Teaching management method and device, terminal equipment and storage medium

Abstract

The invention discloses a teaching management method and apparatus, a terminal device and a storage medium. The method comprises the steps of collecting multi-source teaching data of a preset time length; performing feature extraction on the multi-source teaching data to obtain a plurality of teaching feature vectors, and fusing the teaching feature vectors to generate a comprehensive feature matrix; arranging the multi-source teaching data according to a time sequence, calculating a time feature vector of each preset time point, and generating time sequence features based on the time feature vectors; converting the multi-source teaching data into a fuzzy set, and generating fuzzy output of the multi-source teaching data according to a preset fuzzy library rule and the time sequence features; inputting the comprehensive feature matrix and the fuzzy output into a preset neural network model to enable the neural network model to perform data prediction according to a preset teaching management decision target, and outputting a teaching management decision result; and generating a teaching management scheme according to the teaching management decision result. According to the invention, the teaching management quality and efficiency can be improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and particularly to a teaching management method, device, terminal device and storage medium. Background Art

[0002] Teaching management refers to a series of activities such as planning, organizing, coordinating, controlling and evaluating to ensure the achievement of teaching objectives and the improvement of teaching quality. It covers multiple aspects such as curriculum design, teaching resource allocation, teaching process monitoring, and teaching effect evaluation, aiming to optimize the teaching environment and improve students' learning effects. Teaching management methods play a crucial role in the teaching process, which can help teachers better organize teaching content, reasonably allocate teaching resources, monitor teaching progress, evaluate teaching effects, and make adjustments according to feedback. Effective teaching management methods can improve teaching efficiency, promote the all-round development of students, and enhance teaching quality.

[0003] Currently, the commonly used teaching management methods still stay in the traditional classroom management stage, that is, teaching management is carried out through classroom discipline, attendance, homework assignment, etc. Some schools also allocate teaching resources and track students' learning progress through information technology, online examination systems, etc., and make teaching decisions and adjustments by collecting and analyzing students' learning data. However, traditional methods and information-based management methods are often difficult to carry out personalized management according to the specific needs of each student and cannot respond to students' learning needs in a timely manner. Moreover, teachers need to invest a lot of time and energy in the management process and it is difficult to focus on teaching itself. How to formulate a more effective teaching management method is a technical problem to be solved urgently. Summary of the Invention

[0004] Embodiments of the present invention provide a teaching management method, device, terminal device and storage medium to solve the above problems and improve the quality and efficiency of teaching management.

[0005] An embodiment of the present invention provides a teaching management method, including:

[0006] Collecting multi-source teaching data of a preset time length;

[0007] Extracting features from the multi-source teaching data to obtain a number of teaching feature vectors, and fusing the teaching feature vectors to generate a comprehensive feature matrix;

[0008] Arranging the multi-source teaching data in chronological order, calculating time feature vectors at each preset time point, and generating time series features based on the time feature vectors; wherein, the time series features are used to describe the changing trends of students' learning behaviors, teachers' teaching effects and course contents within the preset time length;

[0009] Convert the multi-source teaching data into a fuzzy set, and generate a fuzzy output of the multi-source teaching data according to the preset fuzzy library rules and the time series characteristics; wherein, the fuzzy output is the teaching evaluation result of the preset time length.

[0010] Input the comprehensive feature matrix and the fuzzy output into a preset neural network model, so that the neural network model performs data prediction according to the preset teaching management decision-making goal and outputs a teaching management decision result.

[0011] Generate a teaching management plan according to the teaching management decision result.

[0012] Further, the collection of multi-source teaching data of a preset time length specifically includes:

[0013] Obtain initial multi-source teaching data through a teaching management system; wherein, the initial multi-source teaching data includes student behavior data, teacher behavior data, and course content data.

[0014] Call corresponding preprocessing strategies according to the data types of the initial multi-source teaching data, and perform data processing on the initial multi-source teaching data based on the preprocessing strategies to obtain the multi-source teaching data; wherein, the multi-source teaching data includes a student behavior feature table, a teacher teaching feature table, and a course content feature table.

[0015] Further, the calling of corresponding preprocessing strategies according to the data types of the initial multi-source teaching data and performing data processing on the initial multi-source teaching data based on the preprocessing strategies to obtain the multi-source teaching data includes:

[0016] The preprocessing strategies include a numerical data preprocessing strategy, a text data preprocessing strategy, and an image data preprocessing strategy.

[0017] The numerical data preprocessing strategy is to perform normalization processing on the numerical data in the initial multi-source teaching data and unify the numerical data within a preset numerical range.

[0018] The text data preprocessing strategy is to perform word segmentation and word vectorization operations on the text data in the initial multi-source teaching data and convert the text data into text feature vectors.

[0019] The image data preprocessing strategy is to extract the key frames of the image data in the initial multi-source teaching data, perform visual feature extraction on the key frames based on a convolutional neural network model, and generate the concentration feature of the student behavior feature table according to the extracted visual features.

