, the difference value of the number of the desizing target object and the active target object is smaller than a preset threshold value, and the preset threshold value can be selected according to actual conditions. The number of sinkage type target objects and active type target objects can be selected to be equal.

According to the thinking mode difference of the desizing type target object and the active type target object when processing problems, an accurate and efficient excitation mode needs to be determined so that different performances of the target objects in different learning styles appear, individual differences of learners in learning style processing dimensions are effectively excited, and the designed excitation mode is ensured to generate as few invalid signals as possible in a real excitation process, for example, less time for information processing of the learners, limb actions interfering with the acquisition quality of brain electrical signals of the learners and the like. In the embodiment of the application, a Raven high-level reasoning test question (Raven's Advanced Progressive Matrices, RAPM) is adopted as a test question for testing the target object, so that different performances of different learning style target objects are stimulated through the Raven Wen Gao-level reasoning test question. In stimulating the learning style of the target object, the main task of RAPM is to ask the learner to logically think according to the rule of symbols in the matrix diagram and fill the appropriate options into the vacant positions. RAPM is often used to evaluate learner thinking ability, observability, and ability to solve problems using information required by themselves. RAPM is used as an excitation source, so that a learner can be prompted to conduct various logic thinking. The brain will process information when the learner thinks, so as to excite the processing process of the learner's brain. Too simple test questions shorten the cognitive processing time of the learner, but too difficult test questions lead to tiredness and cognitive load of the learner, thereby affecting the quality of the processing signals excited later. The RAPM is used as the excitation source, so that the overall difficulty is moderate, and the signal quality of the excitation source for exciting the information processing process of the learner on each problem is ensured. In the aspect of reducing the generation of invalid signals, the RAPM test question is a non-text test, so that a learner does not need to read text, the time for non-cognitive processing such as reading a stem can be effectively reduced, and the process of thinking processing and processing of brains is guaranteed to the greatest extent. The RAPM test questions are presented in the form of multiple choice questions, and the learner can complete the answer by directly clicking the corresponding options, so that unnecessary limb actions are reduced to the greatest extent, and the influence of signals such as limb actions on the subsequent electroencephalogram data acquisition stage is reduced.

In the embodiment of the present application, the first preset number may be set according to actual situations, for example, 36, 60, etc., and the specifically selected first preset number is suitable for being able to obtain enough electroencephalogram feature data without causing the target object to feel boring. The first preset number of test questions may be shown to the target object by E-prime2.0 software, and specifically, one test question may be shown to the target object each time, and after the target object answers, the next test question may be shown. The display time of each question can be set, and if the time exceeds the display time and the target object is not answered, the next test question is continuously displayed. When the target object answers the currently displayed test question or the currently displayed test question reaches the display time, the answer of the test question can be directly displayed, and the next question is continuously displayed after the answer of the test question is displayed.

S120, acquiring the brain electrical characteristic data of the target object in the test period, and taking the brain electrical characteristic data as sample brain electrical characteristic data.

For example, the wireless electroencephalograph can be controlled to collect electroencephalogram characteristic data of the target object during the test, and the wireless electroencephalograph can be a head-mounted electroencephalogram device used for collecting the electroencephalogram characteristic data of the target object. The test questions may be divided into a plurality of groups, each group including a plurality of test questions, each group forming a stimulation sequence, as shown in fig. 2, which illustrates a timing diagram of a single group of stimulation sequences. Each set of tests may begin with presentation of a prompt preparation interface to the target object, and the upper limit of the answer time for each test question may be set to 60 seconds, and if the answer time of the target object exceeds 60 seconds, the answer may be presented directly, or after the answer of the target object is completed. And aiming at each stimulation sequence, during the test of the target object, acquiring the brain electrical characteristic data of the target object, and taking the brain electrical characteristic data as the brain electrical characteristic data corresponding to the stimulation sequence.

S130, determining training data according to the sample electroencephalogram characteristic data and the learning style label of the target object.

For example, for each target object, the corresponding learning style label is known in advance, and the learning style label may be obtained by determining the learning style of each target object by using the ILS. In order to enable the target object to intuitively and completely understand the content in the ILS and avoid the target object from generating contradiction emotion to the ILS and generating messy answering, the ILS can be translated and interpreted in detail to indicate the meaning of each topic, and the test willingness of the target object is acquired in advance, so that an accurate answering result is acquired under the voluntary condition of the target object, and the learning style label of the target object is determined according to the answering result.

In the embodiment of the application, training data is determined according to the sample electroencephalogram characteristic data of the target object and the learning style label of the target object so as to train the model according to the training data.

And S140, performing model training according to the training data to obtain a learning style recognition model.

