Intelligent seat arrangement method and system and readable storage medium

Abstract

The invention belongs to the field of education, and particularly provides an intelligent seat arrangement method and system and a readable storage medium, and the intelligent seat arrangement method comprises the steps: obtaining the data information of each student; seat information of each student is acquired; processing the data information; inputting the data information of each student and the corresponding seat information into a seat arrangement regression model trained in advance according to a convolutional self-attention network to obtain a fitness matrix of each student; setting a seat arrangement priority rule; and performing seat arrangement according to a seat arrangement priority rule and an optimal value in the fitness matrix of each student, and selecting a seat arrangement sequence with the maximum overall fitness value of the class as a final seat arrangement sequence. According to the method, the seat arrangement mode is associated with the score, seat arrangement is carried out according to the optimal fitness value, the highest score of the whole class is ensured, the average score of the class is improved, meanwhile, a priority rule is set during seat arrangement, and the seat arrangement requirements of different students are met.

Description

Technical Field

[0001] The present invention relates to the field of education, and in particular to an intelligent seating arrangement method, system and readable storage medium. Background Art

[0002] At present, the school's seating arrangement system is mainly divided into examination seating and regular classroom seating. For examination seating, the teaching administrator selects all student data and then sorts them by student number. This method is suitable for student examinations.

[0003] Regular classroom seating is intelligent, and this seating system includes a data input module, a data sorting module, an intelligent seating module, a display module, a cloud server, a storage module, a login module, and a verification module. The input module is used to input student and classroom information. Student information includes features such as name, photo, height, vision, gender, and seat preference. The data sorting module is used to sort student and classroom information, and intelligently sort student features according to seating preferences. The display module is used to display the results of intelligent seating and associate student information with seats. The cloud server provides cloud operation services. The storage module is used to store relevant data. The login module is used for user login. The verification module is used to verify the user's login qualifications. If the verification is successful, the user successfully logs into the cloud server.

[0004] The existing technology has two main shortcomings: (1) The seating arrangement system has a single function. Most of the users are class teachers, and other subject teachers are not involved. Therefore, there is a disconnect between the seats and the students' learning situation. (2) The seats are disconnected from the students' learning situation. Students' learning will be affected by the arrangement of seats, and for students, whether the current seat arrangement is conducive to their learning is data that cannot be observed by the existing seating arrangement system.

[0005] Therefore, the current seating arrangement system still cannot meet the personalized needs of teachers, students and parents, and the intelligent dynamic functions of the seating arrangement system are insufficient. Summary of the invention

[0006] The object of the present invention is to provide an intelligent seating arrangement method, system and readable storage medium.

[0007] In order to solve the above technical problems, the present invention provides an intelligent seating arrangement method, which comprises:

[0008] Get data information of each student;

[0009] Get each student's seat information

[0010] Processing data information;

[0011] Input each student’s data information and corresponding seat information into the seating regression model pre-trained based on the convolutional self-attention network to obtain the fitness matrix of each student;

[0012] Set seating priority rules;

[0013] The seats will be arranged according to the seating priority rules and the optimal value in each student's fitness matrix, and the seating order with the largest overall fitness value of the class will be selected as the final seating order.

[0014] Furthermore, the student's data information includes:

[0015] Basic information of students, including: grade, gender, age, height, weight, personality, classroom behavior and learning style;

[0016] Students’ academic performance.

[0017] Furthermore, the step of processing the data information includes:

[0018] The students’ basic information and academic performance are standardized, that is, the minimum-maximum method is used to scale the range of the students’ basic information and academic performance to [0,1].

[0019] Furthermore, the steps for constructing the seating regression model based on the convolutional self-attention network training are as follows:

[0020] Get the historical performance data of each class;

[0021] Calculate the average score for each class's history grade;

[0022] Collect classes ranked within a preset range to construct training data;

[0023] The training data is input into the convolutional self-attention network structure to train the seat ranking regression model.

[0024] Furthermore, the convolutional self-attention network structure includes: a convolutional layer, a self-attention layer and a fully connected layer;

[0025] The convolutional layer is used to extract local features of the input data;

[0026] The self-attention layer is used to capture the relationship between input data;

[0027] The fully connected layer maps the features of the previous layer to the output space, and obtains the final regression result through nonlinear transformations of multiple hidden layers.

[0028] Furthermore, the convolution introduces residual connections in each layer of the self-attention network structure, thereby forming a residual between the input and output of each layer.

[0029] Furthermore, the seating priority rule is academic performance priority, independent learning style priority, cooperative learning style priority or height priority.