[0020] Further, the feature extraction of the multi-source teaching data to obtain a number of teaching feature vectors, and the fusion of the teaching feature vectors to generate a comprehensive feature matrix includes:

[0021] Taking the column elements of each feature table as feature objects for feature extraction respectively, and constructing the teaching feature vectors based on the feature values of each feature object; wherein, the teaching feature vectors include student behavior feature vectors, teacher teaching feature vectors, and course content feature vectors;

[0022] Concatenating the student behavior feature vectors, the teacher teaching feature vectors, and the course content feature vectors to obtain the comprehensive feature matrix.

[0023] Further, the calculation of the time feature vectors at each preset time point and the generation of time series features based on the time feature vectors include:

[0024] Arranging the multi-source teaching data in chronological order, calculating the teaching feature vectors at each preset time point and concatenating them to obtain the time feature vectors;

[0025] Arranging each of the time feature vectors in sequence to obtain time series data;

[0026] Performing a sliding window operation on the time series data based on a window of a preset size to generate the time series features.

[0027] Further, converting the multi-source teaching data into a fuzzy set, and generating a fuzzy output of the multi-source teaching data according to preset fuzzy library rules and the time series features, includes:

[0028] Defining fuzzy output targets according to the feature table types of the student behavior feature table and the teacher teaching feature table; wherein, the fuzzy output targets include student learning status and teacher teaching effect;

[0029] Taking the column elements of the student behavior feature table and the teacher teaching feature table as fuzzy input objects respectively, and constructing fuzzy sets corresponding to each column element based on the values of the column elements in the time feature sequence;

[0030] Defining the mapping relationship between each column element in the feature table and the corresponding fuzzy output target, and generating the fuzzy output target of the fuzzy set based on the mapping relationship;

[0031] Defuzzifying the fuzzy output target to obtain the fuzzy output; wherein, the fuzzy output includes a learning status score and a teaching effect score.

[0032] Further, inputting the comprehensive feature matrix and the fuzzy output into a preset neural network model, so that the neural network model performs data analysis according to a preset teaching management decision-making goal and outputs a teaching management decision result, includes:

[0033] Concatenate the comprehensive feature matrix and the fuzzy output, and input the obtained concatenated data into the preset neural network model; wherein, the preset neural network model is trained according to historical multi-source teaching data;

[0034] Control the preset neural network model to predict the concatenated data based on a preset teaching management decision-making goal, and output a teaching management decision result; wherein, the teaching management decision-making goal includes student class assignment suggestions, teaching, teaching resource recommendations, and teaching improvement suggestions.

[0035] Based on the above method item embodiments, the present invention correspondingly provides device item embodiments;

[0036] An embodiment of the present invention provides a teaching management device, including: a data acquisition module, a first feature module, a second feature module, a fuzzy processing module, a data prediction module, and a solution generation module;

[0037] The data acquisition module is used to acquire multi-source teaching data of a preset time length;

[0038] The first feature module is used to extract features from the multi-source teaching data to obtain several teaching feature vectors, and fuse the teaching feature vectors to generate a comprehensive feature matrix;

[0039] The second feature module is used to arrange the multi-source teaching data in chronological order, calculate the time feature vectors of each preset time point, and generate time series features based on the time feature vectors; wherein, the time series features are used to describe the change trends of students' learning behaviors, teachers' teaching effects, and course contents within the preset time length;

[0040] The module processing module is used to convert the multi-source teaching data into a fuzzy set, and generate a fuzzy output of the multi-source teaching data according to preset fuzzy library rules and the time series features; wherein, the fuzzy output is the teaching evaluation result of the preset time length;

[0041] The data prediction module is used to input the comprehensive feature matrix and the fuzzy output into a preset neural network model, so that the neural network model performs data prediction according to a preset teaching management decision-making goal and outputs a teaching management decision result;

[0042] The solution generation module is used to generate a teaching management solution according to the teaching management decision result.

[0043] Based on the above method item embodiments, the present invention provides another embodiment;

[0044] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the teaching management method provided in any one of the above method item embodiments of the present application.

[0045] Based on the method item embodiments of the present invention, another embodiment is provided:

[0046] Another embodiment of the present invention provides a storage medium. The storage medium includes a stored computer program. When the computer program runs, it controls the device where the storage medium is located to execute the teaching management method provided in any one of the above method item embodiments of the present invention.