For example, 80% of training data can be selected as a training set, 20% of training data is selected as a testing set, and the recognition model is trained, for example, network models such as a Support Vector Machine (SVM) and a back propagation neural network (BP) can be trained, so as to obtain a learning style recognition model.

Fig. 3 is a flowchart of a learning style recognition model training method according to another embodiment of the present application. The embodiments of the present application are further optimized for the above embodiments, and details not described in detail in the embodiments of the present application are detailed in the above embodiments. Referring to fig. 3, the learning style recognition model training method provided by the embodiment of the application may include:

S210, displaying a first preset number of test questions to a target object so as to test the target object; the test questions are Rayleigh advanced reasoning test questions.

S220, recording the response starting time and the response finishing time of each test question of the target object according to the triggering operation of the target object when each test question is responded.

For example, the time when the test question starts to be displayed may be taken as the answer starting time, and the answer finishing time may be recorded according to a triggering operation when the target object answers the test question, for example, an operation of clicking the answer option. The start time and end time of the response may be recorded by a device that presents the test question to the target object.

In the embodiment of the application, according to the triggering operation of the target object when answering each test question, recording the answering start time and the answering finish time of the target object when answering each test question, including: if the duration of the target object for answering the current test question exceeds the preset duration, the ending time of the preset duration is used as the answering completion time for answering the test question.

For example, a preset duration may be preset, if the duration of the display of the test question reaches the preset duration, but the target object is not yet answered, the display of the test question is ended, and the time when the display of the test question is ended is taken as the answering completion time.

S230, acquiring the brain electrical characteristic data of the target object in the test period, and taking the brain electrical characteristic data as sample brain electrical characteristic data.

S240, extracting the sample electroencephalogram characteristic data according to the answering start time and the answering finish time of each test question answered by the target object and the time sequence information of the sample electroencephalogram characteristic data to obtain the sample electroencephalogram characteristic data corresponding to the time period of each test question answered by the target object.

For example, in the acquired sample electroencephalogram feature data, not all sample electroencephalogram feature data may be electroencephalogram feature data of the target object when answering the test question, and some sample electroencephalogram feature data may be electroencephalogram feature data generated when the target object views the answer or views the operation prompt of the preparation page. Therefore, the sample characteristic data needs to be extracted to obtain sample electroencephalogram characteristic data corresponding to the time period for which the target object answers each test question. The electroencephalogram characteristic data corresponds to corresponding time sequence information, and can be extracted according to the response starting time and the response finishing time when the target object is used for responding to each test question, so that the sample electroencephalogram characteristic data corresponding to the time period when the target object is used for responding to each test question is obtained. For example, as shown in fig. 4, if the answer start time of the first question is 15 seconds and the answer finish time is 60 seconds, the sample electroencephalogram feature data is extracted from the data between 15 seconds and 60 seconds, and the sample electroencephalogram feature data corresponding to the time period of the target object for answering the first question is obtained. If the answer starting time of the second question is 60.02 seconds and the answer finishing time is 70 seconds, extracting the sample electroencephalogram characteristic data from the data between 60.02 seconds and 70 seconds to obtain the sample electroencephalogram characteristic data corresponding to the time period of the second question for the target object. And for other non-extracted sample brain electrical characteristic data, the brain electrical characteristic data generated by the target object when the target object is used for answering the question is described, so that the brain electrical characteristic data can be omitted.

S250, continuously segmenting sample electroencephalogram characteristic data corresponding to a time period for answering each test question of the target object according to a preset unit time period.

Different target objects have different thinking time, so that the problem of unequal training data length is solved, and the number of training samples is increased, so that the electroencephalogram characteristics of the samples are segmented. For example, according to a preset unit time period, sample electroencephalogram feature data corresponding to a time period for answering each test question of a target object is continuously segmented, for example, the sample electroencephalogram feature data corresponding to each question is segmented into a plurality of sample electroencephalogram feature data, the time length of each sample electroencephalogram feature data is the preset unit time period, as shown in fig. 5, from 0 second, the sample electroencephalogram feature data of the first question is segmented, and the segmentation is performed into a plurality of sample feature data with the time length of 2 seconds. For sample brain electrical characteristic data less than 2 seconds, no use may be made.

S260, selecting a sample electroencephalogram characteristic data fragment with the time length meeting the preset unit time period.

By way of example, through the steps described above, a sample electroencephalogram feature fragment satisfying a preset unit time period is selected.