[0030] Furthermore, the step of arranging seats according to the seating priority rule and the optimal value in the fitness matrix of each student, and selecting the seating order with the largest overall fitness value of the class as the final seating order includes:

[0031] Initialization: According to the fitness matrix of each student, obtain the suitable seats for all students and sort them by fitness;

[0032] Priority selection: prioritize students according to set priority rules;

[0033] Assign seats: Check the most suitable seat for each student based on priority, combine proximity and social preferences, complete preliminary seating, and obtain the overall fitness value;

[0034] Adjustment optimization: adjust the positions of students with the same priority and obtain the adjusted overall fitness value;

[0035] Among them, the neighboring relationship includes:

[0036] Position matrix: Model the seats and determine the adjacent seats for each seat;

[0037] Adjacent weight: Set weights for adjacent seats based on actual needs;

[0038] Quantify students' social preferences, including:

[0039] Social network graph: Build a social network among students, where each student is a node and the weight of the edge represents the strength of their social relationship;

[0040] Intimacy score: Calculate an intimacy score for each pair of students based on their personal information and interests.

[0041] The present invention also provides an intelligent seating arrangement system, the system comprising:

[0042] A first acquisition module is configured to acquire data information of each student;

[0043] The second acquisition module is configured to obtain the seat information of each student

[0044] A data processing module is configured to process data information;

[0045] The fitness matrix acquisition module is configured to input the data information of each student and the corresponding seat information into the seating regression model pre-trained according to the convolutional self-attention network to obtain the fitness matrix of each student;

[0046] A setting module, configured to set a seating priority rule;

[0047] The seating module is configured to arrange seats according to the seating priority rules and the optimal value in the fitness matrix of each student, and select the seating order with the largest overall fitness value of the class as the final seating order.

[0048] The present invention also provides a computer-readable storage medium, wherein at least one instruction is stored in the computer-readable storage medium, and wherein when the instruction is executed by a processor, the intelligent seating arrangement method as described above is implemented.

[0049] The beneficial effect of the present invention is that the present invention provides an intelligent seating method, system and readable storage medium, wherein the intelligent seating method includes: obtaining data information of each student; obtaining seat information of each student; processing the data information; inputting the data information of each student and the corresponding seat information into a seating regression model pre-trained according to a convolutional self-attention network to obtain a fitness matrix for each student; setting seating priority rules; seating is performed according to the seating priority rules and the optimal value in the fitness matrix of each student, and the seating order with the largest overall fitness value of the class is selected as the final seating order. The seating method is associated with the grades, and the seats are arranged according to the optimal value of the fitness to ensure that the overall grade of the class is the highest, thereby improving the average grade of the class. At the same time, priority rules are set when seating, so as to meet the seating needs of different individual students. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0051] Figure 1 It is a flow chart of the intelligent seating arrangement method provided by an embodiment of the present invention.

[0052] Figure 2 It is a principle block diagram of the intelligent seating arrangement system provided in an embodiment of the present invention.

[0053] Figure 3 It is a partial principle block diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0055] Example 1

[0056] See also Figure 1 This embodiment adopts an intelligent seating arrangement method to associate the seating arrangement with the grades, and arranges the seats according to the optimal value of fitness to ensure that the overall grade of the class is the highest, thereby improving the average grade of the class. At the same time, priority rules are set when arranging seats to meet the seating needs of different individual students.

[0057] Specifically, the method comprises the following steps:

[0058] S110: Obtain data information of each student.

[0059] Specifically, the student data information includes: basic information of the student and the student's academic performance. It should be noted that the basic information of the student includes: grade, gender, age, height, weight, personality, classroom behavior and learning style.

[0060] Among them, personality is divided into introversion and extroversion, classroom behavior includes whether it is active and whether the attention is focused, and learning style includes independent learning and cooperative learning.

[0061] Academic performance: Based on the different requirements of each grade, the academic performance of major subjects is collected, such as Chinese, mathematics and English scores of K12 students.

[0062] S120: Obtain the seat information of each student.

[0063] The seat information is represented by a matrix.

[0064] For example, a class has 45 students, which can be represented as follows:

[0065] [1,2,3,4,5],

[0066] [6,7,8,9,10],

[0067] [11,12,13,14,15],

[0068] [16,17,18,19,20],

[0069] [21,22,23,24,25],

[0070] [26,27,28,29,30],

[0071] [31,32,33,34,35],

[0072] [36,37,38,39,40],

[0073] [41,42,43,44,45]

[0074] It should be noted that the above matrix represents the seating arrangement of the classroom, and the rows and columns represent the specific positions of the students. In this example, there are 9 rows and 5 columns, totaling 45 seats. The actual seating arrangement can be adjusted according to the design of the classroom. For example, some classrooms may have 6 rows and 8 columns, 5 rows and 9 columns, etc. The specific forms vary, so the seating matrix needs to be designed according to the actual situation.