[0047] By implementing the embodiments of the present invention, the following beneficial effects are achieved:

[0048] Embodiments of the present invention provide a teaching management method, apparatus, terminal device, and storage medium. The method collects multi-source teaching data of a preset time length, covering a wider range of information, which can more comprehensively reflect the behaviors and states of teachers and students during the teaching process, providing a rich and diversified basis for subsequent analysis and helping to discover teaching problems and student needs that are difficult to detect by traditional methods. By feature extraction, the multi-source teaching data is transformed into teaching feature vectors, which can focus on the key information in the data. Then, these feature vectors are fused to generate a comprehensive feature matrix, enabling data from different sources to complement and correlate with each other, forming a more representative and comprehensive teaching data representation form, which helps to more accurately grasp the overall situation of teaching. Further, time series features are introduced to describe the changing trends of students' learning behaviors, teachers' teaching effects, and course content within the preset time length. By analyzing these changing trends, teachers can timely discover problems such as the fluctuations in students' learning states and the effectiveness changes of teachers' teaching methods, thereby timely adjusting teaching strategies, better responding to students' learning needs, and providing more forward-looking information for teaching decisions. Converting the multi-source teaching data into a fuzzy set and generating a fuzzy output using fuzzy library rules and time series features can more reasonably process this uncertain information and provide a more comprehensive and objective teaching evaluation result. Finally, a neural network model is used to deeply analyze and learn the comprehensive feature matrix and the fuzzy output, make data predictions according to the preset teaching management decision-making goals, utilize the learning ability and adaptive ability of the neural network model to discover potential rules and patterns from a large amount of data, and output more scientific and reasonable teaching management decision results. At the same time, teachers do not need to spend a lot of time and energy on data collection, analysis, and decision-making, so they can invest more energy in teaching itself and improve teaching quality. Based on the teaching management decision results obtained from the previous steps, personalized teaching management plans for different students and teaching scenarios can be generated, overcoming the defect that traditional methods are difficult to perform personalized management according to the specific needs of each student, providing a more comprehensive, accurate, and personalized teaching management plan, and improving the efficiency and quality of teaching management. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 FIG. is a schematic flowchart of a teaching management method provided by an embodiment of the present invention.

[0050] Figure 2 FIG. is a schematic structural diagram of a teaching management apparatus provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0052] See Figure 1 , Figure 1 which is a schematic flowchart of a teaching management method provided by an embodiment of the present invention, including steps S101 to S106. The specific steps are as follows:

[0053] Step S101: Collect multi-source teaching data of a preset time length.

[0054] In one embodiment, the collecting of multi-source teaching data of a preset time length specifically includes: obtaining initial multi-source teaching data through a teaching management system; wherein, the initial multi-source teaching data includes student behavior data, teacher behavior data, and course content data; calling corresponding preprocessing strategies according to the data types of the initial multi-source teaching data, and performing data processing on the initial multi-source teaching data based on the preprocessing strategies to obtain the multi-source teaching data; wherein, the multi-source teaching data includes a student behavior feature table, a teacher teaching feature table, and a course content feature table.

[0055] In the embodiments of the present invention, student behavior data, teacher behavior data, and course content data are obtained from the teaching management system as research data for the teaching management method. Generally, student behavior data includes attendance rate, classroom participation rate, homework completion rate, and exam scores; teacher behavior data includes teaching content, teaching methods, and classroom interaction frequency; course content data includes course difficulty, knowledge point distribution, and teaching resources. Further, classroom video and audio data can also be collected through cameras and microphones.

[0056] In one embodiment, the corresponding preprocessing strategy is called according to the data type of the initial multi-source teaching data, and the initial multi-source teaching data is processed based on the preprocessing strategy to obtain the multi-source teaching data, including: the preprocessing strategy includes a numerical data preprocessing strategy, a text data preprocessing strategy, and an image data preprocessing strategy; the numerical data preprocessing strategy is to normalize the numerical data in the initial multi-source teaching data and unify the numerical data within a preset numerical range; the text data preprocessing strategy is to perform word segmentation and word vectorization operations on the text data in the initial multi-source teaching data and convert the text data into text feature vectors; the image data preprocessing strategy is to extract the key frames of the image data in the initial multi-source teaching data, extract visual features from the key frames based on a convolutional neural network model, and generate the concentration feature of the student behavior feature table according to the extracted visual features.

[0057] In the embodiment of the present invention, the collected initial multi-source teaching data needs to be subjected to certain data cleaning and preprocessing for subsequent use. The collected multi-source teaching data is processed through several data preprocessing strategies. The numerical data such as exam scores and attendance rates is scaled to the range of [0,1] through the standard deviation formula. The expression of the normalization formula is In the formula, X is the numerical value. For text data such as teacher teaching records and student feedback texts, operations such as word segmentation, stop word removal (such as "de", "shi"), and converting the text into a vector using the TF-IDF model (Term Frequency-Inverse Document Frequency) are performed to convert the text into text feature vectors. For image data such as classroom videos and images, the video key frames are extracted (such as extracting 1 frame per minute), and the VGG16 (Visual Geometry Group 16) convolutional neural network model is used to extract visual features (such as student concentration and facial expressions). Based on the above preprocessing strategies, the collected initial multi-source teaching data can be cleaned and sorted to obtain a student behavior feature table, a teacher teaching feature table, and a course content feature table. Generally, the student behavior feature table includes learning duration, exam scores, concentration, and homework completion rate; the teacher teaching feature table includes interaction frequency and knowledge point coverage; the course content feature table includes teaching method vectors, knowledge point complexity, and resource richness.