In the embodiment of the application, the effective range of the brain electrical characteristic signal of the human brain is 0-30Hz, and the Delta wave band (0-4 Hz) in the brain electrical characteristic signal only appears in deep sleep, so that the method further comprises the step of filtering the sample brain electrical characteristic data by adopting a band-pass filtering technology before the sample brain electrical characteristic data are segmented, so as to obtain the effective sample brain electrical characteristic data. In addition, in the acquisition process, actions such as blinking, hand movement and the like of the target object also interfere with the electroencephalogram characteristic signals, so that independent component analysis (Independent Component Analysis, ICA) is carried out on the filtered sample electroencephalogram characteristic data so as to remove artifact signals such as electrooculogram, myoelectricity and the like in the sample electroencephalogram characteristic signals.

S270, taking the sample electroencephalogram characteristic data fragment and the learning style label of the corresponding target object as training data.

Exemplary, the sample electroencephalogram characteristic data fragments of the target object and the learning style labels corresponding to the target object are combined to obtain training data.

And S280, performing model training according to the training data to obtain a learning style recognition model.

The other technical scheme has the following technical effects: recording the response starting time and the response finishing time of each test question by the target object according to the triggering operation of the target object when each test question is responded, and extracting the sample electroencephalogram characteristic data according to the response starting time and the response finishing time of each test question by the target object and the time sequence information of the sample electroencephalogram characteristic data to obtain the sample electroencephalogram characteristic data corresponding to the time period of each test question by the target object. According to a preset unit time period, sample electroencephalogram characteristic data corresponding to a time period for answering each test question by a target object are continuously segmented to obtain sample electroencephalogram characteristic data fragments, and learning style labels are combined to serve as training data, so that accuracy and objectivity of the training data are improved, data volume of the training data is improved, and accuracy of a training model is further improved.

In the embodiment of the application, the determining process for taking the Rayleigh high-level reasoning test questions as the test questions comprises the following steps: displaying the target Wen Gao-level reasoning test questions to tested objects with different learning style labels; counting the average accuracy and average time of the target object with different learning style labels for answering the Wen Gao-level reasoning test questions; and if the personnel characteristics corresponding to different learning styles determined according to the average accuracy and the average time are consistent with the personnel characteristics corresponding to the actual different learning styles, taking the grade Wen Gao reasoning test questions as test questions. And if the personnel characteristics corresponding to different learning styles determined according to the average accuracy and the average time are consistent with the personnel characteristics corresponding to the actual different learning styles, taking the grade Wen Gao reasoning test questions as test questions, wherein the grade Wen Gao reasoning test questions comprise: carrying out single-factor variance analysis on the average correct rate to obtain a first F statistic; carrying out single-factor variance analysis on the average time to obtain a second F statistic; and if the difference data of the personnel corresponding to different learning styles are determined according to the first F statistic and the second F statistic and are consistent with the difference data of the personnel corresponding to the actual different learning styles, taking the grade Wen Gao reasoning test question as a test question.

For example, since the explicit behavior of the jettisoning type target object tends to be deliberate when it appears to face a problem, the problem is considered and examined with sufficient time, the method of solving various problems is weighed, and then an optimal scheme satisfying various conditions is selected from them, so that the error is less, the explicit behavior of the active type target object appears to be a quick test hypothesis, a decision is made on a prompt according to partial information of the problem or without thorough analysis of the problem, and the reaction speed is faster but error is easy to occur. Therefore, the theoretical answer accuracy of the jettisoning type target object is larger than that of the active type target object, and the theoretical answer time of the jettisoning type target object is larger than that of the active type target object. Therefore, the average accuracy and average time of the tested objects in different learning styles in answering the Rayleigh reasoning questions are counted, and if the characteristics of the theoretical accuracy and time are met, the Rayleigh Wen Gao reasoning questions can excite the performance characteristics of the tested objects in different learning styles and display different apparent behaviors, so that the Rayleigh reasoning test questions are used as test questions.

In addition, single-factor variance analysis can be performed on the average correct rate to obtain a first F statistic; carrying out single-factor variance analysis on the average time to obtain a second F statistic; and if the difference data of the personnel corresponding to different learning styles are determined according to the first F statistic and the second F statistic and are consistent with the difference data of the personnel corresponding to the actual different learning styles, taking the grade Wen Gao reasoning test question as a test question. For example, the first F statistic=5.17 and p=0.0421 <0.05 determined according to the average accuracy rate, so that it can be explained that the average accuracy rates of the measured objects in different learning styles show significant differences, that is, significant differences in the question-making accuracy rates of the measured objects in 2 groups of processing dimensions in different learning styles under each experimental condition are verified. And (3) carrying out single-factor analysis of variance on the average time to obtain a second F statistic=10.54 and P=0.007 <0.01, so that the extremely obvious difference exists in the average time of the tested objects in different learning styles, and the obvious difference in the response time of learners in 2 groups of different learning style processing dimensions under each experimental condition is verified. Therefore, the explanatory Rayleigh Wen Gao reasoning test question can excite the performance characteristics of the tested objects with different learning styles and display different apparent behaviors, so that the Rayleigh reasoning test question is taken as the test question.