[0075] S130: Process the data information.

[0076] Specifically, the basic information and academic performance of students are standardized, that is, the minimum-maximum method is used to scale the range of the basic information and academic performance of students to [0,1].

[0077] For a given feature F, its maximum and minimum values ​​are denoted as F max and F min . Each value F of feature F i The normalized value Vi of is calculated as follows:

[0078]

[0079] S140: Input the data information of each student and the corresponding seat information into the seating regression model pre-trained based on the convolutional self-attention network to obtain the fitness matrix of each student.

[0080] For example, if there are 40 students in a new class and the student seats are arranged in 4 rows and 10 columns, the model's output for the first student is as follows: [

[0082] [0.8,0.6,0.7,0.9,0.5,0.4,0.3,0.2,0.7,0.6], / / first row

[0083] [0.5,0.8,0.7,0.6,0.9,0.4,0.3,0.8,0.7,0.6], / / second line

[0084] [0.4,0.5,0.8,0.7,0.6,0.9,0.5,0.4,0.3,0.5], / / The third line

[0085] [0.3,0.2,0.4,0.5,0.6,0.7,0.8,0.9,0.4,0.3] / / The fourth row ]

[0087] Note: Each element indicates the recommended suitability of the position for a specific student. For example, the first position (0.8) in the first row indicates that the first student's suitability at seat (1,1) is 0.8, which may be a good match for their grades, study habits, etc. Seats with lower values ​​(such as 0.2 or 0.3) may not be recommended seats, and you can consider arranging these students to other seats that are more suitable for them.

[0088] The steps for constructing the seat ranking regression model based on convolutional self-attention network training are as follows:

[0089] S141: Obtain historical performance data for each class.

[0090] S142: Calculate the average history score for each class.

[0091] S143: Collecting classes ranked within a preset range to construct training data.

[0092] Specifically, the data information, academic performance and seat information of all students in the class with the top 30 average scores are collected and standardized, and all values ​​are scaled to [0,1]. Records with scores above the average in the training data are marked as 1, and records with scores below the average are marked as 0.

[0093] S144: Input the training data into the convolutional self-attention network structure to train the seat ranking regression model.

[0094] Specifically, the convolutional self-attention network structure includes: a convolutional layer, a self-attention layer and a fully connected layer.

[0095] The convolution layer is used to extract local features of the input data. The input data is convolved through multiple convolution kernels to generate a series of high-dimensional feature maps. These feature maps capture the feature information of different levels of the input data, thus helping to improve the model's ability to represent the input data. Combining convolution and self-attention mechanisms, the convolutional self-attention network can capture both local and global features at the same time, thereby improving the model's expressive power.

[0096] The self-attention layer is used to capture the relationship between input data. It can learn the importance weights of different positions in the input data and adjust the representation of the input data according to these weights. Specifically, the self-attention layer calculates the attention weight matrix of the input data and multiplies it with the original input data to obtain a weighted representation. This weighted representation can better capture the dependencies between input data, thereby improving the expressiveness and reasoning effect of the model. The fully connected layer is used to convert high-dimensional features into outputs predicted by the model. The self-attention layer can dynamically adjust the weight of each feature according to the context of the input feature, providing an adaptive feature weighting method and improving the modeling of dependencies between surrounding elements.

[0097] The fully connected layer maps the features of the previous layer to the output space, and obtains the final regression result through nonlinear transformations of multiple hidden layers.

[0098] In this embodiment, the convolution introduces a residual connection in each layer of the self-attention network structure, thereby forming a residual between the input and output of each layer.

[0099] In deep networks, due to the large number of layers, the process of calculating the gradient during back propagation may cause the gradient to become smaller layer by layer and eventually approach zero, which is called the vanishing gradient problem. The vanishing gradient will cause the weights of some layers to be almost not updated, thus affecting the learning ability of the entire model. This application introduces residual connections between each layer, which helps to alleviate the vanishing gradient problem in deep networks, allowing information to be effectively propagated in the network and improving the training stability and prediction accuracy of the model.