[0058] Exemplarily, the initial multi-source teaching data of a certain class for one week is collected as follows:

[0059] Student behavior data:

[0060]

[0061] Teacher behavior data:

[0062]

[0063] Course content data:

[0064]

[0065] For numerical data, normalize the exam scores to the range [0, 1], i.e., 85 → 0.85, 75 → 0.75. For text data, segment the teacher's teaching method "discussion-based teaching" into ["discussion", "teaching"], and convert it into a vector [0.6, 0.4] through the TF-IDF method; for image data, extract key frames from the video (such as the student's facial image at the 10th minute), and use the VGG16 model to extract the concentration features [0.8, 0.7]. After data processing based on the above data preprocessing strategy, the final multi-source teaching data can be obtained, including the student behavior feature table, the teacher teaching feature table, and the course content feature table. The examples are as follows:

[0066] Student behavior feature table:

[0067]

[0068] Teacher teaching feature table:

[0069]

[0070] Course content feature table:

[0071]

[0072] In the table, the concentration level is a concentration score generated by analyzing students' facial expressions (such as line of sight direction, head posture) and behaviors (such as frequency of raising hands, note-taking) in classroom videos using computer vision algorithms (such as OpenCV or deep learning models), with a range of 0% - 100%. The concentration score of student 001 is 0.8, indicating a high level of concentration in class. The homework completion rate is the ratio of the number of completed homework by students to the total number of homework assigned by teachers. Exemplarily, student 001 completed 9 assignments, and the total number of assignments was 10, with a completion rate of 0.9. The interaction frequency is the number of interaction behaviors (such as questions, discussions, group activities) of teachers in each class, and is divided into high, medium, and low levels according to a preset threshold. The knowledge point coverage is the ratio of the number of knowledge points explained by teachers in class to the total number of knowledge points in the course. The knowledge point complexity is as follows: combining the course difficulty (high, medium, low) and the knowledge point distribution (extensive, concentrated), a complexity rating is generated through expert scoring or an algorithm model. The resource richness is the ratio of the number of teaching resources (such as courseware, videos, exercise questions) in the course to the number of knowledge points, reflecting the sufficiency of resources. Exemplarily, course C01 has 10 teaching resources and 20 knowledge points, with a resource richness of 0.5.

[0073] It should be noted that the column elements of the generated student behavior feature table, teacher teaching feature table, and course content feature table can be customized and adjusted according to specific teaching management requirements. The above is only for illustration and is not used to limit each feature table. Through the above operations of data collection and preprocessing, useful data features can be extracted from multi-source teaching data, providing high-quality input data for the prediction of subsequent models and improving the accuracy of teaching management results.

[0074] Step S102: Extract features from the multi-source teaching data to obtain a number of teaching feature vectors, and fuse the teaching feature vectors to generate a comprehensive feature matrix.

[0075] In one embodiment, the extracting features from the multi-source teaching data to obtain a number of teaching feature vectors and fusing the teaching feature vectors to generate a comprehensive feature matrix includes: respectively taking the column elements of each feature table as feature objects for feature extraction, and constructing the teaching feature vectors based on the feature values of each feature object; wherein, the teaching feature vectors include student behavior feature vectors, teacher teaching feature vectors, and course content feature vectors; splicing the student behavior feature vectors, the teacher teaching feature vectors, and the course content feature vectors to obtain the comprehensive feature matrix.

[0076] In the embodiments of the present invention, feature extraction is performed on the column elements of each feature table of the student behavior feature table, the teacher teaching feature table, and the course content feature table to generate corresponding student behavior feature vectors, teacher teaching feature vectors, and course content feature vectors. Among them, the student behavior feature vector includes classroom participation, learning duration (the average value is calculated by statistically counting the daily learning duration of students), concentration, homework completion rate, etc.; the teacher teaching feature vector includes interaction frequency, teaching method vector, etc.; the course content feature vector includes knowledge point complexity, resource richness, etc. The student behavior feature vector, the teacher teaching feature vector, and the course content feature vector are concatenated by rows to generate a comprehensive feature matrix. Each row of the comprehensive feature matrix corresponds to a student and contains the comprehensive features of the student, teacher, and course.