In the embodiment of the application, in order to further improve the accuracy of the existing model identification, and in combination with the situation that EEG signals contain high-dimensional characteristic data of time, space and frequency, the experiment optimizes and constructs a one-dimensional multi-scale space-time convolutional neural network model (1-DCNN) based on a convolutional neural network (Convolutional Neural Network, CNN) model to optimize the accuracy of the existing electroencephalogram identification model, and specific improvement measures are as follows: firstly, a one-dimensional time-space convolution kernel is used for replacing a traditional two-dimensional convolution kernel, so that the extraction of the electroencephalogram time domain features and the inter-channel space domain features can be realized, and the parameter quantity of model training is reduced; secondly, a parallel multi-scale convolution module is constructed, so that more abundant brain electrical characteristics can be effectively obtained; third, the model replaces the full connection layer through global average pooling, so that the training speed of the model can be effectively improved, and the overfitting effect can be minimized. Finally, the SVM, BP and 1-DCNN respectively obtain the accuracy rates of 61.4%, 65.6% and 71.2%, and the recognition accuracy rate reaches a higher level in the field, so that the effectiveness of the recognition method provided by the application is verified.

Fig. 6 is a flowchart of a learning style recognition method according to an embodiment of the present application. The learning style identification method provided by the embodiment of the application can be suitable for the situation of identifying the learning style of the object to be identified. The method may in particular be performed by a learning-style recognition device, which may be implemented in software and/or hardware, which may be integrated in an electronic apparatus capable of implementing the learning-style recognition method. Details not described in detail in the embodiments of the application are referred to in the above embodiments. Referring to fig. 6, the method in the embodiment of the present application specifically includes:

S310, displaying a second preset number of test questions to the object to be identified so as to test the object to be identified; the test questions are Rayleigh advanced reasoning test questions.

The object to be identified is an object of which the learning style needs to be identified. The second preset number can be determined according to actual situations, and can be one, two or more, so long as the learning style of the object to be identified can be identified by responding to the object to be identified by the second preset number of test questions. The manner of presentation is the same as in the above-described embodiments.

S320, acquiring the brain electrical characteristic data of the object to be identified in the test period, and taking the brain electrical characteristic data as target brain electrical characteristic data.

The wireless electroencephalograph can be controlled to collect electroencephalogram characteristic data of the object to be identified in the test period, and can be head-mounted electroencephalogram equipment used for collecting the electroencephalogram characteristic data of the object to be identified as target electroencephalogram characteristic data.

S330, inputting the target brain electrical characteristic data into a learning style recognition model to obtain a learning style recognition result of the object to be recognized. The learning style recognition model is obtained by training based on the learning style recognition model training method according to any one of the embodiments.

Inputting the target brain electrical characteristic data into a learning style recognition model, comprising: extracting target electroencephalogram characteristic data according to response starting time and response finishing time when the object to be identified answers each test question and time sequence information of the target electroencephalogram characteristic data to obtain target electroencephalogram characteristic data corresponding to a time period when the object to be identified answers each test question; continuously segmenting target electroencephalogram characteristic data corresponding to a time period for answering each test question of an object to be identified according to a preset unit time period; selecting a target electroencephalogram characteristic data fragment with the time length meeting the preset unit time period; and inputting the target brain electrical characteristic data segment into a learning style recognition model. The extraction and segmentation of the target electroencephalogram feature data are as shown in fig. 7, and the whole process of learning style recognition model training and learning style recognition is as shown in the above embodiment.

In the embodiment of the application, the target electroencephalogram characteristic data is input into a learning style recognition model to obtain a learning style recognition result of an object to be recognized, and the method comprises the following steps: if any learning style exists, the learning style identification model identifies that the target brain electrical characteristic data segment of the learning style is the largest, and the learning style is taken as the learning style corresponding to the object to be identified.

Exemplary, as shown in fig. 8, target electroencephalogram feature data of the object to be identified when the i-th test question is answered is obtained, a plurality of target electroencephalogram feature segments with the time length of 2 seconds are obtained after extraction and segmentation, the target electroencephalogram feature segments are input into a learning style identification model, and a learning style result is obtained for each target electroencephalogram feature segment. If the number of target electroencephalogram characteristic data fragments identified as the jettison type is larger than the number of target electroencephalogram characteristic data frequency bands identified as the active type, determining that the learning style of the object to be identified is jettison type, and if the number of target electroencephalogram characteristic data fragments identified as the jettison type is smaller than the number of target electroencephalogram characteristic data frequency bands identified as the active type, determining that the learning style of the object to be identified is active type.