[0100] Introducing residual connections in the Convolutional Self-Attention Network structure mainly involves making appropriate connections between the convolutional layer, the self-attention layer, and the fully connected layer to form residuals between the input and output of each layer. These connections can effectively alleviate the gradient vanishing problem, enhance information flow, and thus improve the learning ability of the model. The following is a specific implementation method:

[0101] (1) Introducing residual connections in convolutional layers

[0102] In a convolutional layer, the input (X) is added to the output (Y) after the convolution operation:

[0103] #Suppose there is an input X and a convolutional layer conv

[0104] Y = conv(X) #convolution operation

[0105] output=Y+X#residual connection

[0106] output = ReLU (output) # through activation function

[0107] (2) Introducing residual connections in the self-attention layer

[0108] In the self-attention layer, the attention representation is first calculated, and then the input and output are added:

[0109] #Suppose there is an input X and a self-attention layer self_attention

[0110] Z = self_attention(X) #Self-attention operation

[0111] output=Z+X#residual connection

[0112] output = ReLU (output) # through activation function

[0113] (3) Introducing residual connections in the fully connected layer

[0114] In the fully connected layer, the same approach is used to add the input (X) to the fully connected layer output (W):

[0115] #Assume there is an input X and a fully connected layer fc

[0116] W = fc(X) # Fully connected operation

[0117] output=W+X#residual connection

[0118] output = ReLU (output) # through activation function

[0119] (4) Overall connection example

[0120] Combining all of the above, the residual connection in the entire convolutional self-attention network can be divided into multiple modules, which may be as follows:

[0121] defresidual_block(input_tensor):

[0122] #Convolutional layer

[0123] conv_output=conv_layer(input_tensor)

[0124] #Residual connection

[0125] conv_residual=conv_output+input_tensor

[0126] conv_residual=ReLU(conv_residual)

[0127] #Self-attention layer

[0128] attention_output=self_attention(conv_residual)

[0129] #Residual connection

[0130] attention_residual=attention_output+conv_residual

[0131] attention_residual=ReLU(attention_residual)

[0132] #Fully connected layer

[0133] fc_output=fully_connected_layer(attention_residual)

[0134] #Residual connection

[0135] final_output=fc_output+attention_residual

[0136] returnfinal_output

[0137] S150: Setting seating priority rules.

[0138] Specifically, the seating priority rules are academic performance priority, independent learning style priority, cooperative learning style priority, or height priority.

[0139] S160: Seating is performed according to the seating priority rule and the optimal value in each student's fitness matrix, and the seating order with the largest overall class fitness value is selected as the final seating order.

[0140] Specifically, step S160 includes:

[0141] Initialization: According to the fitness matrix of each student, get the suitable seats for all students and sort them by fitness.

[0142] Priority selection: Prioritize students based on set priority rules.

[0143] Assign seats: For each student based on priority, check the most suitable seat, combine the proximity and social preferences, complete the preliminary seating, and obtain the overall fitness value. Specifically, for each student, check the most suitable seat (obtained from the fitness matrix). If a seat is already occupied by other students with higher priority, check the second best seat for that student. This process continues until a seat is assigned to each student. During the selection process, if a student has a strong requirement for his or her social preference and his or her suitable seat is not occupied, give priority to meeting his or her social needs. If a conflict occurs, you can choose to maintain contact between adjacent students to enhance class cohesion.

[0144] Adjustment optimization: Adjust the positions of students with the same priority and obtain the adjusted overall fitness value.

[0145] At the same time, after the seating arrangement is completed, feedback from students and teachers is collected regularly, and continuous optimization is carried out based on classroom atmosphere, learning outcomes, etc. The fitness matrix and social preference model are constantly updated to adapt to new student characteristics or changes in demand.

[0146] Among them, the neighboring relationship includes:

[0147] Position matrix: Model the seats and determine the adjacent seats of each seat; for example, for each element S(i,j) in the seat matrix S, its adjacent positions can be S(i±1,j) and S(i,j±1), paying attention to the boundary conditions.

[0148] Proximity weight: Setting weights for adjacent seats based on actual needs will affect the interaction between students. For example, if two students have similar grades and complementary personalities, they can be given a higher weight.

[0149] Quantify students' social preferences, including:

[0150] Social network graph: Build a social network among students, with each student as a node and the edge weight representing the strength of their social relationship.

[0151] Intimacy score: Calculate an intimacy score for each pair of students based on their personal information and interests.

[0152] In this embodiment, the intelligent seating arrangement method further includes:

[0153] Export data and export class seating charts based on time, class and other conditions.