[0077] Exemplarily, if the learning duration of student 001 is 2.5 hours, the concentration is 80%, and the homework completion rate is 90%, then the student behavior feature vector is [2.5, 0.8, 0.9]. If the learning duration of student 002 is 2.0 hours, the concentration is 70%, and the homework completion rate is 80%, then the student behavior feature vector of student 002 is [2.0, 0.7, 0.8]. If the classroom interaction frequency of teacher T01 is high (by default, a single-classroom interaction frequency exceeding 10 times is considered high), the knowledge point coverage is 80%, and the teaching method vector is [0.6, 0.4], then the teacher teaching feature vector of teacher T01 is [10, 0.8, 0.6, 0.4]. If the classroom interaction frequency of teacher T02 is medium (by default, a single-classroom interaction frequency exceeding 5 times is considered medium), the knowledge point coverage is 50%, and the teaching method vector is [0.6, 0.4], then the teacher teaching feature vector of teacher T02 is [5, 0.5, 0.6, 0.4]. If the knowledge point complexity of course C01 is 0.9 and the resource richness is 0.5, then the course content feature vector of course C01 is [0.9, 0.5]. If the knowledge point complexity of course C02 is 0.6 and the resource richness is 0.53, then the course content feature vector of course C02 is [0.6, 0.53]. By concatenating the above feature vectors respectively, the comprehensive feature vector of each student can be obtained, and then the comprehensive feature matrix can be obtained. Exemplarily, the comprehensive feature vector of student 001 is [2.5, 0.8, 0.6, 0.4, 0.9, 10, 0.8, 0.9, 0.5]; the comprehensive feature vector of student 002: [2.0, 0.7, 0.8, 5, 0.5, 0.6, 0.4, 0.6, 0.53]. The comprehensive feature matrix is:

[0078]

[0079] Step S103: Arrange the multi-source teaching data in chronological order, calculate the time feature vectors at each preset time point, and generate time series features based on the time feature vectors; wherein, the time series features are used to describe the changing trends of students' learning behaviors, teachers' teaching effects, and course content within the preset time length.

[0080] In one embodiment, the calculating the time feature vectors at each preset time point and generating the time series features based on the time feature vectors includes: arranging the multi-source teaching data in chronological order, calculating and splicing the teaching feature vectors at each preset time point to obtain the time feature vectors; arranging each of the time feature vectors in sequence to obtain time series data; and performing a sliding window operation on the time series data based on a window of a preset size to generate the time series features.

[0081] In the embodiments of the present invention, the student behavior feature table, the teacher teaching feature table, and the course content feature table are sorted in chronological order. It should be noted that, for the convenience of description, the following teaching feature vectors are only exemplified by the values of one column element in the student behavior feature table, the teacher teaching feature table, and the course content feature table within a certain time length respectively.

[0082] Exemplarily, the learning duration sequence of student 001 is [2.5, 2.6, 2.7, 2.8, 2.9] (hours); the classroom interaction frequency sequence of teacher T01 is [10, 11, 12, 13, 14] (times / lesson); the knowledge point coverage sequence of course C01 is [0.8, 0.85, 0.9, 0.95, 1.0]. Each element in each sequence corresponds to a teaching feature vector. Based on the above sequences, the teaching feature vectors of the multi-source teaching data at each time point are generated. Then, the time feature vector at time point 1 is [2.5, 10, 0.8]; the time feature vector at time point 2 is [2.6, 11, 0.85]; the time feature vector at time point 3 is [2.7, 12, 0.9]... Arrange the teaching feature vectors in chronological order to generate time series data. The expression of the time series data is:

[0083]

[0084] Set the size of sliding window 1 to 3, and perform a window sliding operation on the time series data based on this sliding window, that is, one sliding window can be used to view the time series data of 3 time points at the same time. Exemplarily, window 1 can view the time series data of time points 1 - 3 at the same time, that is:

[0085]

[0086]

[0087] Slide the sliding window 1 with a step size of 1, and name the slid window as window 2. Then window 2 can view the time series data at time points 2 - 4 simultaneously, that is:

[0088]

[0089] Similarly, slide window 2 and name it as window 3. Then window 3 can view the time series data at time points 3 - 5 simultaneously. In summary, the time series features in different window cases can be obtained as follows:

[0090]

[0091] Based on the above time series features, the changing trends of students' learning behaviors, teachers' teaching effects, and the coverage of course content can be analyzed. It should be noted that the size of the sliding window and the setting of the sliding step size can be custom - set according to factors such as the data volume and data processing requirements in each feature table of the multi - source teaching data, and are not limited here.

[0092] Step S104: Convert the multi - source teaching data into a fuzzy set, and generate a fuzzy output of the multi - source teaching data according to the preset fuzzy library rules and the time series features; wherein, the fuzzy output is the teaching evaluation result of the preset time length.

[0093] In an embodiment, define a fuzzy output target according to the feature table types of the student behavior feature table and the teacher teaching feature table; wherein, the fuzzy output target includes students' learning status and teachers' teaching effects; respectively use the column elements of the student behavior feature table and the teacher teaching feature table as fuzzy input objects, and construct fuzzy sets corresponding to each column element based on the values of the column elements in the time feature sequence; define the mapping relationship between each column element in the feature table and the corresponding fuzzy output target, and generate the fuzzy output target of the fuzzy set based on the mapping relationship; defuzzify the fuzzy output target to obtain the fuzzy output; wherein, the fuzzy output includes a learning status score and a teaching effect score.

[0094] In the embodiment of the present invention, convert the feature values corresponding to each column element (such as class participation rate, homework completion rate, interaction frequency) in the student behavior feature table and the teacher teaching feature table in the time feature sequence into fuzzy sets (such as low, medium, high). Then define the mapping relationship between each column element and the output target, and this relationship is the fuzzy library rule. Based on the defined mapping relationship, reason about the column element values to generate a fuzzy output target. Then convert the fuzzy output target into specific numerical values or categories to achieve defuzzification.