One of the above technical solutions has the following technical effects: displaying a second preset number of test questions to the object to be identified so as to test the object to be identified; the test questions are Rayleigh advanced reasoning test questions. And acquiring the electroencephalogram characteristic data of the object to be identified in the test period, and taking the electroencephalogram characteristic data as target electroencephalogram characteristic data. And inputting the target brain electrical characteristic data into a learning style recognition model to obtain a learning style recognition result of the object to be recognized. The learning style recognition model is obtained by training based on the learning style recognition model training method according to any one of the embodiments. Through the scheme, the accuracy of learning style identification can be effectively improved.

Fig. 9 is a schematic structural diagram of a learning style recognition model training device according to an embodiment of the present application. The device can be applied to the case of training a learning style recognition model for recognizing a learning style. The apparatus may be implemented in software and/or hardware, and the apparatus may be integrated in an electronic device. Referring to fig. 9, the apparatus specifically includes:

the test question display module 410 is configured to display a first preset number of test questions to a target object, so as to test the target object; the test questions are Rayleigh advanced reasoning test questions;

an electroencephalogram feature acquisition module 420, configured to acquire electroencephalogram feature data of the target object during a test, as sample electroencephalogram feature data;

a training data determining module 430, configured to determine training data according to the sample electroencephalogram feature data and the learning style label of the target object;

the recognition model training module 440 is configured to perform model training according to the training data to obtain a learning style recognition model.

In the embodiment of the application, the target object comprises target objects with different learning style labels, and the number difference of the target objects corresponding to the different learning style labels is smaller than a preset threshold.

In an embodiment of the present application, the apparatus further includes:

and the time recording module is used for recording the answering starting time and the answering finishing time of each test question answered by the target object according to the triggering operation of the target object when answering each test question.

In the embodiment of the application, the time recording module is specifically used for:

if the duration of the target object for answering the current test question exceeds the preset duration, the ending time of the preset duration is used as the answering completion time for answering the test question.

In an embodiment of the present application, the apparatus further includes:

the extraction module is used for extracting the sample electroencephalogram characteristic data according to the response starting time and the response finishing time when the target object is used for responding to each test question and the time sequence information of the sample electroencephalogram characteristic data to obtain the sample electroencephalogram characteristic data corresponding to the time period when the target object is used for responding to each test question.

In an embodiment of the present application, the training data determining module 430 includes:

the segmentation unit is used for continuously segmenting the sample electroencephalogram characteristic data corresponding to the time period for answering each test question of the target object according to the preset unit time period;

The selecting unit is used for selecting sample electroencephalogram characteristic data fragments with the time length meeting the preset unit time period;

and the data determining unit is used for taking the sample electroencephalogram characteristic data fragment and the learning style label of the corresponding target object as training data.

In an embodiment of the present application, the apparatus further includes:

the question display module is used for displaying the grade Wen Gao reasoning test questions to the tested objects with different learning style labels;

the statistics module is used for counting the average accuracy and average time of the target of the Wen Gao-level reasoning test questions responded by the tested objects with different learning style labels;

and the comparison module is used for taking the grade Wen Gao reasoning test questions as test questions if the personnel characteristics corresponding to different learning styles determined according to the average accuracy and the average time are consistent with the personnel characteristics corresponding to the actual different learning styles.

In an embodiment of the present application, the comparison module includes:

the first F statistic determining unit is used for carrying out single-factor variance analysis on the average correct rate to obtain first F statistic;

the second F statistic determining unit is used for carrying out single-factor variance analysis on the average time to obtain a second F statistic;

And the statistic comparison unit is used for taking the Rayleigh Wen Gao-level reasoning test question as the test question if the difference data of the personnel corresponding to different learning styles are determined according to the first F statistic and the second F statistic and are consistent with the difference data of the personnel corresponding to the actual different learning styles.

The learning style recognition model training device provided by the embodiment of the application can execute the learning style recognition model training method provided by any embodiment of the application, and has the corresponding functional modules and beneficial effects of the execution method.

Fig. 10 is a schematic structural diagram of a learning style recognition model training device according to an embodiment of the present application. The device can be suitable for the situation of identifying the learning style of the object to be identified. The apparatus may be implemented in software and/or hardware, and the apparatus may be integrated in an electronic device. Referring to fig. 10, the apparatus may include:

the test module 510 is configured to display a second preset number of test questions to an object to be identified, so as to test the object to be identified; the test questions are Rayleigh advanced reasoning test questions;

the data acquisition module 520 is configured to acquire electroencephalogram feature data of the object to be identified during a test, as target electroencephalogram feature data;

The recognition result determining module 530 is configured to input the target electroencephalogram feature data into a learning style recognition model, so as to obtain a learning style recognition result of the object to be recognized;

the learning style recognition model is obtained by training based on the learning style recognition model training method provided by any one of the embodiments.