[0154] Data analysis, record and analyze students’ learning performance after seating arrangement, and form a weekly classroom learning summary.

[0155] Example 2

[0156] See also Figure 2 This embodiment also provides an intelligent seating arrangement system, which includes: a first acquisition module, a second acquisition module, a data processing module, a fitness matrix acquisition module, a setting module and a seating arrangement module.

[0157] In this embodiment, the first acquisition module is configured to acquire data information of each student; specifically, the data information of the student includes: basic information of the student and the student's academic performance. It should be noted that the basic information of the student includes: grade, gender, age, height, weight, personality, classroom behavior and learning style.

[0158] Among them, personality is divided into introversion and extroversion, classroom behavior includes whether it is active and whether the attention is focused, and learning style includes independent learning and cooperative learning.

[0159] Academic performance: Based on the different requirements of each grade, the academic performance of major subjects is collected, such as Chinese, mathematics and English scores of K12 students.

[0160] In this embodiment, the second acquisition module is configured to obtain the seat information of each student.

[0161] The seat information is represented by a matrix.

[0162] For example, a class has 45 students, which can be represented as follows:

[0163] [1,2,3,4,5],

[0164] [6,7,8,9,10],

[0165] [11,12,13,14,15],

[0166] [16,17,18,19,20],

[0167] [21,22,23,24,25],

[0168] [26,27,28,29,30],

[0169] [31,32,33,34,35],

[0170] [36,37,38,39,40],

[0171] [41,42,43,44,45]

[0172] It should be noted that the above matrix represents the seating arrangement of the classroom, and the rows and columns represent the specific positions of the students. In this example, there are 9 rows and 5 columns, totaling 45 seats. The actual seating arrangement can be adjusted according to the design of the classroom. For example, some classrooms may have 6 rows and 8 columns, 5 rows and 9 columns, etc. The specific forms vary, so the seating matrix needs to be designed according to the actual situation.

[0173] In this embodiment, the data processing module is configured to process data information; specifically, the basic information and academic performance of students are standardized, that is, the minimum-maximum method is used to scale the range of the basic information and academic performance of students to [0,1].

[0174] For a given feature F, its maximum and minimum values ​​are denoted as F max and F min . Each value F of feature F i The normalized value V i The calculation is as follows:

[0175]

[0176] The fitness matrix acquisition module is configured to input the data information of each student and the corresponding seat information into a seating regression model pre-trained based on a convolutional self-attention network to obtain the fitness matrix of each student.

[0177] For example, if there are 40 students in a new class and the seats are arranged in 4 rows and 10 columns, the model's output for the first student is as follows: [

[0179] [0.8,0.6,0.7,0.9,0.5,0.4,0.3,0.2,0.7,0.6], / / first row

[0180] [0.5,0.8,0.7,0.6,0.9,0.4,0.3,0.8,0.7,0.6], / / second line

[0181] [0.4,0.5,0.8,0.7,0.6,0.9,0.5,0.4,0.3,0.5], / / The third line

[0182] [0.3,0.2,0.4,0.5,0.6,0.7,0.8,0.9,0.4,0.3] / / The fourth row ]

[0184] Note: Each element indicates the recommended suitability of the position for a specific student. For example, the first position (0.8) in the first row indicates that the first student's suitability at seat (1,1) is 0.8, which may be a good match for their grades, study habits, etc. Seats with lower values ​​(such as 0.2 or 0.3) may not be recommended seats, and you can consider arranging these students to other seats that are more suitable for them.

[0185] The steps for constructing the seat ranking regression model based on convolutional self-attention network training are as follows:

[0186] S141: Obtain historical performance data for each class.

[0187] S142: Calculate the average history score for each class.

[0188] S143: Collecting classes ranked within a preset range to construct training data.

[0189] Specifically, the data information, academic performance and seat information of all students in the class with the top 30 average scores are collected and standardized, and all values ​​are scaled to [0,1]. Records with scores above the average in the training data are marked as 1, and records with scores below the average are marked as 0.

[0190] S144: Input the training data into the convolutional self-attention network structure to train the seat ranking regression model.

[0191] Specifically, the convolutional self-attention network structure includes: a convolutional layer, a self-attention layer and a fully connected layer.

[0192] The convolution layer is used to extract local features of the input data. The input data is convolved through multiple convolution kernels to generate a series of high-dimensional feature maps. These feature maps capture the feature information of different levels of the input data, thus helping to improve the model's ability to represent the input data. Combining convolution and self-attention mechanisms, the convolutional self-attention network can capture both local and global features at the same time, thereby improving the model's expressive power.