[0095] Defuzzify the eigenvalues of each column element to obtain a fuzzy set. Exemplarily, for a class participation rate of 0 - 30%, it is determined that the class participation is low; for a class participation rate of 31% - 70%, it is determined that the class participation is medium; for a class participation rate of 71% - 100%, it is determined that the class participation is high. Based on the above defuzzification operation, a fuzzy set of class participation can be obtained. An example form of the fuzzy set is: [low, medium, medium, high]. Similarly, fuzzy sets of the remaining column elements can be obtained, which will not be elaborated here. Define the mapping relationship between each column element and the fuzzy output target. It should be noted that this mapping relationship can be determined according to the weight ratio of each column element in teaching management, or can be custom-set according to actual task requirements, which is not limited here. Exemplarily, define the mapping relationship between column elements and students' learning status as follows: If the "class participation" is high and the "homework completion rate" is high, then the "learning status" is excellent; if the "class participation" is medium and the "homework completion rate" is medium, then the "learning status" is good; if the "class participation" is low and the "homework completion rate" is low, then the "learning status" is poor. Define the mapping relationship between column elements and teachers' teaching effectiveness as follows: If the "course difficulty" is high and the "teaching resources" are insufficient, then the "teaching effectiveness" is poor; if the "course difficulty" is medium and the "teaching resources" are sufficient, then the "teaching effectiveness" is good. If student 001's class participation = high and homework completion rate = high, then it is determined that his learning status is excellent; if student 002's class participation = medium and homework completion rate = medium, then it is determined that his learning status is good. If teacher T01's course difficulty = high and teaching resources = insufficient, then it is determined that his teaching effectiveness is poor; if teacher T02's course difficulty = medium and teaching resources = sufficient, then it is determined that his teaching effectiveness is good. Convert the results (excellent, good, poor) of the fuzzy output target into scores according to the custom conversion relationship. Exemplarily, if a student's learning status is excellent, the score is 90, for good it is 75, and for poor it is 60. Similarly, the teaching effectiveness scores of teachers are obtained. Further, generate the learning status scores of students at different time points. For example, if student 001's learning status scores are [85, 88, 90], then it can be determined that student 001's learning status trend is steadily improving. If student 002's learning status scores are [70, 72, 75], it can be determined that student 002's learning status score is slowly improving.

[0096] Step S105: Convert the multi-source teaching data into a fuzzy set, and generate a fuzzy output of the multi-source teaching data according to the preset fuzzy library rules and the time series characteristics; wherein, the fuzzy output is the teaching evaluation result of the preset time length.

[0097] In one embodiment, inputting the comprehensive feature matrix and the fuzzy output into a preset neural network model, so that the neural network model performs data analysis according to a preset teaching management decision-making objective and outputs a teaching management decision result, includes: splicing the comprehensive feature matrix and the fuzzy output, and inputting the obtained spliced data into the preset neural network model; wherein, the preset neural network model is trained according to historical multi-source teaching data; controlling the neural network model to predict the spliced data based on a preset teaching management decision-making objective, and outputting a teaching management decision result; wherein, the teaching management decision-making objective includes student class division suggestions, teaching, teaching resource recommendations, and teaching improvement suggestions.

[0098] In an embodiment of the present invention, after splicing the comprehensive feature matrix and the fuzzy output, they are input into a neural network model. Exemplarily, the comprehensive feature matrix is as follows:

[0099]

[0100] The fuzzy output result is that the learning status score of student 001 is 90, and the teaching effect score of teacher T01 is 85. The spliced data obtained by splicing the comprehensive feature matrix and the fuzzy output is [2.5, 0.8, 0.9, 10, 0.8, 0.9, 0.5, 90, 85]. Input this spliced data into the trained neural network model. Preferably, in an embodiment of the present invention, a random forest model is used as the initial neural network model, and through historical data, that is, the historical comprehensive feature matrix and the corresponding fuzzy output as the training data set, and class division suggestions, resource recommendations, etc. as labels for training, the final neural network model is obtained. The neural network model makes data predictions according to a preset teaching management decision-making objective (i.e., the labels used during training) to generate the final teaching management decision result. Exemplarily, the teaching management decision result output by the model is: (1) Student class division suggestion: Class A; (2) Teaching resource recommendation: videos, documents; Teaching improvement suggestion: It is recommended to increase classroom interaction.

[0101] Step S106: Generate a teaching management plan according to the teaching management decision result.

[0102] In an embodiment of the present invention, a corresponding teaching management plan is generated according to the teaching management decision result output by the model, including student class division results, a teaching resource recommendation list, and a teaching effect evaluation report. Exemplarily, the student class division result is: Student 001 is assigned to Class A; the teaching resource recommendation list is the recommended resource types, including videos, documents, etc.; the teaching effect evaluation report includes the teaching effect score and teaching improvement suggestions.