In the embodiment of the present application, the recognition result determining module 530 includes:

the data extraction unit is used for extracting the target electroencephalogram characteristic data according to the response starting time and the response finishing time when the object to be identified answers each test question and the time sequence information of the target electroencephalogram characteristic data to obtain target electroencephalogram characteristic data corresponding to the time period when the object to be identified answers each test question;

the data segmentation unit is used for continuously segmenting target electroencephalogram characteristic data corresponding to the time period for answering each test question of the object to be identified according to the preset unit time period;

the segment selection unit is used for selecting target electroencephalogram characteristic data segments with the time length meeting the preset unit time period;

and the input unit is used for inputting the target electroencephalogram characteristic data fragment into a learning style recognition model.

In the embodiment of the present application, the recognition result determining module 530 is specifically configured to:

if any learning style exists, the learning style identification model identifies that the target brain electrical characteristic data segment of the learning style is the largest, and the learning style is taken as the learning style corresponding to the object to be identified.

The learning style recognition device provided by the embodiment of the application can execute the learning style recognition method provided by any embodiment of the application, and has the corresponding functional modules and beneficial effects of the execution method.

Fig. 11 is a schematic structural diagram of an electronic device according to an embodiment of the present application. Fig. 11 illustrates a block diagram of an exemplary electronic device 612 suitable for use in implementing embodiments of the application. The electronic device 612 depicted in fig. 11 is merely an example, and should not be construed as limiting the functionality and scope of use of embodiments of the present application.

As shown in fig. 11, the electronic device 612 may include: one or more processors 616; a memory 628 for storing one or more programs that, when executed by the one or more processors 616, cause the one or more processors 616 to implement a learning style recognition model training method provided by an embodiment of the present application, comprising:

Or the learning style recognition method provided by the embodiment of the application comprises the following steps:

Components of the electronic device 612 may include, but are not limited to: one or more processors or processors 616, a memory 628, and a bus 618 that connects the various device components, including the memory 628 and the processor 616.

Bus 618 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures. By way of example, and not limitation, such architectures include Industry Standard Architecture (ISA) bus, micro channel architecture (MAC) bus, enhanced ISA bus, video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

The electronic device 612 typically includes a variety of computer device readable storage media. Such storage media can be any available storage media that can be accessed by the electronic device 612 and includes both volatile and nonvolatile storage media, removable and non-removable storage media.

Memory 628 may include computer device-readable storage media in the form of volatile memory, such as Random Access Memory (RAM) 630 and/or cache memory 632. The electronic device 612 may further include other removable/non-removable, volatile/nonvolatile computer device storage media. By way of example only, storage system 634 can be used to read from or write to non-removable, nonvolatile magnetic storage media (not shown in FIG. 11, commonly referred to as a "hard disk drive"). Although not shown in fig. 11, a magnetic disk drive for reading from and writing to a removable non-volatile magnetic disk (e.g., a "floppy disk"), and an optical disk drive for reading from or writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical storage medium), may be provided. In such cases, each drive may be coupled to bus 618 through one or more data storage medium interfaces. Memory 628 may include at least one program product having a set (e.g., at least one) of program modules configured to carry out the functions of embodiments of the application.

A program/utility 640 having a set (at least one) of program modules 642 may be stored in, for example, the memory 628, such program modules 642 including, but not limited to, an operating device, one or more application programs, other program modules, and program data, each or some combination of which may include an implementation of a network environment. Program modules 642 generally perform the functions and/or methods of the described embodiments of the present application.

The electronic device 612 may also communicate with one or more external devices 614 (e.g., keyboard, pointing device, display 624, etc.), one or more devices that enable a user to interact with the electronic device 612, and/or any device (e.g., network card, modem, etc.) that enables the electronic device 612 to communicate with one or more other computing devices. Such communication may occur through an input/output (I/O) interface 622. Also, the electronic device 612 may communicate with one or more networks such as a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the Internet, through a network adapter 620. As shown in fig. 11, network adapter 620 communicates with other modules of electronic device 612 over bus 618. It should be appreciated that although not shown in fig. 11, other hardware and/or software modules may be used in connection with the electronic device 612, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID devices, tape drives, data backup storage devices, and the like.

Processor 616 performs various functional applications and data processing by running at least one of the other programs among the plurality of programs stored in memory 628, for example, to implement a learning style recognition model training method provided by embodiments of the present application.