[0193] The self-attention layer is used to capture the relationship between input data. It can learn the importance weights of different positions in the input data and adjust the representation of the input data according to these weights. Specifically, the self-attention layer calculates the attention weight matrix of the input data and multiplies it with the original input data to obtain a weighted representation. This weighted representation can better capture the dependencies between input data, thereby improving the expressiveness and reasoning effect of the model. The fully connected layer is used to convert high-dimensional features into outputs predicted by the model. The self-attention layer can dynamically adjust the weight of each feature according to the context of the input feature, providing an adaptive feature weighting method and improving the modeling of dependencies between surrounding elements.

[0194] The fully connected layer maps the features of the previous layer to the output space, and obtains the final regression result through nonlinear transformations of multiple hidden layers.

[0195] In this embodiment, the convolution introduces a residual connection in each layer of the self-attention network structure, thereby forming a residual between the input and output of each layer.

[0196] In deep networks, due to the large number of layers, the process of calculating the gradient during back propagation may cause the gradient to become smaller layer by layer and eventually approach zero, which is called the vanishing gradient problem. The vanishing gradient will cause the weights of some layers to be almost not updated, thus affecting the learning ability of the entire model. This application introduces residual connections between each layer, which helps to alleviate the vanishing gradient problem in deep networks, allowing information to be effectively propagated in the network and improving the training stability and prediction accuracy of the model.

[0197] Introducing residual connections in the Convolutional Self-Attention Network structure mainly involves making appropriate connections between the convolutional layer, the self-attention layer, and the fully connected layer to form residuals between the input and output of each layer. These connections can effectively alleviate the gradient vanishing problem, enhance information flow, and thus improve the learning ability of the model. The following is a specific implementation method:

[0198] (1) Introducing residual connections in convolutional layers

[0199] In a convolutional layer, the input (X) is added to the output (Y) after the convolution operation:

[0200] #Suppose there is an input X and a convolutional layer conv

[0201] Y = conv(X) #convolution operation

[0202] output=Y+X#residual connection

[0203] output = ReLU (output) # through activation function

[0204] (2) Introducing residual connections in the self-attention layer

[0205] In the self-attention layer, the attention representation is first calculated, and then the input and output are added:

[0206] #Suppose there is an input X and a self-attention layer self_attention

[0207] Z = self_attention(X) #Self-attention operation

[0208] output=Z+X#residual connection

[0209] output = ReLU (output) # through activation function

[0210] (3) Introducing residual connections in the fully connected layer

[0211] In the fully connected layer, the same approach is used to add the input (X) to the fully connected layer output (W):

[0212] #Assume there is an input X and a fully connected layer fc

[0213] W = fc(X) # Fully connected operation

[0214] output=W+X#residual connection

[0215] output = ReLU (output) # through activation function

[0216] (4) Overall connection example

[0217] Combining all of the above, the residual connection in the entire convolutional self-attention network can be divided into multiple modules, which may be as follows:

[0218] defresidual_block(input_tensor):

[0219] #Convolutional layer

[0220] conv_output=conv_layer(input_tensor)

[0221] #Residual connection

[0222] conv_residual=conv_output+input_tensor

[0223] conv_residual=ReLU(conv_residual)

[0224] #Self-attention layer

[0225] attention_output=self_attention(conv_residual)

[0226] #Residual connection

[0227] attention_residual=attention_output+conv_residual

[0228] attention_residual=ReLU(attention_residual)

[0229] #Fully connected layer

[0230] fc_output=fully_connected_layer(attention_residual)

[0231] #Residual connection

[0232] final_output=fc_output+attention_residual

[0233] returnfinal_output

[0234] In this embodiment, the setting module is configured to set a seating priority rule; specifically, the seating priority rule is academic performance priority, independent learning style priority, cooperative learning style priority, or height priority.

[0235] In this embodiment, the seating arrangement module is configured to arrange seats according to the seating priority rules and the optimal value in the fitness matrix of each student, and select the seating order with the largest overall fitness value of the class as the final seating order.

[0236] Specifically, step S160 includes:

[0237] Initialization: According to the fitness matrix of each student, get the suitable seats for all students and sort them by fitness.

[0238] Priority selection: Prioritize students based on set priority rules.

[0239] Assign seats: For each student based on priority, check the most suitable seat, combine the proximity and social preferences, complete the preliminary seating, and obtain the overall fitness value. Specifically, for each student, check the most suitable seat (obtained from the fitness matrix). If a seat is already occupied by other students with higher priority, check the second best seat for that student. This process continues until a seat is assigned to each student. During the selection process, if a student has a strong requirement for his or her social preference and his or her suitable seat is not occupied, give priority to meeting his or her social needs. If a conflict occurs, you can choose to maintain contact between adjacent students to enhance class cohesion.