[0103] Based on the above method item embodiment, a corresponding device item embodiment is provided;

[0104] As shown in the figure, another embodiment of the present invention provides a teaching management device, including: a data acquisition module, a first feature module, a second feature module, a fuzzy processing module, a data prediction module, and a solution generation module;

[0105] The data acquisition module is used to acquire multi-source teaching data of a preset time length;

[0106] The first feature module is used to extract features from the multi-source teaching data to obtain a number of teaching feature vectors, and fuse the teaching feature vectors to generate a comprehensive feature matrix;

[0107] The second feature module is used to arrange the multi-source teaching data in chronological order, calculate the time feature vectors at each preset time point, and generate time series features based on the time feature vectors; wherein, the time series features are used to describe the change trends of students' learning behaviors, teachers' teaching effects, and course contents within the preset time length;

[0108] The module processing module is used to convert the multi-source teaching data into a fuzzy set, and generate a fuzzy output of the multi-source teaching data according to preset fuzzy library rules and the time series features; wherein, the fuzzy output is the teaching evaluation result of the preset time length;

[0109] The data prediction module is used to input the comprehensive feature matrix and the fuzzy output into a preset neural network model, so that the neural network model performs data prediction according to the preset teaching management decision-making goal and outputs a teaching management decision result;

[0110] The solution generation module is used to generate a teaching management solution according to the teaching management decision result.

[0111] It can be understood that the above device item embodiment corresponds to the method item embodiment of the present invention, and it can implement the teaching management method provided by any one of the above method item embodiments of the present invention.

[0112] It should be noted that the device embodiments described above are merely illustrative. The units / modules described as separate components may or may not be physically separated, and the components shown as units / modules may or may not be physical units / modules. That is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the attached drawings of the device embodiments provided by the present invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art can understand and implement without creative efforts. The schematic diagram is only an example of the teaching management device and does not constitute a limitation on the teaching management device. It may include more or fewer components than shown in the figure, or combine some components, or different components.

[0113] Based on the above method item embodiments, another embodiment is provided;

[0114] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the teaching management method provided in any one of the above method item embodiments of the present invention.

[0115] Exemplarily, the computer program can be divided into one or more modules / units. The one or more modules / units are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, and these instruction segments are used to describe the execution process of the computer program in the terminal device.

[0116] The terminal device can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The terminal device may include, but is not limited to, a processor and a memory. Those skilled in the art can understand that, for example, the terminal device may further include input / output devices, network access devices, a bus, etc.

[0117] The so-called processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The processor is the control center of the terminal device, and connects various parts of the entire terminal device through various interfaces and circuits.

[0118] The memory can be used to store the computer programs and / or modules. The processor realizes various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory, and by calling the data stored in the memory. The memory mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system, application programs required for at least one function (such as a sound display function, an image display function, etc.); the data storage area can store data created according to the use of the mobile phone (such as audio data, phone book, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, at least one magnetic disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0119] Based on the above-mentioned invention embodiments, corresponding embodiments of the storage medium are provided;

[0120] Another embodiment of the present invention provides a storage medium. The storage medium includes a stored computer program, wherein when the computer program runs, it controls the device where the storage medium is located to execute the teaching management method provided by any one of the above method embodiments of the present invention.

[0121] Among them, the storage medium is a computer storage medium. If the modules / units integrated in the device / terminal device are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such understanding, to implement all or part of the processes in the above-described embodiment methods of the present invention, it can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-described various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0122] The above is the preferred implementation manner of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements are also regarded as the protection scope of the present invention.

Claims

1. A teaching management method, characterized in that: include: Collect multi-source teaching data of preset time length; Extracting features from the multi-source teaching data to obtain a number of teaching feature vectors, and fusing the teaching feature vectors to generate a comprehensive feature matrix; Arrange the multi-source teaching data in chronological order, calculate the time feature vector of each preset time point, and generate time series features based on the time feature vector; wherein the time series features are used to describe the changing trends of students' learning behaviors, teachers' teaching effects and course content within the preset time length; Convert the multi-source teaching data into a fuzzy set, and generate a fuzzy output of the multi-source teaching data according to a preset fuzzy library rule and the time series characteristics; wherein the fuzzy output is a teaching evaluation result of the preset time length; Inputting the comprehensive feature matrix and the fuzzy output into a preset neural network model, so that the neural network model performs data prediction according to a preset teaching management decision-making target and outputs a teaching management decision-making result; A teaching management plan is generated according to the teaching management decision result.

2. A teaching management method as claimed in claim 1, characterized in that: The collecting of multi-source teaching data of a preset time length specifically includes: Acquire initial multi-source teaching data through a teaching management system; wherein the initial multi-source teaching data includes student behavior data, teacher behavior data and course content data; The corresponding preprocessing strategy is called according to the data type of the initial multi-source teaching data, and the initial multi-source teaching data is processed based on the preprocessing strategy to obtain the multi-source teaching data; wherein the multi-source teaching data includes a student behavior feature table, a teacher teaching feature table and a course content feature table.