One embodiment of the present application provides a storage medium containing computer-executable instructions for performing a learning style recognition model training method when executed by a computer processor, comprising:

Or for performing a learning style recognition method, comprising:

The computer storage media of embodiments of the present application may take the form of any combination of one or more computer-readable storage media. The computer readable storage medium may be a computer readable signal storage medium or a computer readable storage medium. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor device, apparatus, or means, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In embodiments of the present application, a computer-readable storage medium may be any tangible storage medium that can contain, or store a program for use by or in connection with an instruction execution apparatus, device, or means.

The computer readable signal storage medium may include a propagated data signal with computer readable program code embodied therein, either in baseband or as part of a carrier wave. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal storage medium may also be any computer readable storage medium that is not a computer readable storage medium and that can transmit, propagate, or transport a program for use by or in connection with an instruction execution apparatus, device, or apparatus.

Program code embodied on a computer readable storage medium may be transmitted using any appropriate storage medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

Computer program code for carrying out operations of the present application may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, smalltalk, C ++ and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or device. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (for example, through the Internet using an Internet service provider).

Note that the above is only a preferred embodiment of the present application and the technical principle applied. It will be understood by those skilled in the art that the present application is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the application. Therefore, while the application has been described in connection with the above embodiments, the application is not limited to the embodiments, but may be embodied in many other equivalent forms without departing from the spirit or scope of the application, which is set forth in the following claims.

Claims

1. A method for training an identification model of a Felder-Silverman learning style processing dimension, the method comprising:

displaying a first preset number of test questions to a target object so as to test the target object; the test questions are Rayleigh advanced reasoning test questions; the first preset quantity is set according to the condition that the target object is not bored;

recording the answering starting time and the answering finishing time of each test question by the target object according to the triggering operation of the target object when each test question is answered;

If the duration of the target object for answering the current test question exceeds the preset duration, taking the termination time of the preset duration as the answering completion time for answering the test question;

extracting the sample electroencephalogram characteristic data according to the answering starting time and the answering finishing time when the target object answers each test question and the time sequence information of the sample electroencephalogram characteristic data to obtain sample electroencephalogram characteristic data corresponding to the time period when the target object answers each test question;

continuously segmenting sample electroencephalogram characteristic data corresponding to a time period for answering each test question by the target object according to a preset unit time period;

selecting a sample electroencephalogram characteristic data fragment with the time length meeting the preset unit time period;

processing dimension labels by using the sample electroencephalogram characteristic data fragments and Felder-Silverman learning styles of corresponding target objects as training data;

model training is carried out according to the training data to obtain a recognition model of Felder-Silverman learning style processing dimensionality; the recognition model is constructed based on a one-dimensional multi-scale space-time convolution neural network model, a one-dimensional time-space convolution kernel is used for replacing a traditional two-dimensional convolution kernel, a parallel multi-scale convolution module is constructed, and a full connection layer is replaced through global average pooling.

2. The method of claim 1, wherein the determining of the rayleigh-level reasoning test topics as test topics comprises:

displaying the Rayleigh Wen Gao level reasoning test questions to the tested object with different Felder-Silverman learning style processing dimension labels;

counting the average accuracy and average time of the tested objects with different Felder-Silverman learning styles and dimension labels for answering the Rayleigh Wen Gao-level reasoning test questions;

if the personnel characteristics corresponding to the processing dimensions of different Felder-Silverman learning styles determined according to the average accuracy and the average time consumption are consistent with the personnel characteristics corresponding to the processing dimensions of the actual different Felder-Silverman learning styles, taking the Rayleigh Wen Gao level reasoning test questions as test questions;

if the personnel characteristics corresponding to the processing dimensions of different Felder-Silverman learning styles determined according to the average accuracy and the average time consumption are consistent with the personnel characteristics corresponding to the processing dimensions of different practical Felder-Silverman learning styles, taking the Rayleigh Wen Gao level reasoning test questions as test questions, wherein the steps comprise:

carrying out single-factor variance analysis on the average correct rate to obtain a first F statistic;

Carrying out single-factor variance analysis on the average time to obtain a second F statistic;

and if the difference data of the personnel corresponding to the different Felder-Silverman learning style processing dimensions are determined according to the first F statistic and the second F statistic and are consistent with the difference data of the personnel corresponding to the actual different Felder-Silverman learning style processing dimensions, taking the Rayleigh Wen Gao level reasoning test question as the test question.