[0240] Adjustment optimization: Adjust the positions of students with the same priority and obtain the adjusted overall fitness value.

[0241] At the same time, after the seating arrangement is completed, feedback from students and teachers is collected regularly, and continuous optimization is carried out based on classroom atmosphere, learning outcomes, etc. The fitness matrix and social preference model are constantly updated to adapt to new student characteristics or changes in demand.

[0242] Among them, the neighboring relationship includes:

[0243] Position matrix: Model the seats and determine the adjacent seats of each seat; for example, for each element S(i,j) in the seat matrix S, its adjacent positions can be S(i±1,j) and S(i,j±1), paying attention to the boundary conditions.

[0244] Proximity weight: Setting weights for adjacent seats based on actual needs will affect the interaction between students. For example, if two students have similar grades and complementary personalities, they can be given a higher weight.

[0245] Quantify students' social preferences, including:

[0246] Social network graph: Build a social network among students, with each student as a node and the edge weight representing the strength of their social relationship.

[0247] Intimacy score: Calculate an intimacy score for each pair of students based on their personal information and interests.

[0248] In this embodiment, the intelligent seating arrangement system also includes a display module, which is set on each seat to facilitate interaction between teachers and students. Students can press buttons to interact with teachers in real time on different issues. The seating chart after seating arrangement is intelligently displayed and exported. By setting a display module to display the seating arrangement results, it is convenient for students and teachers to clearly understand the basic information of students in their class and intuitively display the seating arrangement of students. It can also be linked to the parent's mobile phone so that the parent can also know the class student seating chart.

[0249] In this embodiment, the intelligent seating arrangement system also includes a file generation module, which is configured to generate student seating arrangements per week, month, and semester, and calculate the number of student seat rotations, thereby ensuring that students are in different positions in the classroom to the greatest extent possible and avoiding missing students who have no seats.

[0250] In this embodiment, the intelligent seating arrangement system also includes a feedback module, which is configured to provide efficient and rapid real-time feedback on students' learning effects by recording students' classroom performance and teachers' real-time evaluation of students.

[0251] In this embodiment, the intelligent seating system further includes a dynamic adjustment module configured to dynamically adjust seats according to changes in vision.

[0252] Example 3

[0253] This embodiment provides a computer-readable storage medium, in which at least one instruction is stored. When the instruction is executed by a processor, the intelligent seating arrangement method provided in Embodiment 1 is implemented.

[0254] The intelligent seating arrangement method includes: obtaining the data information of each student; obtaining the seat information of each student; processing the data information; inputting the data information of each student and the corresponding seat information into the seating regression model pre-trained based on the convolutional self-attention network to obtain the fitness matrix of each student; setting the seating priority rules; seating is arranged according to the seating priority rules and the optimal value in the fitness matrix of each student, and the seating order with the largest overall fitness value of the class is selected as the final seating order. The seating arrangement method is associated with the grades, and the seats are arranged according to the optimal value of the fitness to ensure that the overall grade of the class is the highest, thereby improving the average grade of the class. At the same time, priority rules are set when seating, so as to meet the seating needs of different individual students.

[0255] Example 4

[0256] See also Figure 3 This embodiment provides an electronic device, including: a memory 502 and a processor 501; the memory 502 stores at least one program instruction; the processor 501 loads and executes the at least one program instruction to implement the cross-project instant software defect prediction method provided in Example 1.

[0257] The memory 502 and the processor 501 are connected in a bus manner, and the bus may include any number of interconnected buses and bridges, and the bus connects various circuits of one or more processors 501 and the memory 502 together. The bus can also connect various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be one element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices on a transmission medium. The data processed by the processor 501 is transmitted on a wireless medium via an antenna, and further, the antenna also receives data and transmits the data to the processor 501.

[0258] The processor 501 is responsible for managing the bus and general processing, and can also provide various functions, including timing, peripheral interfaces, voltage regulation, power management and other control functions. The memory 502 can be used to store data used by the processor 501 when performing operations.