3. A teaching management method as claimed in claim 2, characterized in that: The calling of a corresponding preprocessing strategy according to the data type of the initial multi-source teaching data, and performing data processing on the initial multi-source teaching data based on the preprocessing strategy to obtain the multi-source teaching data includes: The preprocessing strategies include numerical data preprocessing strategies, text data preprocessing strategies and image data preprocessing strategies; The numerical data preprocessing strategy is to normalize the numerical data in the initial multi-source teaching data to unify the numerical data into a preset numerical range; The text data preprocessing strategy is to perform word segmentation and word vectorization operations on the text data in the initial multi-source teaching data to convert the text data into a text feature vector; The image data preprocessing strategy is to extract key frames of the image data in the initial multi-source teaching data, extract visual features of the key frames based on a convolutional neural network model, and generate concentration features of the student behavior feature table based on the extracted visual features.

4. A teaching management method as claimed in claim 2, characterized in that: The feature extraction of the multi-source teaching data is performed to obtain a plurality of teaching feature vectors, and the teaching feature vectors are fused to generate a comprehensive feature matrix, including: Taking the column elements of each feature table as feature objects for feature extraction, the teaching feature vector is constructed based on the feature value of each feature object; wherein the teaching feature vector includes the student behavior feature vector, the teacher teaching feature vector and the course content feature vector; The student behavior feature vector, the teacher teaching feature vector and the course content feature vector are concatenated to obtain the comprehensive feature matrix.

5. A teaching management method as claimed in claim 1, characterized in that: The calculating of the time feature vector of each preset time point and generating the time series feature based on the time feature vector includes: Arranging the multi-source teaching data in chronological order, calculating the teaching feature vector of each preset time point and concatenating them to obtain the time feature vector; Arrange each of the time feature vectors in sequence to obtain time series data; A sliding window operation is performed on the time series data based on a window of a preset size to generate the time series features.

6. A teaching management method as claimed in claim 2, characterized in that: The method converts the multi-source teaching data into a fuzzy set, and generates a fuzzy output of the multi-source teaching data according to a preset fuzzy library rule and the time series feature, including: Defining a fuzzy output target according to the feature table types of the student behavior feature table and the teacher teaching feature table; wherein the fuzzy output target includes the student learning status and the teacher teaching effect; The column elements of the student behavior feature table and the teacher teaching feature table are respectively used as fuzzy input objects, and the fuzzy sets corresponding to the column elements are constructed based on the values ​​of the column elements in the time feature sequence; Defining a mapping relationship between each of the column elements in the feature table and a corresponding fuzzy output target, and generating a fuzzy output target of the fuzzy set based on the mapping relationship; The fuzzy output target is defuzzified to obtain the fuzzy output; wherein the fuzzy output includes a learning status score and a teaching effect score.

7. A teaching management method as claimed in claim 1, characterized in that: The step of inputting the comprehensive feature matrix and the fuzzy output into a preset neural network model so that the neural network model performs data analysis according to a preset teaching management decision-making target and outputs a teaching management decision-making result includes: The comprehensive feature matrix and the fuzzy output are spliced, and the obtained spliced ​​data is input into the preset neural network model; wherein the preset neural network model is trained based on historical multi-source teaching data; The preset neural network model is controlled to predict the spliced ​​data based on the preset teaching management decision-making objectives, and output the teaching management decision results; wherein the teaching management decision objectives include student class placement suggestions, teaching, teaching resource recommendations and teaching improvement suggestions.

8. A teaching management device, characterized in that: include: Data acquisition module, first feature module, second feature module, fuzzy processing module, data prediction module and solution generation module; The data acquisition module is used to collect multi-source teaching data of a preset time length; The first feature module is used to extract features from the multi-source teaching data to obtain a number of teaching feature vectors, and fuse the teaching feature vectors to generate a comprehensive feature matrix; The second feature module is used to arrange the multi-source teaching data in chronological order, calculate the time feature vector of each preset time point, and generate a time series feature based on the time feature vector; wherein the time series feature is used to describe the change trend of students' learning behavior, teachers' teaching effect and course content within the preset time length; The module processing module is used to convert the multi-source teaching data into a fuzzy set, and generate a fuzzy output of the multi-source teaching data according to a preset fuzzy library rule and the time series characteristics; wherein the fuzzy output is a teaching evaluation result of the preset time length; The data prediction module is used to input the comprehensive feature matrix and the fuzzy output into a preset neural network model, so that the neural network model performs data prediction according to a preset teaching management decision-making target and outputs a teaching management decision-making result; The scheme generating module is used to generate a teaching management scheme according to the teaching management decision result.

9. A terminal device, characterized in that: The teaching management method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the teaching management method according to any one of claims 1 to 7 when executing the computer program.

10. A storage medium, characterized in that: The storage medium includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute the teaching management method according to any one of claims 1 to 7.

Images