3. A method for identifying a dimension of a Felder-Silverman learning style process, the method comprising:

displaying a second preset number of test questions to the object to be identified so as to test the object to be identified; the test questions are Rayleigh advanced reasoning test questions; wherein the second preset number is set according to the condition that the target object does not generate boring feeling;

recording the answering starting time and the answering finishing time of the object to be identified for answering each test question according to the triggering operation of the object to be identified when answering each test question;

if the duration of the to-be-identified object answering the current test question exceeds the preset duration, taking the termination time of the preset duration as the answering completion time for answering the test question;

extracting target electroencephalogram characteristic data according to response starting time and response finishing time when the object to be identified answers each test question and time sequence information of the target electroencephalogram characteristic data to obtain target electroencephalogram characteristic data corresponding to a time period when the object to be identified answers each test question;

continuously segmenting target electroencephalogram characteristic data corresponding to a time period for answering each test question by the object to be identified according to a preset unit time period;

selecting a target electroencephalogram characteristic data fragment with the time length meeting the preset unit time period;

inputting the target electroencephalogram characteristic data fragment into a recognition model of Felder-Silverman learning style processing dimension;

if any learning style exists, the Felder-Silverman learning style processing dimension identification model identifies that the target electroencephalogram characteristic data fragment of the Felder-Silverman learning style processing dimension is the largest, and the Felder-Silverman learning style processing dimension is used as the learning style corresponding to the object to be identified;

the recognition model of the Felder-Silverman learning style processing dimension is trained and obtained based on the recognition model training method of the Felder-Silverman learning style processing dimension of any one of claims 1-2, the recognition model is constructed based on a one-dimensional multi-scale space-time convolution neural network model, a one-dimensional time-space convolution kernel is used for replacing a traditional two-dimensional convolution kernel, a parallel multi-scale convolution module is constructed, and a full-connection layer is replaced by global average pooling.

4. An identification model training device for a Felder-Silverman learning style processing dimension, the device comprising:

the test question display module is used for displaying a first preset number of test questions to a target object so as to test the target object; the test questions are Rayleigh advanced reasoning test questions; the first preset quantity is set according to the condition that the target object is not bored;

the time recording module is used for recording the answering starting time and the answering finishing time of each test question answered by the target object according to the triggering operation of the target object when answering each test question;

the time recording module is specifically configured to take a termination time of a preset duration as a response completion time for responding to the test question if the duration of the target object for responding to the current test question exceeds the preset duration;

the extraction module is used for extracting the sample electroencephalogram characteristic data according to the response starting time and the response finishing time when the target object answers each test question and the time sequence information of the sample electroencephalogram characteristic data to obtain sample electroencephalogram characteristic data corresponding to the time period when the target object answers each test question;

A training data determination module comprising:

the data determining unit is used for processing dimension labels of Felder-Silverman learning styles of the sample electroencephalogram characteristic data fragments and corresponding target objects as training data;

the recognition model training module is used for carrying out model training according to the training data to obtain a recognition model of Felder-Silverman learning style processing dimensionality; the recognition model is constructed based on a one-dimensional multi-scale space-time convolution neural network model, a one-dimensional time-space convolution kernel is used for replacing a traditional two-dimensional convolution kernel, a parallel multi-scale convolution module is constructed, and a full connection layer is replaced through global average pooling.

5. A device for identifying a dimension of a Felder-Silverman learning style process, the device comprising:

the test module is used for displaying a second preset number of test questions to the object to be identified so as to test the object to be identified; the test questions are Rayleigh advanced reasoning test questions; wherein the second preset number is set according to the condition that the target object does not generate boring feeling;

The time determining module is used for recording the answering starting time and the answering finishing time of the object to be identified for answering each test question according to the triggering operation of the object to be identified when the object to be identified answers each test question;

the time determining module is specifically configured to, if the duration of the response of the object to be identified to the current test question exceeds a preset duration, take the termination time of the preset duration as the response completion time of the response to the test question;

the identification result determining module comprises:

the input unit is used for inputting the target electroencephalogram characteristic data fragment into an identification model of Felder-Silverman learning style processing dimensionality;

the recognition result determining module is specifically configured to, if any learning style exists, identify that a target electroencephalogram feature data segment of a Felder-Silverman learning style processing dimension is the largest in a Felder-Silverman learning style processing dimension recognition model, and use the Felder-Silverman learning style processing dimension as a learning style corresponding to the object to be recognized;

6. An electronic device, the electronic device comprising:

One or more processors;

a memory for storing one or more programs;

when the one or more programs are executed by the one or more processors, the one or more processors implement the method for model training for recognition of the Felder-Silverman learning style processing dimension as claimed in any of claims 1-2, or the method for recognition of the Felder-Silverman learning style processing dimension as claimed in claim 3.

7. A computer readable storage medium having stored thereon a computer program, which when executed by a processor, implements the method for model training for recognition of a Felder-Silverman learning style process dimension according to any of claims 1-2, or the method for recognition of a Felder-Silverman learning style process dimension according to claim 3.