[0259] In summary, the present invention provides an intelligent seating method, system and readable storage medium, wherein the intelligent seating method includes: obtaining data information of each student; obtaining seat information of each student; processing the data information; inputting the data information of each student and the corresponding seat information into a seating regression model pre-trained based on a convolutional self-attention network to obtain a fitness matrix for each student; setting seating priority rules; seating is performed based on the seating priority rules and the optimal value in the fitness matrix of each student, and the seating order with the largest overall fitness value of the class is selected as the final seating order. The seating method is associated with the grades, and the seats are arranged according to the optimal value of the fitness to ensure that the overall grade of the class is the highest, thereby improving the average grade of the class. At the same time, priority rules are set when seating, so as to meet the seating needs of different individual students.

[0260] In the description of the embodiments of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0261] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0262] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0263] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0264] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0265] Based on the above ideal embodiments of the present invention, the relevant staff can make various changes and modifications without departing from the technical concept of the present invention through the above description. The technical scope of the present invention is not limited to the contents of the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. An intelligent seating arrangement method, characterized in that: The method comprises: Get data information of each student; Get each student's seat information; Processing data information; Input each student’s data information and corresponding seat information into the seating regression model pre-trained based on the convolutional self-attention network to obtain the fitness matrix of each student; Set seating priority rules; The seats will be arranged according to the seating priority rules and the optimal value in each student's fitness matrix, and the seating order with the largest overall fitness value of the class will be selected as the final seating order.

2. The intelligent seating arrangement method according to claim 1, characterized in that: The student data information includes: Basic information of students, including: grade, gender, age, height, weight, personality, classroom behavior and learning style; Students’ academic performance.

3. The intelligent seating arrangement method according to claim 2, characterized in that: The step of processing the data information comprises: The students’ basic information and academic performance are standardized, that is, the minimum-maximum method is used to scale the range of the students’ basic information and academic performance to [0,1].

4. The intelligent seating arrangement method according to claim 1, characterized in that: The steps for constructing the seat ranking regression model based on convolutional self-attention network training are as follows: Get the historical performance data of each class; Calculate the average score for each class's history grade; Collect classes ranked within a preset range to construct training data; The training data is input into the convolutional self-attention network to train the seat ranking regression model.

5. The intelligent seating arrangement method according to claim 4, characterized in that: The convolutional self-attention network structure includes: a convolutional layer, a self-attention layer and a fully connected layer; The convolutional layer is used to extract local features of the input data; The self-attention layer is used to capture the relationship between input data; The fully connected layer maps the features of the previous layer to the output space, and obtains the final regression result through nonlinear transformations of multiple hidden layers.

6. The intelligent seating arrangement method according to claim 5, characterized in that: The convolution introduces residual connections in each layer of the self-attention network structure, thereby forming a residual between the input and output of each layer.

7. The intelligent seating arrangement method according to claim 1, characterized in that: The seating priority rules are academic performance priority, independent learning style priority, cooperative learning style priority or height priority.

8. The intelligent seating arrangement method according to claim 7, characterized in that: The steps of arranging seats according to the seating priority rule and the optimal value in the fitness matrix of each student and selecting the seating order with the largest overall fitness value of the class as the final seating order include: Initialization: According to the fitness matrix of each student, obtain the suitable seats for all students and sort them by fitness; Priority selection: prioritize students according to set priority rules; Assign seats: Check the most suitable seat for each student based on priority, combine proximity and social preferences, complete preliminary seating, and obtain the overall fitness value; Adjustment optimization: adjust the positions of students with the same priority and obtain the adjusted overall fitness value; Among them, the neighboring relationship includes: Position matrix: Model the seats and determine the adjacent seats for each seat; Adjacent weight: Set weights for adjacent seats based on actual needs; Quantify students' social preferences, including: Social network graph: Build a social network among students, where each student is a node and the weight of the edge represents the strength of their social relationship; Intimacy score: Calculate an intimacy score for each pair of students based on their personal information and interests.

9. An intelligent seating arrangement system, characterized in that: The system comprises: A first acquisition module is configured to acquire data information of each student; The second acquisition module is configured to obtain the seat information of each student A data processing module is configured to process data information; The fitness matrix acquisition module is configured to input the data information of each student and the corresponding seat information into the seating regression model pre-trained according to the convolutional self-attention network to obtain the fitness matrix of each student; A setting module, configured to set a seating priority rule; The seating module is configured to arrange seats according to the seating priority rules and the optimal value in the fitness matrix of each student, and select the seating order with the largest overall fitness value of the class as the final seating order.

10. A computer-readable storage medium, wherein at least one instruction is stored in the computer-readable storage medium, characterized in that: When the instructions are executed by a processor, the intelligent seating arrangement method described in any one of claims 1 to 8 is implemented.

Images